Jensen v. Shinn

Full title: Shawn JENSEN, et al., Plaintiffs, v. David SHINN, et al., Defendants.

Court: United States District Court, D. Arizona.

Date published: Jun 30, 2022

609 F. Supp. 3d 789 (D. Ariz. 2022)

Opinion

No. CV-12-00601-PHX-ROS

2022-06-30

Shawn JENSEN, et al., Plaintiffs, v. David SHINN, et al., Defendants.

Roslyn O. Silver, Senior United States District Judge

ORDER

Roslyn O. Silver, Senior United States District Judge

Prisoners held in government custody have a basic right to health care and humane conditions of confinement protected by the Eighth Amendment prohibiting cruel and unusual punishment.¹ These Eighth Amendment rights do not require the highest quality of health care, the community standard of health care, or the most pleasant accommodations possible. But the health care and conditions of confinement must reflect basic common decency and a recognition of the dignity the government must accord all human beings. This suit was brought by prisoners in the custody of the Arizona Department of Corrections, Rehabilitation, and Reentry ("ADCRR"), alleging systemic violations in providing minimally sufficient health care and minimally humane conditions in maximum custody units.

1 See Estelle v. Gamble , 429 U.S. 97, 103, 97 S.Ct. 285, 50 L.Ed.2d 251 (1976) (holding government has "obligation to provide medical care for those whom it is punishing by incarceration").

While this case involves an unusually large amount of evidence, there are only two basic questions. Are Defendants violating the constitutional rights of Arizona's prisoners through the existing medical and mental health care system? And are Defendants violating the constitutional rights of a subset of Arizona's prisoners by almost round-the-clock confinement in their cells? The answer is yes to both questions.

Defendants have failed to provide, and continue to refuse to provide, a constitutionally adequate medical care and mental health care system for all prisoners. Defendants’ health care system is plainly grossly inadequate. Defendants have been aware of their failures for years and Defendants have refused to take necessary actions to remedy the failures. Defendants’ years of inaction, despite Court intervention and imposition of monetary sanctions, establish Defendants are acting with deliberate indifference to the substantial risk of serious harm posed by the lack of adequate medical and mental health care affecting all prisoners.

Further, Defendants keep thousands of prisoners in restrictive housing units where they are not provided adequate nutrition, nor are they provided meaningful out-of-cell time for exercise or social interaction. Defendants’ treatment of prisoners in restrictive housing units results in the deprivation of basic human needs. For years, Defendants have known of the deficiencies, highlighted by Court intervention and direction, and refused to take meaningful remedial actions. Therefore, Defendants are acting with deliberate indifference to the substantial risk of serious harm posed to prisoners in restrictive housing units.

Based on the 15-day bench trial that commenced on November 1, 2021, as well as the parties post-trial briefing, the Court makes the following Findings of Fact and Conclusions of Law. Fed. R. Civ. P. 52(a)(1).

BACKGROUND

There are over 4,300 docket entries in this case, reflecting a wide variety of motions, hearings, and orders. A full recital of that background would be lengthy and would serve little purpose. However, some background information is necessary to place in context the events leading up to trial and to demonstrate Defendants’ recognition, for over a decade, that they are systematically violating the constitutional rights of the prisoners.

I. The Complaint and Class Certification

ADCRR operates ten prison complexes across Arizona. In 2012, a group of prisoners filed this class action, seeking only injunctive relief and challenging the provision of medical, dental, and mental health care throughout the prison system. Plaintiffs also challenged the conditions of confinement in "isolation units," where prisoners are confined to their cells for 22 or more hours each day. Plaintiffs named as defendants the Director of Arizona's prisons, currently David Shinn, and the Assistant Director of ADCRR's Medical Services Contract Monitoring Bureau, currently Larry Gann.

The complaint presented five claims for relief on behalf of what would subsequently be identified as a main class and one subclass. The class consisted of all prisoners at ADCRR's ten complexes. According to the complaint, Defendants were subjecting class members to a substantial risk of serious harm and injury from constitutionally deficient health care, dental care, and mental health care. The subclass consisted of all prisoners confined in their cells for 22 hours or more per day. Defendants allegedly were subjecting subclass members to a substantial risk of serious harm and injury from various policies, including constitutionally inadequate out-of-cell time, social isolation, nutrition, and mental health treatment.

The Court certified the class and subclass in March 2013. (Doc. 372). The Ninth Circuit affirmed certification of the class and subclass. Parsons v. Ryan , 754 F.3d 657 (9th Cir. 2014). In doing so the court held the prisoners’ claims had sufficient commonality because the class and subclass were allegedly subjected to practices that could be addressed "in one stroke." Id. at 678. In particular, the Ninth Circuit reasoned this Court could determine whether the "statewide [ADCRR] policies and practices to which all members of the class are subjected, and [particular] statewide [ADCRR] policies and practices which affect all members of the subclass" exposed the class and subclass to "substantial risk of serious future harm to which the defendants are allegedly deliberately indifferent." Id.

II. Settlement and Stipulation

Before the mandate issued in the class certification appeal, the parties negotiated a settlement agreement—the Stipulation—that was intended by the parties to eventually resolve all claims. (Doc. 1185). The Stipulation contained 103 health care and 10 maximum custody performance measures.² With respect to health care, performance measures were to be assessed pursuant to a "Monitoring Guide" and reported monthly for each of the ten prison complexes. (Doc. 1185 at 3). Under the Stipulation, a performance measure, for health care or maximum custody, could terminate (i.e. , free Defendants from monitoring and reporting) if Defendants achieved, and maintained, specified thresholds: 75% compliance the first 12 months, 80% the second 12 months, and 85% thereafter. The Stipulation was accepted by the Court in February 2015. (Doc. 1458). III. Stipulation Monitoring History

2 As it happened, the Performance Measures were not comprehensive enough and under-evaluated the alleged constitutional violations. Defendants used this to their advantage to both subvert the Stipulation and continue violating prisoners’ constitutional rights.

In an Order dated July 16, 2021, the Court provided a detailed recital of the history of the monitoring and enforcement of the Stipulation for the five years from 2016 through 2021. (Doc. 3921). Rather than repeat the history, the Court incorporates the factual findings contained in the July 16, 2021, Order.

In brief, between 2016 and July 2021, Plaintiffs filed twelve motions to enforce the Stipulation, the Court held multiple evidentiary hearings and status conferences, the Court issued dozens of Orders with specific directions mandating Defendants comply with the Stipulation, the Court issued three Orders to Show Cause why Defendants should not be held in contempt, the Court appointed two experts, and the Court held Defendants in contempt twice, resulting in millions of dollars in fines, which were upheld on appeal. At the end of the five-year period, the Court concluded Defendants had consistently refused to perform the obligations under the Stipulation and had offered baseless legal and factual theories for their failures. Imposition of substantial fines, and threats of even more, did not prompt Defendants to make required efforts to perform as they agreed under the Stipulation.

Accordingly, as of July 2021, the Court resolved that attempting to enforce the Stipulation had proven futile and rescission of the Stipulation was required.³ (Doc. 3921 at 36-37). The Court vacated the settlement under Federal Rule of Civil Procedure 60(b) and set the matter for trial to begin November 1, 2021. (Doc. 3921).

3 Of significance, the Stipulation substantially favored Defendants because it expressly provided by its terms that the Court was precluded from ordering Defendants "hire a specific number or type of staff." (Doc. 1185 at 14-15). This provision was addressed in one of Defendants’ appeals. The Ninth Circuit held the provision meant the Court could issue a "general staffing order" but could not issue an order requiring a "specific number and type of personnel." Parsons v. Ryan , 912 F.3d 486, 498 (9th Cir. 2018). This interpretation made enforcement difficult, and eventually completely impossible, because any effective staffing plan would necessarily address the specific needs of specific complexes. Further, the Court was consistently reminded Defendants placed great importance on this limitation and were unwilling to hire additional staff. As this Order demonstrates, failure to maintain adequate staff has been the fundamental cause of Defendants’ unconstitutional actions.

IV. Trial Procedures

The parties were given time after the Order vacating the Stipulation to engage in discovery and prepare for trial. (Docs. 4006, 4009, 4011, 4012, 4019, 4019, 4021). In early August 2021, Plaintiffs moved to present direct expert testimony by declaration to preserve trial time. On September 2, 2021, the Court granted the motion over Defendants’ objection and directed both sides to present most of their experts’ direct testimony by way of written declaration.⁴ Defendants, however, if necessary, were allowed to file a motion seeking permission to present all their experts’ testimony in-court instead of through written declarations. (Doc. 3952 at 3 n.1). Defendants chose not to file such a motion.

4 A similar practice was used in another case involving challenges to prisoner healthcare. See Madrid v. Gomez , 889 F. Supp. 1146, 1158 n.5 (N.D. Cal. 1995) ("The direct testimony of all experts was submitted by way of declaration, supplemented by two hours of live testimony. The parties were also permitted unlimited live cross examination and redirect examination.").

After considering the nature of the evidence for both parties, the Court directed the parties to plan on 50 hours of trial time for each side, with the caveat that "adjustments, as necessary" would be made to ensure "a fair trial to both parties." (Doc. 3952 at 3). Neither party objected to the 50-hour limit nor asked for additional time at the conclusion of the trial.

Trial commenced on November 1, 2021 and proceeded for fifteen non-consecutive days. The parties filed lengthy written testimony for each expert and they testified at trial. Beyond the experts, the Court also heard testimony from class and subclass members, the named defendants, deputy wardens in charge of isolation units, and others.⁵ Neither side had exhausted their allotted time on the last day of trial or asked the Court to continue the trial for additional days. The parties subsequently filed approximately 800 pages of proposed findings of fact and conclusions of law as well more than 350 pages of responses to the opposing side's filing. (Docs. 4308, 4309, 4314, 4315).

5 The Court heard from the following witnesses: Prisoners Kendall Johnson, Ronald Slavin, Laura Redmond, Rahim Muhammad, Jason Johnson, and Dustin Brislan; Staffing Expert Robert Joy; Paralegal Jessica Carns; Deputy Wardens Travis Scott, Lori Stickley, and Anthony Coleman; Mental Health Care Expert Pablo Stewart; Solitary Confinement Expert Craig Haney; former Centurion Regional Mental Health Care Director Dr. Stefanie Platt; Corrections Expert Martin Horn; Medical Care Expert Dr. Todd Wilcox; Defendants ADCRR Director David Shinn and Assistant Director Larry Gann; ADCRR's Mental Health Director Dr. Bobbie Pennington-Stallcup; Yuma Site Director Dr. Elijah Jordan; Warden Jeffrey Van Winkle; ADCRR Medical Director Dr. Grant Phillips; Mental Health Care Expert Dr. Joseph Penn; Medical Care Expert Dr. Owen Murray; and Centurion vice president Tom Dolan.

V. Evidentiary Issues

Most of the evidentiary issues were raised and resolved during trial but a few are pending.

A. Motion to Strike (Doc. 4219)

After trial, Defendants filed a motion seeking to strike "portions of Plaintiffs’ experts’ declarations." (Doc. 4219). That motion is confusing. For example, Defendants argue the written testimony of one of Plaintiffs’ experts was incorrectly presented such that it must be excluded but then add, without explanation, that the expert's "entire trial testimony" should also be stricken. (Id. at 2). Defendants also claim there is hearsay within the experts’ written testimony, that some offered legal conclusions, and maybe they offered undisclosed opinions. Defendants’ motion is baseless because the motion was untimely and, as such, Plaintiffs were unable to cure any alleged deficiencies.

Defendants’ motion was filed over a month after Defendants received the last written testimony by one of Plaintiffs’ experts. Federal Rule of Evidence 103 requires objections or motions to strike be "timely." As a general matter, "[a]n objection is ‘timely’ if it is made as soon as the opponent knows, or should know, that the objection is applicable." Jerden v. Amstutz , 430 F.3d 1231, 1236-37 (9th Cir. 2005) (quotation marks and citation omitted). The Ninth Circuit has found an objection two days after a witness testified untimely. Id. In another case, the Ninth Circuit found untimely an objection "not raised until the end of the witness's direct examination." San Antonio Cmty. Hosp. v. S. California Dist. Council of Carpenters , 125 F.3d 1230, 1238 (9th Cir. 1997). Defendants’ objections raised a month or more after they learned of the testimony are not timely. Thus, the objections will be denied.

What is more, by waiting until the conclusion of trial to bring their belated motion, Defendants deprived Plaintiffs of an opportunity to cure the alleged flaws in the testimony. The Ninth Circuit made clear it is improper to exclude testimony based on an untimely objection when the proponents of the testimony lack an "opportunity to cure the objection." Jerden , 430 F.3d at 1238. Defendants’ motion to strike will be denied. However, the Court's findings of fact are not reliant on patently inadmissible evidence, whether contained in written testimony or elsewhere.⁶

6 It is self-evident that both the prisoners and Defendants cannot be considered completely neutral witnesses because prisoners as well as Defendants and their staff are obviously interested in the outcome of this case.

B. Motions to Admit Evidence (Docs. 4215, 4225)

Plaintiffs moved to admit numerous court filings by both sides, a report by a court-appointed expert, and previous reports by experts. (Doc. 4215). Plaintiffs also moved to admit a report from the "Arizona Auditor General" and an excerpt of a prisoner's electronic medical records. (Doc. 4225). Defendants objected to most of these exhibits but agreed that one exhibit, the electronic medical records, can be admitted provided additional information is included. (Doc. 4247).

It is unnecessary to consider the documents and objections in detail because the Court has not materially relied on the documents at issue. Plaintiffs’ motion regarding court filings will be denied as moot but the motion regarding medical records will be granted, subject to Defendants’ request to include additional information.

C. Joint Motion to Admit (Doc. 4224) ⁷

7 Plaintiffs motion to exclude evidence and testimony due to Defendants’ Rule 30(b)(6) designee's lack of knowledge is moot because no contrary testimony was offered during trial. (Doc. 4163).

The parties have agreed to admit certain exhibits, admit with modifications other exhibits, and withdraw some exhibits. The stipulated motion will be granted and the exhibits admitted or withdrawn as agreed to by the parties.

D. Admissibility of Expert Witnesses’ Testimony

"Under Rule 702 of the Federal Rules of Evidence, [a] witness who is qualified as an expert by knowledge, skill, experience, training, or education may testify in the form of an opinion, provided the testimony meets certain criteria." United States v. Gadson , 763 F.3d 1189, 1202 (9th Cir. 2014) (quotation marks and citation omitted). An expert witness may rely on information, such as hearsay evidence, that would not be admissible on its own. Fed. R. Evid. 703. The time to raise a challenge to an expert's testimony is before or occasionally during trial. "Failure to raise a Daubert challenge at trial causes a party to waive the right to raise objections to the substance of expert testimony post-trial." Skydive Arizona, Inc. v. Quattrocchi , 673 F.3d 1105, 1113 (9th Cir. 2012). Here, the parties waived any Daubert challenges to the opposing party's expert witnesses. But even if they had not, any such challenge would have failed.

Expert testimony is sufficiently "reliable if the knowledge underlying it has a reliable basis in the knowledge and experience of [the relevant] discipline." United States v. Sandoval-Mendoza , 472 F.3d 645, 654 (9th Cir. 2006) (quotation marks and citation omitted). The test for admissibility "is not the correctness of the expert's conclusions but the soundness of his methodology." Primiano v. Cook , 598 F.3d 558, 564 (9th Cir. 2010) (quotation marks and citation omitted). Here, all the parties’ experts have relevant knowledge and experience. The parties disagree regarding the soundness of some experts’ methodologies but those disputes, as well as the appropriate weight to assign to each expert's opinions, are addressed below. The Court has considered all the expert opinions as well as the manner and demeanor of every witness's testimony in assessing credibility.

LEGAL STANDARD

The Eighth Amendment "imposes duties on [prison] officials, who must provide humane conditions of confinement; prison officials must ensure that inmates receive adequate food, clothing, shelter, and medical care, and must take reasonable measures to guarantee the safety of the inmates." Farmer v. Brennan , 511 U.S. 825, 832, 114 S.Ct. 1970, 128 L.Ed.2d 811 (1994). Prisons officials "violate[ ] the Eighth Amendment only when two requirement are met." Id. at 834, 114 S.Ct. 1970. "First, the deprivation alleged must be, objectively, sufficiently serious" such that it results "in the denial of the minimal civilized measure of life's necessities." Id. Second, "a prison official must have a sufficiently culpable state of mind." Id. That state of mind is "deliberate indifference," which requires a prison official "knows of and disregards an excessive risk to inmate health or safety." Id. at 837, 114 S.Ct. 1970. A prison official's state of mind "is a question of fact subject to demonstration in the usual ways, including inference from circumstantial evidence, and a factfinder may conclude that a prison official knew of a substantial risk from the very fact that the risk was obvious." Id. at 843, 114 S.Ct. 1970.

HEALTH CARE

I. Privatization and Current Contract

When this lawsuit was filed in March 2012, ADCRR⁸ provided health services (medical and mental health care) directly to its prisoner population. But the Arizona State Legislature enacted legislation requiring "the privatization of all correctional health services, including all medical and dental services." Arizona 2011 HB 2154. That legislation took effect on July 1, 2012, when Wexford Health Services assumed responsibility for all health services. On March 4, 2013, Corizon Health Incorporated took over for Wexford. Centurion of Arizona became the third private health care contractor on July 1, 2019. That contract is valued at approximately $216 million per year. (Doc. 4274 at 32). In June 2021, ADCRR extended its contract with Centurion for an additional fifteen months. (Doc. 4274 at 96). ADCRR is currently soliciting bids for the health care contract to begin in August 2022. (Doc. 4287 at 25).

8 In 2019, the name was changed from "Arizona Department of Corrections" to "Arizona Department of Corrections, Rehabilitation, and Reentry." (Doc. 4279 at 31).

The current contract requires Centurion provide all health services at ADCRR's ten complexes in the state. The table below reflects the prisoner population at each complex as of September 30, 2021, the date discovery closed. The table also reflects the number of "full time equivalents" or "FTEs" (i.e. , employees working 40 hours per week) required by the contract and the number of filled FTE positions.⁹ These FTE numbers cover all Centurion staff at that location, meaning medical, dental, and mental health staff as well as administrative staff.

9 The population figures are as of September 30, 2021, while the staffing figures are as of August 2021.

Prison Population, Contract FTE, and Actual FTE
Location	Population	Contract FTE	Actual FTE
Douglas	1,524	33.75	32.10
Eyman	5,219	132.50	110.85
Florence	2,534	152.00	96.76
Perryville	3,318	136.75	109.35
Phoenix	468	88.05	67.93
Lewis	4,421	150.00	115.00
Safford	1,035	34.75	36.85
Tucson	4,420	180.50	141.75
Winslow	1,481	31.00	37.10
Yuma	3,374	86.50	104.25

These numbers show an obvious and crucial amount of understaffing at certain locations. For example, Tucson has approximately 40 fewer FTEs than required by the contract. Defendants claim understaffing is alleviated by employees working overtime or by recruiting temporary employees. Even if this occurred, the current staffing levels fall significantly beneath what Centurion and Defendants intended at the time they entered into the contract.

II. Differences Among Complexes

The health services at the ten complexes are structured the same, subject to a few caveats. Each complex has a medical unit staffed with Centurion medical personnel. (Doc. 4279 at 59). The medical units at every complex consist of exam rooms allowing for treatment of prisoners on something like an "out-patient" basis. That is, prisoners go to the medical unit, are seen by medical staff, and the prisoners then return to the housing units. Beyond the standard medical units, some complexes have intensive health units for prisoners with more significant medical needs. An inpatient care unit (also known as an "inpatient component" or "IPC") "is staffed 24/7 with a registered nurse" and is where prisoners requiring a high level of care are housed. (Doc. 4279 at 61). A special needs unit ("SNU") provides a lower level of care and may be staffed "on an LPN level." (Doc. 4279 at 61). The typical care in SNUs consists of care for prisoners who require assistance with activities of daily living, or who have mobility or memory issues that preclude their placement in the general population. (Id. at 61). In addition, seven of the complexes are considered "corridor facilities" because they are located near larger population centers where higher levels of off-site medical care are available. (Id. at 60). Finally, four complexes have "mental health units" where prisoners with intensive mental health needs are housed. The table below depicts the medical and mental health capabilities of each complex. (Doc. 4279 at 60-64); (Doc. 4172 at 40-41). Of the ten complexes, nine are for male prisoners and one, Perryville, is for female prisoners.

Medical and Mental Health Capabilities of Complexes
Complex	IPC	SNU	Corridor	Mental Health Units
Douglas	No	No	No	No
Eyman	No	No	Yes	Yes
Florence	Yes	Yes	Yes	No
Perryville	Yes	Yes	Yes	Yes
Phoenix	No	No	Yes	Yes
Lewis	Yes	No	Yes	No
Safford	No	No	No	No
Tucson	Yes	Yes	Yes	Yes
Winslow	No	No	No	No
Yuma	No	No	Yes	No

Based on the facilities and capabilities of each complex, male prisoners may be transferred between complexes depending on their needs. If a prisoner housed at Winslow were to develop a serious medical condition, it is unlikely Winslow would be capable of treating that condition. The prisoner might initially be transferred to a "corridor" facility to be closer to off-site medical resources. If the prisoner was transferred to a "corridor" facility without an IPC or SNU, but the prisoner deteriorated, the prisoner may need to be transferred again to a facility with an IPC or SNU. Similarly, if a prisoner at Yuma needed intensive mental health treatment, that prisoner would need to be transferred to a mental health unit at a different location.

The different capabilities and facilities between complexes provide significant support for examining the ADCRR medical and mental health care system as a whole, instead of on a complex-by-complex basis. For example, in order for a particular prisoner to receive adequate or critical care, transportation among complexes is often necessary, because of the health care diversification in facilities and the overall medical care provided at ADCRR. A prisoner may become seriously injured or develop a serious medical or mental health care condition at any time. If that prisoner is male, he may be assessed at one complex, transferred to almost any other complex, and then transferred again if needed, depending on bed availability and proximity to relevant specialists. Insufficient care provided at the IPC in Lewis may, suddenly, be a matter of life or death to a prisoner previously housed at Douglas who develops a need for a high level of care.

III. Process to Obtain Care

When a prisoner submits a "health needs request" ("HNR") form, the response typically informs the prisoner he is scheduled to be seen on the "nurse line," meaning he will be seen and evaluated by an RN. Alternatively, if the HNR requested mental health services, he will be seen by mental health staff. (Doc. 4276 at 70-71).

A. Overview of Medical Care

When an RN sees a prisoner on the nurse line, the RN uses "protocols embedded in Nursing Encounter Tools (NETS)." (Doc. 4138 ¶ 163). The NETS are "templates for care that are supposed to guide the nurse's exam and assessment of the patient, and help the nurse formulate a plan of care, including whether to send the patient to a PCP [primary care provider]." (Doc. 4138 ¶ 163). As set forth in Defendants’ policies, after examining the prisoner, the RN will determine "[i]f the situation requires that the inmate be seen by another discipline." (Ex. 1305 at 155). If the RN concludes additional care is needed, policy requires the RN "immediately initiate a referral to the appropriate area." (Ex. 1305 at 155) (emphasis added). Significantly, this first evaluation by an RN always occurs regardless of the seriousness of a prisoner's asserted symptoms and regardless of how many times the prisoner has previously been seen by an RN for the very same issue. (Doc. 4138 at 72 ¶ 171).

If the RN determines the prisoner's situation requires evaluation by a higher-level professional, the RN will arrange for the prisoner to see what the parties refer to as a "provider." As used by the parties, "provider" means a nurse practitioner, physician assistant, or physician. A provider may order further testing or prescribe medication. A provider may also determine specialist care is needed. In that situation, the provider is required to send a referral to Centurion's Utilization Management department. (Doc. 4275 at 13). That department, however, does not examine or evaluate the prisoner but based on the medical records, it "look[s] at the clinical indication to make sure [the referral is] appropriate," which takes time. (Doc. 4279 at 84). If the referral is approved, the information is passed to one of Centurion's "clinical coordinators" who will arrange a specialist appointment for the prisoner. (Doc. 4279 at 84). If Centurion denies the referral, the Utilization Management department often states the provider should pursue an "alternative treatment plan." (Doc. 4138 ¶ 400).

The many prisoners who have chronic conditions are enrolled in ADCRR's Chronic Care Clinics. (Doc. 4206 ¶ 201). A policy defines a chronic condition as "an illness or condition that affects an individual's well-being for an extended interval, and generally is not curable, but can be managed to provide optimum functioning within any limitations the condition imposes on the individual." (Ex. 1305 at 151). Prisoners’ chronic care conditions are monitored by chronic care appointments no less than every 180 days, unless a provider documents a longer time frame is appropriate. (Doc. 1185-1 at 27). These appointments are not on the "nurse line." Rather, they are supposed to be with a provider although the overall records do not make clear whether that occurs. (Doc. 4287 at 59-60).

Defendants created a medical classification system to identify the housing needs for individual patients. The designations range from M-1 through M-5 with M-1 being the lowest needs patient and M-5 being the highest needs patient. (Doc. 4279 at 65).¹⁰ Defendants introduced testimony that as of trial there are approximately 22,000 prisoners classified as M-1 and M-2, 2,000 prisoners classified as M-3, between 400 and 500 classified as M-4, and approximately 20 classified as M-5. (Doc. 4279 at 65-66). Those numbers, however, are clearly approximations as their sum is less than the total number of prisoners in the ten complexes.

10 The scores are:
M-1: Maximum sustained physical capacity consistent with age; no special requirements.
M-2: Sustained physical capacity consistent with age; stable physical illness or chronic condition, no special requirements.
M-3: Restricted physical capacity; requires special housing or reasonable accommodations.
M-4: Limited physical capacity and stamina; severe physical illness or chronic condition; requires housing in a corridor institution.
M-5: Severely limited physical capacity and stamina; requires assistance with Activities of Daily Living (ADLS); requires housing in Inpatient Component or Assisted Living area.
(Doc. 4138 ¶ 484).

B. Overview of Mental Health Care

Defendants created a standardized scoring system for classifying prisoners according to their mental health needs. MH-1 reflects the lowest level of need, and MH-5 reflects the highest.¹¹ (Ex. 3025 (Mental Health Technical Manual "MHTM" Ch. 3 § 5.0)). These scores are assigned during "mental health intake assessments" which occur during prisoners’ initial processing upon entering the system. (Doc. 4276 at 43).

11 The scores are the following:
MH-1: Prisoners who have no history of mental health issues or receiving mental health treatment.
MH-2: Prisoners who have received mental health treatment in the past but do not currently have any mental health needs and have demonstrated behavioral and psychological stability for at least six months.
MH-3: Outpatient Treatment. Patients who have current mental health needs that require outpatient treatment. There are five sub-codes to MH-3.
MH-3A: Patients in acute distress who may require substantial intervention in order to remain stable. All patients classified as seriously mentally ill ("SMI") are to be classified as MH-3A (unless admitted to a residential treatment or inpatient treatment program, and then classified as MH-4 or MH-5). Any patient under a Psychiatric Medication Review Board ("PMRB") order for involuntary administration of psychiatric medication are to be classified as MH-3A (unless admitted to a residential treatment or inpatient treatment program, and then classified as MH-4 or MH-5).
MH-3B: Patients who are generally stable but need regular interventions because they are receiving psychiatric and psychological services.
MH-3C: Patients who are stable, have adequate coping skills, and are able to manage their mental health symptoms through medication only, and who need infrequent intervention.
MH-3D: Patients who were recently taken off psychotropic medications and need follow-up for six months thereafter to ensure stability over time.
MH-3E: Patients who recently arrived to ADCRR custody and are generally stable but may benefit from regular contacts with mental health clinicians, or patients participating only in outpatient group psychotherapy.
MH-4: Residential Treatment. Patients who are admitted to a residential mental health program.
MH-5: Inpatient Treatment. Patients who are admitted to the inpatient mental health treatment programs licensed by the Arizona Department of Health Services.
(Ex. 3025 at 27-32).

Prisoners assigned a score of MH-3 or above are considered "on the mental health caseload," and must be housed at facilities with mental health staff. Prisoners assigned scores of MH-1 or MH-2 are eligible to be housed at Douglas, Safford, or Winslow, which each have a single staff person each to address any mental health needs that may arise.

In addition to learning of mental health needs through the HNR process, mental health staff may learn of a prisoner in need of services through the prisoner's friends or families, or through prison officials who observe the prisoner. (Doc. 4276 at 71). In theory, the system provides that if mental health services are requested through the HNR process, an "urgent concern" will result in the prisoner being seen the same day the HNR is processed. (Doc. 4276 at 70). If the HNR presents a "clinical mental health need," the prisoner is seen within five days. And if there is a "psychiatric concern" that is deemed "non-urgent," the prisoner is seen within 14 days. The record does not establish who is responsible for categorizing the mental health HNRs.

MEDICAL CARE

It is undisputed the class members suffer from a wide variety of serious medical needs. In fact, Bureau of Justice Statistics data show people who are "incarcerated tend to be less healthy, to have more chronic illnesses including substance use disorder, and to have additional stressors in their lives than people who live in the general community." (Doc. 4138 ¶ 252). Dr. Elijah Jordan, the Site Medical Director for the Yuma Complex, testified prisoners often present unusual, and very serious, medical conditions. He explained working in correctional health care "afforded [him] the opportunity to see a lot of things that [he] read about and saw going through medical school that [he] was not able to see during medical school as well as residency, because [he is] able to – [he] take[s] care of patients that do not regularly go to the doctor. So these guys, when you do get a chance to evaluate and see them, they are really bad off." (Doc. 4277 at 56.) And Defendants readily admit approximately two-thirds of class members are prescribed medications, and approximately one quarter receive medication for mental health conditions. (Doc. 4275 at 16, 18). As a general matter, the class consists of thousands of individuals with significant and very serious health needs requiring a high level of care to adequately manage and treat.

I. Overview of Expert Testimony

The parties presented competing experts to address the overall adequacy of the medical care system. Dr. Todd Wilcox testified for Plaintiffs and Dr. Owen Murray testified for Defendants. Dr. Wilcox's testimony was persuasive and credible. Dr. Murray's overall opinions were generally unpersuasive though portions of his testimony confirmed the medical care system is grossly insufficient.

A. Dr. Wilcox

Dr. Wilcox earned a bachelor's degree in biological psychology from Duke University in 1988 and a medical degree from Vanderbilt University School of Medicine in 1992. (Doc. 4138, Ex. A). After graduating from medical school, Dr. Wilcox completed a residency in orthopedic surgery. (Doc. 4270 at 82). During that residency, Dr. Wilcox began working as a staff physician for the Salt Lake County jail. In 1996, he became the medical director for the Salt Lake County Jail System, where he remains today. He has also performed other work, including serving as the Medical Director for the Maricopa County Jail System. He is board certified in Urgent Care Medicine and board certified by the American Academy of HIV Medicine. And he holds an advanced certification from the National Commission on Correctional Healthcare as a Certified Correctional Health Professional Administrator and Correctional Health Physician.

Dr. Wilcox has 27 years of experience in correctional health care and is well-qualified to provide expert opinions on the quality of Defendants’ health care delivery. (Doc. 4138 ¶¶ 2, 4). He has served as Plaintiffs’ expert since 2013, shortly after the lawsuit was filed. (Id. ¶ 6). Defendants did not file a pretrial motion contesting Dr. Wilcox as an expert witness, and he is qualified to offer expert opinions on the adequacy of the medical care provided.

Dr. Wilcox was requested to "evaluate the adequacy of the current medical care delivery system in the ADCRR." (Id. ¶ 18). His current opinion is based on his experience serving in this case since 2013, having submitted multiple reports and declarations regarding the delivery of healthcare in the Arizona state prisons which document that since 2012 ADCRR's healthcare delivery system has been seriously deficient. In preparing for his testimony, he thoroughly reviewed many documents including the written policies and procedures of ADCRR as well as the Department Orders and Medical Service Technical Manual. He examined medical records for 120 patients. He visited all four IPCs and interviewed a number of patients in IPCs, special needs units, and other housing units.

Dr. Wilcox explained why he did not review a completely random sample of prisoners:

I did not review a random sample of all patients in ADCRR custody. This is because when evaluating a healthcare delivery system, it is generally not as helpful to examine care for healthy people as it is to look at the treatment of sick patients, particularly those with complex or chronic conditions that require coordination, communication, and judgment. Healthy patients or those with minimal needs can more readily get their needs met even from systems that offer poor clinical care and lack basic organizational structures; examining their records tells us little. It is the complex patients who test the capacity of staff and systems alike.¹²

12 Dr. Wilcox elaborated on this point in his live testimony.

(Doc. 4138 at ¶ 25). However, Dr. Wilcox's tour of various locations and discussions with prisoners at those locations meant he did not dictate the exact prisoners he would review:

I had no control over who was housed in those units. We tried to speak to everybody in those units who was willing to speak to us. So we got as big of a sample as we could with regards to those random patients who were housed in those units and then we went back and looked at their charts.

(Doc. 4270 at 12).

Overall, Dr. Wilcox attempted to evaluate the care provided to prisoners by looking to prisoners "at the sicker end of the spectrum." (Doc. 4270 at 11). He explained why this approach was useful:

I also wanted to identify patients who are high-end utilizers of the health care system. And the reason that those patients are important is because when you are doing a systems evaluation, what you really want to evaluate are patients who put a little strain on the system. They have a lot of transactions in their health care. It's not particularly informative about a system if you have a patient that has a one-time episodic visit for a sprained ankle and then they're never seen again. That doesn't give you very much data. But what does give you data is patients that have frequent and ongoing interactions with the healthcare

system because you're able to see the coordination nation of care, you're able to see all the different elements of the system operating to care for that patient.

(Doc. 4270 at 11). Dr. Wilcox has used this methodology to evaluate care in other cases and other experts use similar methodology. (Doc. 4270 at 15).

In his written testimony, Dr. Wilcox stated ADCRR's health care system provides "grossly inadequate healthcare." (Doc. 4138 at ¶ 28). The entire health care system "continues to harm many patients and continues to place all at substantial risk of serious harm." (Id. ). And "ADCRR has provided dangerously substandard care for years, both before and during the pandemic, continuing to the present day." (Id. at ¶ 27). He testified this problem stems from a variety of factors but the most fundamental being the lack of appropriate qualified staffing. "By design, healthcare decisions in the [ADCRR system] are pushed down to the lowest possible level – nurses who are practicing poorly and far outside the scope of their licenses." (Id. ¶ 28). In Dr. Wilcox's view,

There is a clear pattern of failure by nurses to complete an adequate nursing assessment, take patient reports seriously, recognize dangerous symptoms, and elevate concerns to providers. Too often, nurses simply send patients back to their housing unit and tell them to submit another written sick call request if symptoms worsen.

(Doc. 4138 at ¶ 28).

ADCRR's system also results in care for complex issues being "scattered throughout the system so no one provider or physician is ultimately responsible for the patient, resulting in serious deficits in care." (Doc. 4138 at ¶ 29). In addition, the system lacks "the foundation of any credible healthcare practice: differential diagnosis." (Doc. 4138 at ¶ 30). Medical staff fail "to identify, test, and otherwise evaluate the underlying cause of the symptom." (Id. ). Staff order testing "without a diagnostic strategy" and they "fail to follow up on significantly abnormal results." (Id. ).

In his written testimony Dr. Wilcox identified the cause of the system's failures:

[ADCRR's] problems appear to the result of a combination of factors, including inadequate staffing, inadequate physician-level attention to problems, and poor attitude of medical staff,¹³ which probably is itself related to inadequate staffing and demanding workloads. In addition, the electronic health record in use within the ADCRR is poorly designed and greatly impairs the clinician's capacity to synthesize a comprehensive picture of a patient's healthcare.

13 Dr. Wilcox testified "[p]roviders’ notes suggest over and over that patients are lying and malingering." (Doc. 4271 at 79). In Dr. Wilcox's opinion, this was evidence of "bias against patients" which often "interferes with the providers’ capacity to recognize serious medical conditions."

(Id. ¶ 31).

During his in-court testimony, Dr. Wilcox summed up his opinion of the medical care system in a simple and stark way:

My opinion is that the system as it exists is terrible. There are so many blocks to care, barriers to care. There's so many areas where the care can get snagged up and not flow appropriately, and the end result of that is that some patients within the healthcare system, in fact many patients who have higher end needs, end up getting terrible care.

(Doc. 4270 at 20). To ensure it had understood, the Court asked Dr. Wilcox to clarify. Dr. Wilcox repeated prisoners "end up getting terrible care." (Doc. 4270 at 20).

B. Dr. Murray

Dr. Murray earned a bachelor's degree in biology from Boston College in 1983 and a doctor of osteopathy degree from Chicago College of Osteopathic Medicine in 1988. (Doc. 4206, Ex. 1). After medical school, Dr. Murray completed a residency in family practice. From 1991 to 1994, Dr. Murray worked as medical director at various locations in the Illinois Department of Corrections. From 1995 to the present, he has worked at the University of Texas Medical Branch. (Doc. 4285 at 104). At present, he is Vice President of Offender Health Services and oversees the provision of health care in the Texas Department of Criminal Justice and the Texas Juvenile Justice Department. (Doc. 4206 ¶¶ 3-4).

Dr. Murray understood he was to evaluate the "overall health care system" for prisoners in ADCRR custody and determine whether there are "systemic deficiencies which render the delivery of health care services below the community standard of care."¹⁴ (Id. ¶ 10). He performed site visits at all ten complexes and focused his review on the following components of medical care: (1) staffing, (2) access to care, (3) diagnostic services, (4) records and facilities, (5) pharmacy services, and (6) quality monitoring. (Id. ¶¶ 15, 17). Dr. Murray reviewed charts for 80 prisoners, focusing on prisoners with at least two documented chronic care diagnoses. Dr. Murry evaluated those charts based on the quality of chronic care documentation, quality of chronic care, and quality of episodic care. (Id. ¶ 207). Dr. Murray's overall conclusion was that:

14 The relevant standard is the Eighth Amendment, not the community standard of care. Balla v. Idaho , 29 F.4th 1019, 1026 (9th Cir. 2022) (noting Eighth Amendment liability must be predicated on something "beyond malpractice"). However, "[t]he community standard of care outside of the prison context is highly relevant in determining what care is medically acceptable and unacceptable." Id. (quotation marks and citation omitted).

ADCRR is delivering care that meets or exceeds the community standard. While I have identified some areas for improvement, most notably with eOMIS [the electronic medical records system], PM [performance measure] reporting, and the mortality review process, I did not find any systemic deficiencies with the provision of health care to ADCRR inmates. Even in the areas where I have noted opportunities for improvement, it is clear that ADCRR recognizes the issues and has been working towards improving them.

(Id. ¶ 1082). Thus, Dr. Murray believes the medical care provided to class members is the same or better than what members of the public receive from their medical providers. As set forth below, this opinion is contradicted by the evidence produced at trial.

II. Staffing is Significantly Below Contract Levels

The core issue is that staffing levels are so inadequate that the provision of constitutionally mandated care is impossible. Dr. Wilcox testified "staffing is the root of the ADCRR's healthcare deficiencies" (Doc. 4138 ¶ 160), which opinion was established by the evidence at trial. The current staffing levels illustrate ADCRR does not have the ability to address the varied and often complex needs of Arizona's prisoners.

The contract between ADCRR and Centurion requires 1052.75 full-time staff members ("FTEs") to provide health care, mental health care, dental care, and perform all related administrative tasks. (Doc. 4287 at 17). Plaintiffs contend, and the evidence establishes, the contractually allotted staffing for each complex is definitely too low and, because of current vacancies, each facility is profoundly understaffed. In fact, the evidence establishes Centurion admitted staffing at the contractual level would be insufficient, even if such staffing levels could be obtained.

Although ADCRR claimed that 1052.75 FTEs were necessary at some point, Defendants failed to explain how they arrived at this number. Defendant Gann stated in his Rule 30(b)(6) deposition "the [staffing] model has been in place for a number of years, and it [is] not something that is negotiated by us ongoing." (Doc. 4146-1 at 221). In other words, the staffing model was already in place at the time Centurion began providing services. Defendant Gann also testified Centurion has been encouraged "to present an alternative [staffing] plan if they so desire." (Doc. 4146-1 at 228). But at trial, Defendant Gann incredibly contradicted his deposition testimony and claimed ADCRR relies on Centurion to report how many health care staff are needed. (Doc. 4275 at 89). The evidence establishes the staffing numbers were set by ADCRR, and it has refused to require increased staffing.

Centurion Vice President Tom Dolan testified that, in July 2019, Centurion performed its own staffing evaluation to determine the number and type of health care staff necessary to perform adequately under the contract and determined Centurion would need to increase the FTEs by 161.5 for a total of 1214.25 FTEs. (Ex. 2166). This was specifically communicated to ADCRR officials. (Doc. 4287 at 29). But ADCRR "was not open to amending the contract to add additional FTEs," despite the insufficiency noted by Centurion. (Doc. 4287 at 30). Of greater significance, Centurion has never even fulfilled the staffing level required by contract. (Doc. 4274 at 99). The current situation falls well below the already-insufficient contract level.

The Court looks to the staffing levels of the most crucial personnel. The following positions are responsible for providing direct patient care: Assistant Director of Nursing, Director of Nursing, Licensed Practical Nurse/Medical Assistant, Medical Assistant, Medical Director, Midlevel Practitioner (nurse practitioners or physician's assistants), Nursing Assistant/Patient Care Technician, Registered Nurse, and Staff Physician. Using these positions, the tables at Appendix 1 set forth the contract level, the hired level, and the variance between the contract and hired levels as of August 2021. (Ex. 2167).

All told, there were 465.45 FTEs providing direct care as compared with the 601.75 FTEs required by the contract. The current staffing figures do not include overtime performed by current staff or temporary staffing. A Centurion employee claims use of overtime and temporary employees has allowed Centurion to work 90% of the hours contemplated by the contract. (Doc. 4287 at 17). But that does not indicate 90% of the hours contemplated by the contract for direct patient care were performed.

As established by the tables at Appendix 1, seven out of ten facilities are understaffed, and some are severely understaffed. The numbers at Tucson, for example, are striking. Tucson houses 4,420 prisoners. According to the Centurion contract, to provide medical care to that population, 40 licensed practical nurses are required but there are only 24.30. The contract requires 36 registered nurses but there are only 18.30. The contract requires 2 staff physicians but there is only 0.75. To be clear, there is not a single full-time physician to cover the 4,420 prisoners housed at Tucson. And even if there were the 2 staff physicians contemplated by the contract, the physician to prisoner ratio would be severely inadequate. Another court noted a physician to prisoner ratio of 1:1,166 was far too large. Madrid v. Gomez , 889 F. Supp. 1146, 1201 (N.D. Cal. 1995). Even if Tucson were "fully staffed" according to the contract, the physician to prisoner ratio would be 1:2,210.

Overall, there are 10 medical directors, 47.15 midlevel providers (who may only work under the direction of a physician), and 5.5 staff physicians for the entire prisoner population of over 27,000. Three complexes do not require a staff physician at all (Douglas, Safford, and Winslow), two more do not have their staff physician positions filled (Lewis and Phoenix), and two more have their staff physician allocations understaffed by 50% (Florence and Tucson). These numbers, and the tables at Appendix 1, demonstrate the contract staffing levels require the vast majority of medical care be provided by nurses, nursing assistants, and medical assistants. Given these undisputed actual staffing numbers, Defendants’ position that there is no cause for concern—and they are providing care that meets or exceeds the community standard—is incredible.

Defendants argue this overview does not account for the various site medical directors. Those positions are filled by physicians and Defendants claim they increase the number of physicians providing direct patient care. First, the increase is insignificant and trial testimony established the medical directors are ill-equipped to provide anything close to full time direct patient care. Dr. Jordan, the site medical director at Yuma, testified he spends between 40 and 50% of his time providing patient care, "because [he is] involved." (Doc. 4277 at 59-60). Dr. Jordan also confirmed he has acted as the site medical director for the Winslow complex in addition to his duties at the Yuma complex on more than one occasion for at least one month. But Dr. Jordan never provided direct patient care at Winslow. (Id. at 70-71). Dr. Wilcox also stated that, during his tours of various locations, he had learned "site medical directors did not do very much clinical work and they substituted in for vacations." (Doc. 4271 at 74).

Even if medical directors were not required, at times, to cover multiple locations, it is inconceivable that each site medical director could provide anything close to full-time direct patient care while supervising a staff of 20-100 and performing administrative responsibilities. As such, based on this record, any participation by the site medical directors in providing direct patient care is marginal and, therefore, inadequate.

Defendants effectively concede the current staffing levels are insufficient. In addressing whether the current staffing levels are adequate, Dr. Murray's opinion was based on information from "facility leadership teams" in terms of the hypothetical staffing contemplated by the contract, not actual staff. They told Dr. Murray if "their vacancies were filled, the current workload would be manageable with the current number of healthcare positions." (Doc. 4206 ¶ 197). Because Centurion has never filled all its vacancies, it is completely unclear how the teams know full staffing would be sufficient. But, more importantly, these comments by "facility leadership teams" effectively admitted the current staffing levels are not sufficient. Moreover, the evidence establishes the contract staffing levels would not be sufficient, even if they could be obtained. Facilities currently considered "overstaffed" according to the contract level are unable to meet basic thresholds for providing care. The Yuma complex illustrates the contract staffing levels are insufficient. Yuma is overstaffed by seventeen positions (nearly thirteen of which are nursing assistants, medical assistants, or LPNs). Despite being "overstaffed" as contemplated by the contract, Yuma was unable to perform numerous health care performance measures during the Stipulation's operative period. (Doc. 4277 at 75-76). From January 2021 through July 2021, Yuma was substantially noncompliant with PM 37, which requires a prisoner to be seen within 24 hours of submitting an HNR, and PM 44, which requires the prison to review a prisoner's hospital discharge instructions within 24 hours of their return from the hospital. In other words, even with Yuma substantially "overstaffed," it was unable to see prisoners within 24 hours or review discharge instructions within 24 hours, both tasks a well-staffed facility should be able to accomplish. It was these types of critical failures that ultimately destined the Stipulation to fail and mandated its rescission. (Id. at 78-83). In addition, Yuma is one of the only complexes where Defendants claimed use of the eConsults service, where prison providers can reach out virtually to specialists, explain a prisoner's symptoms and testing results, and seek specialized advice on how to proceed with a patient. (Id. at 60:8-21). While this may have been helpful, Yuma's performance in obtaining specialty consultations remained substantially noncompliant. (Id. at 75-76).

Significantly, Defendants offered no evidence the contract staffing levels, if achieved, would be sufficient. They did not offer a statistical analysis demonstrating the staffing allocation contemplated by the contract would be appropriate to meet prisoners’ needs. Instead, Defendants relied, almost exclusively, on their National Commission on Correctional Health Care ("NCCHC") accreditation. NCCHC accredits correctional facilities for compliance with NCCHC's standards regarding medical and mental health care. (Doc. 4172 ¶¶ 59-61; Doc. 4279 at 101; Doc. 4270 at 110-111). Correctional facilities initiate the accreditation process and the institution seeking accreditation pays an initial fee, as well as an annual fee to maintain accreditation. (Doc. 4289 at 133-134). NCCHC will accredit a facility if it finds that a facility meets all "essential" NCCHC standards and 85% of all "important" NCCHC standards. (Doc. 4270 at 114:6-15; Doc. 4275 at 61:24-70:11).

ADCRR is fully accredited. But the evidence established that accreditation is of very little value when determining actual performance of the ADCRR system. Dr. Wilcox testified "NCCHC has never and does not now purport to assess the adequacy of staffing levels as part of its accreditation process. In fact, the NCCHC has always declined to offer any guidance on staffing other than to provide an umbrella statement that " ‘[t]he responsible health authority (RHA) ensures sufficient numbers and types of health care staff to care for the inmate population.’ " (NCCHC Standards for Health Services in Prison, 2018, pg 61); (Doc. 4138 ¶ 158 n. 7). This was corroborated by Defendant Gann who testified NCCHC will not provide a benchmark for staffing because every facility is different. (Doc. 4275 at 33). Further, while Defendants claim NCCHC surveyors are neutral and objective, the evidence established Centurion and Corizon are far from neutral because they are prominent NCCHC sponsors. (Doc. 4283 at 5).

The Court finds Dr. Wilcox's opinion that none of the NCCHC essential standards incorporate a substantive review of the health care administered by ADCRR is credible. Rather, these standards are exactly what Dr. Wilcox said they are—standards to ensure some policies and procedures are in place to create a health care delivery system. (Doc. 4138 at page 66 n.7). And those systems do, in fact, exist, so it is not surprising ADCRR was and remains accredited. But it has never been Plaintiffs’ claim that Defendants do not have policies or procedures to provide health care. Rather, Plaintiffs have always claimed the system was providing constitutionally inadequate health care.

If there were any remaining doubt about the utility of NCCHC accreditation, it is dispelled by looking at Defendants’ PM compliance from 2018-2021. As the Court detailed in its July 16, 2021 Order, pervasive noncompliance with the Stipulation's performance measures justified rescission of the Stipulation. Indeed, just one example, PM 44 at Eyman—requiring prisoners returning from a hospital stay shall have the hospital's treatment recommendations reviewed and acted upon within 24 hours—has been noncompliant for years. Given that performance, it is implausible all ten complexes could be 100% compliant with each "essential" NCCHC standard if those standards were based on a qualitative assessment of the actual care provided to prisoners.

Ultimately, the evidence demonstrates current staffing is woefully insufficient. In addition, as Dr. Wilcox stated "[t]hese staffing problems persist and recur because neither the ADCRR nor Centurion adequately analyze, monitor, or take responsibility for addressing them." (Doc. 4138 at ¶ 160). Based on the testimony and evidence, the only reasonable conclusion is the current staffing level is inadequate, and even if all positions contemplated by the contract were filled, it is more probable than not the staffing would be insufficient. The Court need not decide the number of staff that would be necessary. It is enough to know that even Centurion has admitted the contract staffing level, which has never been achieved, is insufficient. The next issue is the impact that lack of staff has on the care provided to prisoners.

III. Intersection of Inadequate Staffing and Nursing

The staffing levels contemplated by the contract tilt sharply toward support staff—licensed practical nurses, medical assistants, nursing assistants, and patient care technicians—and away from providers who can write orders and prescribe medication. All positions are not created equal. Several witnesses attempted to put the staffing numbers into context, and Dr. Wilcox's testimony was the most thorough and persuasive.

Dr. Wilcox explained the process by which prisoners seek care. The ADCRR system requires every prisoner seeking medical care to be evaluated by a nurse who then uses the nursing encounter tools ("NETS") to formulate a plan of care, including whether to send the prisoner to a provider—nurse practitioner, physician's assistant, or physician. (Doc. 4138 ¶ 163).

Dr. Wilcox's undisputed characterization of ADCRR's health care hierarchy is that nurses are tasked with triaging prisoners but quite often prisoners do not obtain care beyond that provided by a nurse. And of greater significance, this system requires nurses to diagnose and treat far beyond their capabilities. Nurses are expected to work closely with higher-level providers, but given the relatively few higher-level providers on staff, that is simply not possible. In Dr. Wilcox's expert opinion, "the poor quality of clinical decision-making demonstrated by nurses and providers in the ADCRR harms patients and places them at an unreasonable and substantial risk of serious harm." (Doc. 4138 ¶ 17). During his in-court testimony, Dr. Wilcox elaborated on this point and pointed out the strangeness of ADCRR using nurses in this manner.

Dr. Wilcox was asked if he was "aware of any other health care settings where the nurse serves as final decider when someone seeks to access their doctor." He responded

No. Considering that it's not really legal, you wouldn't expect to find any others. But, you know, can you imagine in the community if you schedule an appointment with your doctor and you're met in the lobby by the nurse who does a little assessment on you and then turns you around and sends you home and you're not allowed to see your doctor? That just doesn't exist in the scope of healthcare anywhere."

(Doc. 4270 at 74). ADCRR's decision to structure their system in this manner is, in Dr. Wilcox's credible opinion, an extreme outlier.

Dr. Murray attempted to counter Dr. Wilcox's opinion by stating "[t]he nurses and providers I interviewed felt comfortable with the supervision available, and I believe the ratio of providers to nurses in the ADCRR system allows for adequate supervision." But Dr. Murray does not identify a single provider who explained how supervision is adequate to protect prisoners. And with 5.5 staff physicians for the entire prison population of more than 27,000, the evidence does not support a finding that the ratio is sufficient for adequate supervision.

In essence, it is Defendants’ position that access to any care, no matter how poor, satisfies their constitutional obligations. In Defendants’ view, prisoners have access to nurses, and those nurses can in theory escalate matters to higher level providers when appropriate. But the evidence shows that they do not do so. And in practice, it is never enough to provide prisoners with serious medical needs access to health care staff who lack the training and authority to address their needs. Access to health care necessarily involves access to health care performed by competent staff. The staffing levels detailed above make clear it would be impossible for nurses to meaningfully consult with or refer many prisoners to providers because there are simply not enough higher-level providers available to see them.

As credibly stated by Dr. Wilcox, "in a well-functioning system, a high percentage of patients should pass through the triage process to see a provider, who can then diagnose their condition and prescribe appropriate treatment." (Doc. 4138 ¶ 162). Instead, ADCRR relies almost exclusively on nurses to "function outside the scope of their licenses and act as providers." (Doc. 4138 ¶ 169). The basic staffing levels allow for nothing else.¹⁵ The staffing levels require the vast majority of prisoners who submit HNRs to be seen by nurses, and nurses only. And even the nurse lines are cancelled due to lack of staffing. Defendant Gann admitted the nurse line was cancelled in Tucson more than twenty times in January and February of 2021 due to insufficient staffing. (Doc. 4275 at 100:25-101:3).

15 For example, at the Tucson complex there are 18.3 registered nurses, 7.00 midlevel practitioners and 0.75 staff physicians. There are 4,420 prisoners at that complex who submit 200-300 HNRs per day. (Doc. 4206 at 7). It is not remotely possible for each prisoner who submits an HNR to be seen by a midlevel practitioner or staff physician. The ADCRR system documented a total of 300,000 HNRs across all complexes in 2020. (Doc. 4275 at 145). The staffing levels systemwide mean only the tiniest fraction of those HNRs could possibly result in a visit with a midlevel practitioner or staff physician.

Dr. Wilcox opines the inadequacy of ADCRR's system, including the overreliance on nurses, is borne out by reviewing the numerous mortality reviews.¹⁶ Those reviews—which Defendants did not controvert—document a lack of timely access to appropriate care by the appropriately skilled professionals and they reflect numerous preventable deaths.

16 Dr. Wilcox examined 94 mortality reviews of the 322 prisoners who died between January 1, 2019, and September 26, 2021. Dr. Wilcox "focused primarily on more recent deaths to gain the most accurate picture of the current state of affairs in the ADCRR healthcare system." (Doc. 4167 at ¶ 21).

A. Joseph Assyd ¹⁷

17 The Court uses the full names of prisoners who died in ADCRR custody as their deaths are reported publicly. For prisoners who remain in custody, the Court will use their initials unless the prisoner testified at trial.

Joseph Assyd died on April 12, 2020, at the age of 64. On March 13, 2020, he became unresponsive due to hypoglycemia. A nurse measured his glucose at a very low level of 40, and Assyd reported abdominal pain, nausea, vomiting, and diarrhea. He was provided "glucose gel 3 times over the next hour." While he had "a fever of 102.3, hypoglycemia, and abdominal pain," he was transferred from the Lewis complex to the Tucson complex. Assyd was not seen by a provider on March 13. It was not until March 27, 2020, that he saw a provider for complaints of "generalized weakness and pneumonia." He had an altered mental state and was transferred to the hospital. Assyd tested positive for Covid-19 on April 7, 2020, developed sepsis, and died on April 12, 2020.

The mortality review reflected Assyd's vital signs were not properly documented on March 13, 2020, and, consequently, his assessment on that date did not prompt a referral to a provider. In addition, there was no "clear cut documentation" why he was transferred between facilities. Despite this, the mortality review found no issues with the care. Instead, officials merely made "recommendations" that transfers between facilities be accompanied by an explanation why the transfer was occurring and that referrals to providers "must be authorized timely based on level of urgency." (Ex. 155). In other words, the nurse responsible for the assessment on March 13, 2020, should have concluded Assyd needed to be seen by a higher-level provider.

B. Kenneth Barker

Kenneth Barker died on October 5, 2019, at the age of 68. He was housed in the Tucson complex. He was seen by a nurse on May 29, 2019, because of "foot pain with difficulty walking." There is no indication his feet were examined at that time, but he was referred to a provider. Before that provider visit, he was seen again by a nurse on June 10, 2019, for foot pain. Barker was told a provider visit had been scheduled but his feet were not examined.

On July 8, 2019, for the third time, he was seen by a nurse because of foot pain. This time, his right foot was examined. He was seen by a nurse for the fourth and fifth time on July 27 and August 6, with a note on August 6 as "unremarkable." Finally, at his sixth visit on August 21, 2019, a nurse examined his left foot and determined there was a 2.5 cm ulcer on the bottom of his left foot. He was referred a second time to a provider, but the examination did not occur until September 3, 2019, when the wound on his foot had become necrotic and he exhibited edema on his lower left leg between the knee and ankle.

Barker was sent to the hospital, where it was determined the ulcer may have been caused by several months of wearing ill-fitting shoes. Despite aggressive treatment, including amputation of his lower left leg, Barker died on October 5, 2019. His mortality review noted Barker should have been escalated to a provider in a "more timely manner" after two HNRs were submitted by him for the same complaint. (Ex. 159). The mortality review noted "[p]reventative measures not taken" and the diagnosis was "not timely." The mortality review recommended "[a]ll diabetic patients when seen for chronic care must have a visual assessment [of] the feet." The treatment of Barker illustrates the system relies on nurses to assess patients, even when the patient makes repeated complaints for the same issue.

C. Cardinel Barnes

Cardinel Barnes died on April 6, 2021, at the age of 37. He was housed in the Eyman complex. He had a chronic care visit on March 17, 2021 for asthma. The following week, on March 24, 2021, he saw a nurse for shortness of breath and an upper respiratory infection. He told the nurse he had three asthma attacks the previous day, his inhaler was not working, and he was so short of breath that it had taken him 30 minutes to get dressed. The nurse did not refer him to a provider. Instead, the nurse instructed him to call ICS¹⁸ if he had another asthma attack and his inhalers were not working. Barnes collapsed and became unresponsive on April 6, 2021. CPR was performed, but it was unsuccessful. The autopsy found evidence of valley fever. The mortality review questioned whether "the patient [should] have been referred to a provider for the complaints on shortness of breath and diffuse pain," but concluded "[t]he evaluation and recommendations by nursing were appropriate given the clinical presentation." (Ex. 160). Barnes is another example of a prisoner being unable to access a provider and instead being subjected to cursory treatment by a nurse.

18 Incident Command System, the health services’ emergency response system. (Ex. 1305 at 30).

D. Andrew Barnhart

Andrew Barnhart died on July 17, 2020, at the age of 72. He was housed in the Eyman complex. During the two years preceding his death, he lost over thirty pounds and had a body mass index of 15, demonstrating he was extremely underweight. A June 3, 2020, chest x-ray showed "near complete opacification [increased density] of the right upper [lung] lobe." Barnhart underwent a CT scan on June 26, 2020, which revealed a 14x10x12 cm mass. At that time, his blood pressure was 86/64, which was ignored. He was transported to the hospital on July 8, 2020 to evaluate abdominal and back pain with hypotension. He was taken into emergency surgery to address a perforated gastric ulcer, and a lung mass was discovered. Barnhart developed post-surgical complications and died. (Ex. 161). The mortality review stated the CT scan on June 26, 2020, presented "an opportunity for additional questioning, a blood pressure recheck, and elevation to the provider" regarding his blood pressure. But the record does not establish that the results of the CT scan and the low blood pressure were reviewed by anyone, which may or may not have resulted in a different outcome but certainly the lack of review presented a substantial risk of serious harm.

E. James Edwards

James Edwards died on January 1, 2020, at the age of 29. He was housed in the Eyman complex. He submitted his first HNR for leg and stomach pain on October 7, 2019 but did not see a nurse until November 8, 2019 and then a second time on November 17, 2019. It was noted he had a mass on his collar bone and an elevated heart rate. He was referred to a provider who did not see him until November 23, 2019. He complained of severe hip pain and the only diagnosis was mild osteoarthritis in his hip.

Twenty-four days later, on December 17, 2019, Edwards was seen by a provider after an ICS was activated for "shoulder pain," but the provider determined his shoulder was "normal." Another ICS followed on December 20, 2019, because Edwards felt "weakness and discomfort" and he was also "unable to keep water down." He was seen by a nurse who instructed "to drink more water in slow sips." Ten days later, Edwards became unresponsive and was transferred to the hospital where he died of valley fever. The mortality review accurately concluded nurses failed to recognize signs or symptoms, delayed access to care, and failed to follow clinical guidelines. (Ex. 229).

F. Gabriel Figueroa

Gabriel Figueroa died on December 11, 2020, at the age of 55. He was housed in the Tucson complex. He suffered from Hepatitis C and a variety of other conditions. He had a chronic care visit on November 27, 2020, and two days later, on November 29, he saw an RN because an LPN said he "did not sound good." Figueroa reported a cough that started the previous day, shortness of breath, and lethargy. His oxygen saturation was low at 92%. The nurse advised him to submit an HNR or initiate an ICS if he continued to experience shortness of breath. An ICS was initiated the next day when his oxygen saturation was recorded as 62%. He was transported to the hospital, where he died on December 11, 2020. His cause of death was listed as Covid-19 pneumonia.

The mortality review concluded the nurse who examined Figueroa on November 29 should have referred him to a provider "to determine the next steps" to determine the source of his illness (Ex. 241). This reflects another instance where a nurse did not accurately assess the prisoner's needs and prevented him from seeing a provider.

G. Bruce Manthey

Bruce Manthey died on December 31, 2019, at the age of 71. He was housed in the Tucson complex. He suffered from several preexisting conditions. An ICS was called on November 29, 2019, because Manthey was experiencing painful edema bilaterally. Medical staff provided medication and advised him to keep his feet clean. Several days later, he saw health care staff to address a left shoulder abscess, but there was no mention of edema. Thirteen days later, Manthey had a nurse visit on December 12, 2019 for bilateral edema that began when he stood up in the morning. The record does not specify a diagnosis or treatment and the December 18, 2019 provider note does not mention edema or skin lesions. Another ICS was initiated on December 23, 2019, because Manthey could not sit in his wheelchair because of pain from his bilateral edema, which had grown to over 6mm of swelling. He was in pain, and open wounds on both feet were "draining foul discharge." No beds were available in the IPC unit, and he was sent to the emergency room. Manthey became septic, was admitted to the ICU, and he died on December 31, 2019 due to multiorgan failure secondary to sepsis. The mortality review set forth the obvious: Manthey's symptoms were not addressed according to the standard of care, and the documentation in his medical record did not reflect Manthey's actual condition. In particular, Manthey's "condition should have been elevated by nursing ... due to obvious symptoms (edema, odor, drainage) of something wrong." There is no explanation how nurses could have missed these "obvious symptoms." (Ex. 327). This care illustrates the dangers of setting up the system such that nurses are the first, and often only, medical professional available to prisoners.

H. Jose Miranda

Jose Miranda died on August 4, 2019, at the age of 76. He had been housed in the Tucson complex infirmary or SNU since 2015. On July 25, 2019, after he returned from an off-site medical appointment, he was observed sitting up straight but seemed stiff on one side, flaccid on the other, and was not speaking. Despite these issues, staff did not send Miranda to a provider. The next day, he continued to have right side stiffness. He opened his eyes, but he was aphasic, had a weak left grip, and a minimal response to ammonia inhalant. He was sent to the hospital for stroke evaluation and died. The mortality review noted that when nursing evaluated him following his return from his appointment on July 25, 2019, "his altered mental status and vital signs should have triggered the patient to be evaluated to the attention of a provider." (Ex. 344). This reflects yet another example of nurses missing obvious signs that merited referral to a higher-level provider.

I. Gilbert Molina

Gilbert Molina died on February 15, 2021, at the age of 73. He was housed in the Eyman complex. On January 11, 2021, he fell and was evaluated by nursing. At that time, Molina reported "10/10 pain." The nurse did not immediately contact a provider for diagnosis. Instead, an appointment was set and occurred two days later. The provider ordered x-rays, but not expedited and they occurred five days later. The x-rays revealed a fractured femur. Molina underwent surgery, but he contracted Covid-19, developed complications, and subsequently died. The mortality review noted "[a]n elderly patient with severe pain after a fall should be escalated to a higher level of care immediately." In addition, the x-rays should have been ordered "stat" not "routine." (Ex. 346). From the records, it appears the nurse was insufficiently trained and skilled to treat the condition Molina presented with on January 11.

J. David Neville

David Neville died on July 22, 2019, at the age of 59. He was housed in the Tucson complex. He had a history of squamous cell carcinoma at the base of his tongue for which he had received chemotherapy and radiation. In February 2017, a PET scan showed "lymph nodes suspicious for recurrence" of cancer. At that time, Neville complained of neck and throat pain. In October 2018, during a chronic care visit, he complained of "unrelenting pain in [his] throat." He was treated with pain medication for a short time, but the medication was discontinued because the pain "was not a result of cancer." Neville continued to have difficulty swallowing.

Neville hung himself on July 22, 2019. The mortality review noted it was "undetermined" if the death could have been prevented or delayed. The review also stated Neville submitted several HNRs due to unrelenting pain but "[n]ursing staff [did] not appreciate the level and severity of pain symptoms and therefore no referrals [were] made to the health care practitioner." (Ex. 355). This reflects another example of nurses being insufficiently trained and skilled to fully diagnose and treat the presented condition.

K. Pedro Rojas

Pedro Rojas died on June 26, 2019, at the age of 51. He was housed in the Florence complex. On May 6, 2015, he reported he had been suffering from hemorrhoids and rectal bleeding for over a year. Results of three hemoccult tests ordered by a nurse were all positive for occult (hidden) blood, but there was no follow-up ordered. Rojas continued to complain of diarrhea and abdominal pain for years with no diagnosis. On April 19, 2017, hemoccult tests were repeated and, again, all were positive for blood. He was diagnosed with severe anemia and sent to the hospital.

At the hospital, a colonoscopy showed prominent hemorrhoids and a large fungating, nearly obstructing tumor of the colon. "Chemo radiation" was recommended. Rojas was transferred to Florence, and on May 17, 2017, he was seen by an oncologist. The record does not indicate the result of that examination but Rojas continued to complain of weakness. He was sent to the ER again and, upon his release, he was transferred to Lewis. He had been scheduled for radiation on June 21, 2017, but the appointment was overlooked. As of June 2017, his "abdominal pain continued to worsen" and he had to be sent to the hospital repeatedly "for blood transfusions for anemia due to blood loss from rectal bleed." Chemotherapy was finally started on August 7, 2017. He had additional hospitalizations for sepsis, pneumonia, and pyelonephritis, and underwent surgery in October 2017. At that time, he was noted to have Stage 4 cancer.

In April 2019, Rojas completed chemotherapy and radiation treatments and underwent a biopsy of a lung nodule, which showed necrotic adenocarcinoma consistent with metastasis. In May 2019, he developed a small bowel obstruction, and he was admitted to the infirmary on June 14, 2019. On June 21, 2019, Rojas developed hypotension and died. The mortality review noted that Rojas should have been referred to a health care practitioner in a timely manner after the initial positive hemoccult tests in 2015 and there were "significant delays from diagnosis to eventual treatment in addition to delays during treatment of the patient's cancer."

ADCRR and Centurion officials recognize the failures in Rojas’ care. The mortality review notes that prisoner health care had "improved dramatically" in recent years and Rojas’ death "most likely would not have taken place in the current health care delivery climate." It is unclear to what "current health care delivery climate" Defendants refer to as Rojas’ case is relatively recent. (Ex. 398).

L. Kamaka Solo

Kamaka Solo died on September 15, 2019, at the age of 58. He was housed in the Tucson complex. On June 8, 2019, Solo began complaining of constipation. Solo's complaints worsened throughout June, and he complained of gastric distress, bloating, gas, abdominal pain, cramping, and continued constipation. He was seen by nurses and given laxatives. He returned in July 2019, complaining that "even drinking water causes pain," he was unable to eliminate despite trying and straining for hours, and he was unable to eat or sleep. He was referred to a provider, who did not see him until July 9, 2019. The provider ordered x-rays, which were deemed "normal." Again, on July 19, 2019, Solo returned with constipation, gas, bloating, constant pain due to abdominal cramps and inability to sleep or eat. As required, he was seen first by a nurse on July 20 and, after a referral, a provider saw him on July 25, who gave Solo medication for gas and heartburn as well as a laxative.

On August 1, 2019, Solo requested to see a specialist for unresolved pain. He stated, "I am in constant, excruciating pain, and all the treatment I have received so far has been totally ineffective." A nurse saw him and did nothing. On August 3, he submitted an HNR, which stated "[t]he pain is unbelievable, definitely a ‘10’. Could I please have an enema? I need help." Nursing diagnosed "constipation" and gave him "nothing."

On August 12, 2019, Solo submitted an HNR stating "[n]o human being deserves to endure the pain I am in 24 hours a day. Can you help me?" That resulted in a provider appointment on August 13, 2019. The provider repeated the x-ray order and ordered a liquid diet but no diagnosis. Rather, on August 15, 2019, nursing again gave him a laxative. On August 18, 2019, an ICS was called due to Solo's chest pain. A hemoccult was ordered then cancelled, and a helicobacter pylori test was ordered. That same day, a nurse noted Solo's pain was "increasing" and he was sent to the IPC for observation.

While in the IPC, a provider requested a gallbladder ultrasound, but it was not authorized until September 3, 2019, and was never done. Solo was returned to his normal housing unit. On August 20, 2019, Solo reported to a provider that he had not eaten for 2 weeks. The provider once again gave a laxative and pain reliever.

Over the following days, ICS emergencies were called because of vomiting and extreme pain. Solo's heartrate was irregular, and an EKG was done and read with "sinus rhythm with occasional PVC, L anterior fascicular block and nonspecific T wave abnormalities." But Solo was sent back to his cell. On August 26, 2019, Solo was seen by nursing, who noted that Solo was "very uncomfortable, grimacing and moaning." On August 27, 2019, another ICS was called for chest pain, Solo's blood pressure was 96/64, and he was witnessed seizing and throwing up bile. Solo was finally sent to a hospital.

The emergency department performed a CT scan which revealed a large pancreatic mass and multiple hepatic lesions, likely metastatic. On September 10, 2019, Solo suffered a stroke while in the hospital. Solo was sent back to the prison on September 14, 2019 and died on September 15, 2019.

Solo's mortality review noted his "symptoms were worked up routinely and a diagnosis was attempted (as being H. pylori and constipation)." However, "[a]larming symptoms were missed and/or not elevated, and pain was not addressed adequately." The review recommended nursing staff "be provided education regarding multiple HNR[s] for the same complaint to be elevated to the practitioner."

From June 2019 through September 2019, Solo complained of unremitting and extreme pain. The response by nursing staff, on multiple occasions, was to do nothing. And even when elevated to a provider, diagnostic tests were ordered but not performed and Solo was given effectively no treatment for his pain. (Ex. 422).

M. Macario Vela

Macario Vela died on April 21, 2019, at the age of 43. He was housed in the Lewis complex. On April 6, 2019, Vela was seen by a nurse for pain in his left breast radiating to his back at the base of neck on the left side. The nurse identified a palpable lump in Vela's left shoulder. On April 10, 2019, Vela was scheduled to a see a provider, but the record states "refused." On April 14, 2019, he returned to a nurse complaining of increasing back pain in lumbar spine and radiating into left leg, stated he had not eaten in 3 days, and was walking stiffly. The nurse provided ice and "analgesic balm."

Vela was seen by nursing again on April 15, 2019, and after an abnormal urine dip, a urine drug screen was returned positive for opiates. At that visit, Vela had a temperature of 103 degrees. Vela was given fluids for some period but declined additional fluids and refused to see the provider. On April 20, 2019, an ICS was called for a suspected overdose and Vela died the next day.

Vela's autopsy revealed staphylococcus aureus sepsis due to chronic intravascular drug use. His mortality review noted failure to recognize symptoms or signs and patient non-adherence as contributing factors of Vela's death. But the review also noted Vela did not receive a timely diagnosis or treatment. Diagnosis could have potentially been made on the first visit given Vela's complaints, and the subsequent visits should have alerted the nurses that Vela was very ill. The mortality review noted poor documentation, there was no record of labs on April 15, 2019, and failure to closely monitor in that there was growing evidence of possible endocarditis. (Ex. 442). Vela may have refused to see a provider which is a complication, but given his symptoms, nurses should have been able to determine he was at very significant risk.

N. Michael Voden

Michael Voden died on June 16, 2019, at the age of 77. He was housed in the Florence complex. On intake on August 14, 2015, it was noted Voden had a history of hypertension and COPD. A heart murmur was detected and described as aortic valve dysfunction. On October 10, 2018, an ICS was initiated because Voden had "chest pain 10/10." Voden reported recurrent chest pain for two months with recurrent nausea. An EKG showed abnormalities and ischemia or subendocardial injury but no review by a provider. Voden was given antacids.

On December 28, 2018, a second ICS was initialed for "man down," and Voden reported generally feeling unwell and chest discomfort, but there was no record of calling a health care provider. On January 30, 2019, Voden reported chest pain with tightness and difficulty breathing to a nurse and a provider diagnosed COPD exacerbation, GERD, and pleuritic chest pain. On March 5, 2019, Voden had complained to a provider of a progressive leg edema for three years, and bilateral edema was noted, but the provider said it was due to sitting in a wheelchair. Voden's continued shortness of breath was attributed to COPD. On March 21, 2019, a third ICS was initiated for shortness of breath, crackles were heard in Voden's chest, and Voden had lowered pulse oxygen levels. Nursing staff noted shortness of breath on a subsequent appointment, and on March 26, 2019, a provider diagnosed possible congestive heart failure.

On April 14, 2019, Voden was seen by nursing for respiratory symptoms and edema and was instructed to "increase fluid intake." He was referred to a provider, who described him as frail but in no acute distress. The provider recommended that Voden "drink plenty of water," with no follow-up appointment. On April 18, 2019, a fourth ICS was initiated after prison staff witnessed Voden exhibiting labored breathing and nursing staff advised him to "hydrate and ... be seen on nursing line." On April 20, Voden was seen on the nursing line and "diagnosed" with "poor air exchange secondary to Asthma." On April 29, 2019, the provider noted Voden had a "4+ edema" in his left leg with a loud murmur along with lung wheezes and assessed poorly controlled COPD and poor compliance with medication. On May 7, 2019, Voden reported "life threatening emergency respiratory, circulatory function" and requested to be seen by a provider outside of ADCRR. Nursing evaluated Voden and determined he had "fluid weeping from his [lower extremities] along with difficulty breathing and a 25 lb. wt. gain." Nursing referred him to a provider.

Voden was seen several times throughout May, and at the end of May, a cardiology consult was requested. On June 5, 2019, Voden's condition had seriously deteriorated, and he was diagnosed with heart failure and pleural effusions, possible aortic stenosis, cellulitis of legs, and COPD, and was sent to the hospital. At the hospital, Voden was diagnosed with severe non-rheumatic aortic stenosis and deemed a poor candidate for surgery. He was also diagnosed with a number of other conditions, including acute renal failure. Voden died in the hospital on June 16, 2019, after he suffered cardiac arrest.

Voden's mortality review noted a pattern of nursing staff acting beyond their expertise and failing to recognize when referral to a provider was necessary. Nursing should have contacted a provider in October 2018 based on Voden's complaints of chest pains. Nursing should have contacted a provider on December 28, 2018, when Voden again complained of chest pain. The initial diagnosis of congestive heart failure in March 2019 should have been accompanied by other diagnostic tests, such as an echocardiogram, and further follow-up. The recommendation by nursing on April 14 to increase fluid intake very likely was inappropriate because of the serious edema. The provider's exam on April 15, 2019, was inadequate in that the provider failed to note the "significance" of the murmur accompanied by significant edema. Inexplicably, the provider also recommended that Voden increase his fluid intake. During the remainder of April and May 2019, medical staff repeatedly failed to properly diagnose Voden's condition. Overall, Voden's death was possibly avoidable. (Ex. 445).

Voden's case presents a situation where nursing staff ineffectively repeatedly attempted to treat him without referring him to a provider. In doing so, nursing staff missed obvious signs of other conditions and gave wrong treatment advice.

O. Conclusions Derived from Mortality Reviews

The mortality reviews illustrate the harm that routinely befalls prisoners because they do not receive timely and adequate health care. The common theme is nurses repeatedly are unable to properly diagnose health care issues and fail to refer prisoners to a provider. Despite these reviews, Defendants still maintain nurses act reasonably in evaluating prisoners. Dr. Wendy Orm, the statewide medical director, testified during her Rule 30(b)(6) deposition that she is satisfied nurses appropriately assess patients and believes they accurately determine whether and when to refer patients to a provider for care. (Doc. 4148 at 46). Dr. Orm admitted, however, her opinion was not supported by any attempt on her part to validate nurses’ assessments or prisoners’ treatment. (Doc. 4148 at 47-49). Dr. Wilcox accurately found "Dr. Orm's failure to have conducted any quality assurance studies on the accuracy and appropriateness of the nursing assessments and on the limited throughput of patients to providers is an abrogation of her basic duties." (Doc. 4138 ¶ 146).

The reality that nurses do not, in fact, accurately assess prisoners and then refer them to providers is further supported by evidence from the Yuma complex. That facility conducted self-audits to determine whether nurses were selecting the correct NET assessment tool when evaluating a prisoner on the nurse's line. The numbers for selecting the correct assessment tool were as follows. At the La Paz Unit, the correct tool was selected between 0 and 20% of the time from May through August 2021. (Ex. 825 at 25); (Ex. 835 at 14); (Ex. 855 at 12); (Ex. 914 at 12). At other units, the correct tool was selected 43% and 58% of the time in September 2021. (Ex. 914 at 10); (Ex. 914 at 11).

This undisputed evidence shows nurses routinely selected the wrong NET when evaluating a prisoner. That does not necessarily establish the nurses reached inappropriate diagnoses. But if, as the evidence established, nurses are not using the diagnostic templates properly, it is likely the nurses are not properly assessing conditions.

ADCRR's response to the mortality reviews shows deliberate indifference of the deficient practice of placing nurses as the gatekeeper for diagnosis and treatment. Dr. Wilcox explained:

It is clear, through my analysis of mortality reviews authored in the last two years, that there are still significant deficiencies in the process. I did not find the mortality reviews, on the whole, to be honest, thorough, or effective. They minimized the harm caused by healthcare staff, lacked the requisite specificity, failed to identify clear errors in care, failed to offer effective recommendations, and evidenced no staff accountability, even after identification of serious errors that led to a patient's death.

(Doc. 4138 ¶ 131). These reviews do not simply evaluate each prisoner's final interaction with ADCRR's health care. The reviews show systemic failures as well as ADCRR's failure to "identify and address the core cause of many of these problems: nurses practicing outside the scope of practice, insufficient physician-level oversight, and failure to refer the patient to someone qualified to diagnose and treat them." (Doc. 4138 at 56). Dr. Wilcox believes "the response and the write up for the mortality reviews are tepid, and I think they are very blunted and not really directive in any way to try to implement change that would prevent a case like this from happening again." (Doc. 4269 at 91). The mortality reviews demonstrate that, despite ten years of litigation, ADCRR has never created and then implemented a policy to identify systemic issues identified in mortality reviews and has not taken steps to remedy them.

Dr. Orm testified during her deposition if ADCRR identifies an "actionable" recommendation in a mortality review, then staff will develop a corrective action plan ("CAP"). But there was no evidence Defendants developed even one CAP through reviewing mortality reviews. Of significance, Dr. Murray agreed with the assessments of prisoners’ care in the mortality reviews. (Doc. 4286 at 83). As part of his analysis, he also asked Defendants for data to evaluate whether recommendations in the mortality reviews were ever implemented but received none. (Doc. 4286 at 84). Dr. Murray was forced to admit "there remains some room for improvement in terms of consistently identifying and articulating actionable findings." (Doc. 4206 ¶ 1040). That is an understatement. The same failures arise in mortality reviews over and over and over—nurses not recognizing conditions, nurses not referring prisoners to providers, and diagnostic testing not being ordered or promptly obtained and reviewed—and yet not a single CAP was documented or identified. This constitutes systemic conscious disregard of the risk prisoners face.

The deaths outlined above, some very recent, were not aberrations or corrective action would have been implemented. No such evidence was introduced because it does not exist. Defendants’ failure to take any action in response to such obvious deficiencies up to the date of trial is evidence Defendants are content to continue using nurses inappropriately and failing to require medically appropriate follow-up treatment.

Using nurses as the first line, and often last line, for providing care is medically unacceptable. While using nurses in this way is driven by a lack of higher-level staffing, that does not excuse Defendants from adopting a system that leads to preventable deaths. Dr. Wilcox states—and Defendants do not contest—ADCRR's high ratio of physicians to mid-level providers is an extreme outlier among health care systems. Dr. Wilcox was unaware of any "healthcare systems that provide primary care to patients on a large scale where the physicians are overwhelmingly outnumbered by the mid-level providers." (Doc. 4167 at ¶ 224). In Dr. Wilcox's credible opinion, large systems do not have ratios similar to what ADCRR has because "large systems inevitably have numerous patients who require complex care beyond the scope of mid-level providers." (Doc. 4167 at ¶ 224). Indeed, ADCRR recognizes this deficiency by indicating the ratio of mid-level providers to physicians set forth in their pending Request for Proposal is 2:1. (Doc. 4206 ¶ 1071). It is obvious the present system of often allowing prisoners access only to nurses puts prisoners at a substantial risk of serious harm.

IV. Mismanagement of Complex Cases

Another failure that stems from lack of staffing is ADCRR's failure to employ a differential diagnosis approach. Symptoms are treated without developing a "diagnostic strategy" or attempting to test for the underlying cause. Recurring complaints by prisoners are often treated as "new" issues at each appointment without taking account of previous diagnoses and history. (Doc. 4138 ¶ 30). And Dr. Wilcox found "care for complex patients is scattered throughout the system so no one provider or physician is ultimately responsible for the patient[.]" (Doc. 4138 ¶ 29). Because it is impossible for providers with demanding caseloads (due to insufficient staffing) to assess prisoners’ history and symptoms, practitioners miss "with alarming frequency" serious and urgent medical symptoms. (Doc. 4138 ¶¶ 29, 31). "That impossibility translates directly into inadequate care" for all prisoners but particularly for those with severe illness. (Doc. 4138 ¶ 29).

This leads to failures or delays in diagnosing serious conditions. What is entirely absent from nearly every patient interaction is an objective approach to their symptoms. The absence of differential diagnoses is particularly problematic because prisoners often do not see the same health care provider. On top of this multi-provider problem, the electronic health records system used does not provide health care providers with information in an easily usable format to facilitate evaluation allowing an effective course of treatment. (Doc. 4138 ¶ 31). The lack of differential diagnoses and progression through ruling out ailments is pervasive and is shown through the treatment provided to many prisoners. Dr. Wilcox reviewed the following prisoners’ care and made extensive findings. Defendants did not dispute any of them. Given Dr. Wilcox's ten years of experience in this litigation, his expertise and credibility, and Defendants’ failure to address Dr. Wilcox's opinions, the Court adopts Dr. Wilcox's views that the care provided to the following individuals was grossly inadequate.

A. Kendall Johnson

Johnson began her incarceration in 2015. In September 2017, Johnson submitted an HNR complaining her feet and legs had been numb for weeks. (Doc. 4138 ¶ 100). A nurse practitioner assessed her and indicated Johnson should be evaluated to rule out multiple sclerosis vs. idiopathic neuropathy. (Doc. 4270 at 24:18-25). The nurse practitioner, however, did not document Johnson's history or order any diagnostic tests. (Doc. 4270 at 25:13-26:3). Johnson continued to submit HNRs for the next two years and reported her symptoms were getting worse. (Doc. 4138 ¶ 104; Ex. 931 at 1-2). In response, she saw nurses, a NP, and starting in September 2018 a physician, all of whom failed to do proper physical exams or necessary imaging. (Doc. 4138 ¶ 105).

Again, in May 2019, Johnson told a physician she would stumble and fall to the ground and was unable to catch herself. No MRI was ordered, and the physician concluded she did not have MS, which, in Dr. Wilcox's view, was obviously incorrect. (Doc. 4138 ¶ 106). In July 2019, Johnson saw a physician again and recounted an ICS in June when her knees locked up. The physician described her gait with "flopping feet almost as if foot drop" and concluded she was likely suffering from "conversion disorder," which, in essence, means she is delusional. It is an "extraordinarily rare condition ... where somebody believes that they have a disease and they act like they have a disease, but they don't really have the disease." (Doc. 4270 at 26).

Adding to the absurdity, in October 2019, a nurse's encounter note reflected inconsistently "steady and even gait" and "IM [inmate] gait was unstable, IM was holding onto anything she could while she was walking in." (Doc. 4138 ¶ 108). Two months later a physician ordered a diagnostic MRI after documenting that Johnson "walks almost as if you would see in a Frankenstein movie. Very awkward and needs her hands for balance." (Doc. 4138 ¶ 109). Three months later, Johnson received the MRI on January 23, 2020, which "strongly supported a diagnosis of MS." (Doc. 4138 ¶ 110).

Johnson did not see a neurologist until two months later, who recommended additional diagnostic imaging and lab tests and requested that Johnson return in one month. She did not return to the neurologist until November 2020, when she was finally diagnosed with MS. The neurologist recommended a follow up the following month, which did not occur. Rather, Johnson returned to the neurologist at the end of January 2021, who noted "mobility had progressively declined," requiring her to use a wheelchair, as well as urinary incontinence, and that she had not been given an MS medication. (Doc. 4138 ¶ 111). The specialist recommended she be sent to an MS center to develop a treatment plan. Her physician submitted a request to refer Johnson to an MS clinic, noting her "debilitating muscle spasms," but the referral was cancelled because Johnson was prescribed MS medication to be received in the prison. Johnson did not start the medication, Ocrevus, until May 28, 2021. (Doc. 4138 ¶ 108). Had that medication been started earlier, Johnson "might have staved off her more severe symptoms for months or even years." (Doc. 4138 ¶ 113).

Dr. Wilcox met with Johnson on August 31, 2021. He found her "profoundly disabled" and is unable to feed or wash herself, walk, write, and her vision is failing. She is entirely dependent on others to assist her with virtually every activity of daily living. Dr. Wilcox criticized the treating providers who indicated Johnson's symptoms were not suggestive of MS and he opined they were classic MS symptoms. (Doc. 4138 at ¶¶ 100-108). He noted Defendants have Johnson classified as an M3 although she is unable to walk, feed, or care for herself. (Doc. 4138 ¶ 488). It is likely had Johnson received continuity of care by seeing a consistent provider with more ready access to her medical history, her repeated complaints would have resulted in a timely diagnosis and treatment, which would have "staved off her more severe symptoms for months or even years." (Doc. 4138 ¶ 113).

Johnson's testimony at trial was profoundly disturbing. She confirmed she is unable to perform any activities of daily living for herself. She cannot walk, brush her teeth, or wash herself. (Doc. 4256 at 11:19-12:4). She must wear diapers, her vision is impaired, and she can no longer read. (Doc. 4256 at 12:5-10). She testified she paid other prisoners to assist her with self-care activities by ordering things for them from the prison store until three days before her testimony when a medical employee first began to assist her. (Doc. 4256 at 12:11-13:11). Johnson explained she passes time in her housing unit by "count[ing] the ceiling tiles" because she does not go outside. (Doc. 4256 at 13:12-23).

Defendants defend their treatment of Johnson by claiming "ADCRR and Centurion were not, and are not, deliberately indifferent to Johnson's medical needs. To the contrary, they are actively and successfully treating Johnson's multiple sclerosis." (Doc. 4309 ¶ 236). According to Defendants, any delay in diagnosing Johnson was merely negligent.

Johnson's treatment was far from negligent. It was—and may continue to be—a paradigmatic example of the most callous and inhumane indifference. Defendants do not address Johnson's years of attempts without success to seek treatment for her progressing illness. Instead of securing necessary diagnostic treatment, health care staff ignored Johnson's obvious and progressing symptoms, including neglecting her needs up to the day of her pretrial deposition. Only after Johnson testified at her pretrial deposition that she had to pay other prisoners to help her did prison staff begin providing her assistance with her activities of daily living. To conclude that Johnson's treatment has been appalling substantially understates the pain and indifference she has suffered. And in the face of unspeakable hardship, her testimony was clear and compelling. She could explain with precision the evolution of her disease, her futile attempts to receive treatment, and the permanent suffering she endures. Defendants’ failure to rebut this testimony or to attempt to explain why it is an aberration is a clear admission of significant irresponsibility. If medical staff had tested differential diagnoses to evaluate the cause of her symptoms, Johnson's current situation likely would be different.

B. Prisoner A.D.

Last year before trial, in February 2021, A.D., a 44-year-old male, began complaining of severe neck pain. Verbal orders for an EKG, x-ray, and Toradol were noted along with prednisone, diazepam, capsaicin cream, and ibuprofen. (Doc. 4138 ¶¶ 116-117). Dr. Wilcox testified a combination of prednisone, ibuprofen, and Toradol is dangerous in that it could lead to a gastric ulcer or rupture. (Doc. 4138 ¶ 117). Although A.D. reported his pain at a 20 on a scale of 10, a provider, contacted by a nurse, denied the request for diazepam and Toradol and suggested an ice pack and ibuprofen, which he received. (Doc. 4138 ¶ 118). The next day, A.D. was back and seen by a different nurse who determined no referral to a provider or follow-up was needed. (Doc. 4138 ¶ 118). Six days later, an ICS was called for pain and A.D. was scheduled to a see a provider. The following day, A.D. was seen by a nurse practitioner, but she complained that he initiated an ICS "inappropriately." (Doc. 4138 ¶ 119). The nurse practitioner did not perform a neurological exam, did not test reflexes, and did not do an assessment for clonus—a neurological condition causing involuntary muscle contractions—despite A.D.’s claims of limited mobility and twinges of pain to the low back. (Doc. 4138 ¶ 119). On March 4, 2021, a rheumatoid panel was ordered, and the results showed acute inflammation, which strongly suggested infection. The nurse practitioner noted the inflammation but failed to note it was a sign of infection. Additional labs and studies should have been ordered after the test results were received, but the nurse practitioner noted only "[follow-up] in 1-2 wks." (Doc. 4138 ¶ 120).

On March 16, 2021, A.D. was transferred from Florence to Eyman. Over the next two days, three ICSs were called for severe pain that affected his mobility. He was given two Toradol shots and had to be transported via wheelchair or gurney. (Doc. 4138 ¶ 121). On March 19, 2021, a nurse practitioner documented A.D.’s deterioration and increasing neurological dysfunction, including bladder incontinence and inability to ambulate for two weeks. That nurse practitioner sent A.D. by ambulance to a community hospital emergency room. (Doc. 4138 ¶ 122). At the hospital, A.D. was diagnosed with an epidural abscess, which was growing and compressing his spinal cord, causing all of his symptoms. A.D. had surgery and remained in the hospital for two months. (Doc. 4138 ¶ 123).

During the lengthy time the abscess had been allowed to grow, A.D.’s spinal cord was compressed and the nerves began to die. (Doc. 4138 ¶ 123). This resulted in long-term disability. As of September 2021, the close of discovery for trial, A.D. struggled to walk, even using a walker. He probably will never regain normal functioning and will be "substantially impaired the rest of his life." (Doc. 4138 ¶ 124). In Dr. Wilcox's unrebutted opinion, A.D.’s condition was preventable. If Defendants had diagnosed and treated A.D. competently, his disability could have been prevented. Instead, "all [A.D.’s] providers failed him." (Doc. 4138 ¶ 125). It is more likely than not that the failure to test differential diagnoses to understand what was happening with A.D. led to avoidable permanent injuries.

C. Jesse Boldrey

As of 2013, Boldrey weighed 190 pounds. At a chronic care appointment in 2015, he complained of weight loss and weighed only 146 pounds. The only treatment was to provide him a "wasting diet" and liquid supplements. (Doc. 4138 ¶ 33). In 2017, he complained of weight loss again and his records revealed that he had lost another 20 pounds (down to 126 pounds), despite eating all his meals. Boldrey told a doctor he feared he had cancer. Despite obvious warning signs for cancer, the physician ordered simple imaging, stool blood testing, and told him to stop smoking. (Doc. 4138 ¶ 33).

On June 18, 2018, Boldrey again complained of weight loss, but after a chest x-ray and HIV test, the nurse practitioner signed off on his case, continued his wasting diet and ordered liquid nutritional supplements, and scheduled no follow-up. (Doc. 4138 ¶ 34). In December 2018, Boldrey had another chronic care appointment. The records from that appointment do not mention weight loss. (Doc. 4138 ¶ 35). In February 2019, Boldrey submitted an HNR complaining of vomiting, nausea, and bloody bowel movements. Boldrey complained he had lost 10 pounds in a week. When examined, his weight was down to 110 pounds. The nurse practitioner sent Boldrey to the hospital. (Doc. 4138 ¶ 36).

At the hospital Boldrey was diagnosed with a bleeding duodenal ulcer. The hospital obtained CT scan images of Boldrey's abdomen, which showed "scattered nodular opacities in the bilateral lungs concerning for metastatic disease." This was a "new and very serious finding" that demanded follow-up. (Doc. 4138 ¶ 36). When Boldrey returned to prison, a nurse practitioner noted the CT findings but failed to schedule a further consult or follow-up. (Doc. 4138 ¶ 37). And a few weeks later, a doctor discharged Boldrey to return to general population. Inexplicably, the doctor made a note that Boldrey was "doing well." The doctor made no mention of the abnormal lung findings. (Doc. 4138 ¶ 38).

In May 2019, Boldrey twice complained of a very sore throat. In response to the first complaint, the nurse noted his throat was red and neck glands were palpable. The nurse obtained a verbal order for an antibiotic from a nurse practitioner. The provider did not examine Boldrey and prescribing an antibiotic was wrong. (Doc. 4138 ¶ 39). At the second May visit, Boldrey complained of difficulty swallowing. The nurse noted the lung nodules documented in Boldrey's chart and referred him to the provider line on an urgent basis. (Doc. 4138 ¶ 40). On May 29, 2019, the doctor examined Boldrey and stated he suspected cancer. (Doc. 4138 ¶¶ 40-41).

A nurse attempted to schedule a CT of the chest, but the request was denied because "A CT of the chest was performed on 5/23/19," which was wrong. No CT had been ordered or given on 5/23/19. (Doc. 4138 ¶ 42). Likely the reviewing committee had mixed up patients. Despite that denial, somehow Boldrey was given CT scans on May 31 of his brain, abdomen, and neck. The provider who reviewed the results on June 19, 2019, noted probable metastatic disease. The provider noted Boldrey would be scheduled for the earliest possible visit to review the CT results and develop a treatment plan. (Doc. 4138 ¶ 43). That visit did not happen. Instead, Boldrey and his obviously ominous diagnosis were ignored.

Boldrey was not seen until almost three months later when he complained that his throat hurt so badly he could not eat. (Doc. 4138 ¶ 44). At that appointment, Boldrey weighed 100 pounds. Despite the scan results in his file and the loss of 90 pounds from his original weight, the nurse practitioner instructed Boldrey to "eat slow and cut food into small pieces." (Doc. 4138 ¶ 45). The nurse practitioner did request a consult with an ear, nose, and throat specialist because there was a possibility of "neck injury, chronic pharyngitis /mass or tumor." (Doc. 4138 ¶ 45). There was no mention of the CT scan results. (Doc. 4138 ¶ 45).

On October 28, 2019, Boldrey was seen by a physician's assistant. The records from this encounter indicate the notes were completed in a "cut and paste" manner that made no sense. (Doc. 4138 ¶ 46). In one section the PA noted Boldrey allegedly denied weight loss, another section noted Boldrey had "significant weight loss," and yet another section reflected a box indicating "No" regarding "Recent weight/loss/cachexia." (Doc. 4138 ¶ 46). As Dr. Wilcox noted, this level of sloppiness in charting indicates "providers are just cycling patients through these visits as fast as possible with no intent to deliver thorough care." (Doc. 4138 ¶ 46). The lack of adequate staffing affects every interaction with every prisoner.

The ENT appointment requested on September 10, 2019, did not occur until November 5, 2019. At that appointment the ENT diagnosed Boldrey with an otopharyngeal mass suspicious for cancer, and recommended a CT scan, biopsy, and endoscopy . Due to Centurion's policy of not informing prisoners about recommended treatment and possible hospitalization for unidentified "security reasons," Boldrey was not informed of this diagnosis and was unaware of the recommended treatment. On his return to prison, Boldrey inaccurately told a nurse his complaint was throat pain and unintentional weight loss, not the throat and chest cancer he had just been diagnosed with. (Doc. 4138 ¶ 47).

Boldrey was not seen again until November 25, 2019, when an ICS was called for coughing up blood. At that time, Boldrey's oxygen saturation was 77% and an ambulance was called. (Doc. 4138 ¶ 48). At the hospital, Boldrey was diagnosed with malnutrition, lung cancer, cervical lymphadenopathy, hypertension, anemia, and aortic aneurysm. The hospital recommended a referral to oncology for palliative care. (Doc. 4138 ¶ 49). Boldrey returned to prison on December 19, 2019. Upon his arrival, a nurse noted that the liquid nutrition would need to be held "until appropriate tubing is available." (Doc. 4138 ¶ 50). But strangely, a nurse practitioner significantly decreased the recommended pain management. The next day, the physician decided to put Boldrey back on ibuprofen four times a day without protection for his stomach, even though Boldrey previously had a life-threatening GI bleed from NSAIDs. At this time, Boldrey weighed 96 pounds. (Doc. 4138 ¶ 51).

Despite his inability to swallow because of the mass in this throat, Boldrey was prescribed pills (Tylenol #3) for pain. He was also started on another NSAID, Toradol. Thus, he was on ibuprofen and Toradol or "double NSAIDS." (Doc. 4138 ¶ 52). Because he could no longer swallow, Boldrey needed intravenous feeding. A patient receiving intravenous feeding needs weekly electrolytes testing. That testing was not done, leaving Boldrey subject to dangerous swings in electrolytes. (Doc. 4138 ¶ 52).

On December 25, 2019, a physician's assistant indicated Boldrey needed oncology and ENT consults for possible radiation therapy and again recommended palliative care. Centurion cancelled these referrals because Boldrey was designated "Do Not Resuscitate." (Doc. 4138 ¶ 53). This denial requires emphasis. A physician's assistant recommended a palliative care consult, but Centurion cancelled that referral because Boldrey had determined he did not wish to be resuscitated in the event he died. A "Do Not Resuscitate" directive has nothing to do with the need for palliative care. In other words, refusing extraordinary measures in the event of death is not a request to die suffering, in excruciating pain. In short, Centurion "cancelled palliative care for an end-stage cancer patient." (Doc. 4138 ¶ 53). The cruelty of Centurion's behavior is hard to fathom. As Dr. Wilcox testified,

Mr. Boldrey's end of life care does not conform to any standard of care for palliative or hospice care. His cachetic body was racked with pain, and he wasted away with no reasonable assistance from medical science in the form of comfort or compassionate pain control.... [And it was] shocking to see how neglected he was in his end days when he was too weak to advocate for himself.

(Doc. 4167 at ¶ 57).

As of January 2020, Boldrey was still alive, but obviously wasting away. He continued to receive grotesquely inadequate pain medication. (Doc. 4138 ¶ 54). Eventually, on March 16, 2020, medical providers finally ordered IV morphine. (Doc. 4138 ¶ 55). That morphine arrived the following day, but the records show it was administered inadequately. Boldrey was supposed to receive doses four times per day. For the six days prior to his death, he received those four doses on only one day. On one day, he received only one dose. (Doc. 4138 ¶ 56). On March 22, 2020, Boldrey reported his pain was 9/10 and that he had "pain all over." Boldrey died on March 23, 2020. (Doc. 4138 ¶ 57). The mortality review determined the obvious: Boldrey's death could have either been prevented or delayed. (Doc. 4138 ¶ 58).

Unquestionably, Boldrey did not suffer from simple medical malpractice. Basic testing of differential diagnoses would have resulted in a much earlier determination of the cause for his complaints. A functioning health care delivery system would never suffer the magnitude of breakdowns reflected in Boldrey's seven-year odyssey of incompetence, cruelty, and eventual death.

D. Walter Nusz

Nusz, who had long-standing Hepatitis C, died on April 30, 2020 at the age of 60. (Doc. 4138 ¶ 60). On September 21, 2019, he had a lab test showing a fibrosis score of F-3, indicating severe liver scarring. (Doc. 4138 ¶ 61). This should have prompted medical providers to give Nusz liver ultrasounds and beta blockers, a pre-primary prophylaxis for upper GI bleeds, but Nusz was never given a liver ultrasound or beta blockers. (Doc. 4138 ¶ 61). Nusz also had a known platelet count of less than 150,00 per microliter, which should have resulted in routine screening and surveillance for possible esophageal varices via an upper endoscopy at least once per year and to have any varices banded, which would prevent an upper gastrointestinal bleeding episode. Nusz was not given these screenings. (Doc. 4138 ¶ 61).

Nusz had complained of low back pain. (Doc. 4138 ¶ 73). The nurse practitioner overseeing his care managed those complaints with high doses of indomethacin, an NSAID. This medication "is one of the highest risk non-steroidals with respect to causing gastrointestinal bleeding." (Doc. 4138 ¶ 62). Given Nusz's underlying conditions, this medication "was the worst possible choice." (Doc. 4138 ¶ 64).

Between August 17, 2020 and January 15, 2021, Nusz's hematocrit level dropped from 46.3 to 38.2. This lab result suggested Nusz was bleeding internally after starting indomethacin. Nothing was done to address the lab result. (Doc. 4138 ¶ 62). Blood pressure readings also indicated his blood pressure was below the historical baseline, but the providers did not make any note of the issue. (Doc. 4138 ¶ 63). Although Nusz's pain increased after he was prescribed indomethacin, the nurse practitioner tripled the dose on March 23, 2021, without seeing him in person. (Doc. 4138 ¶ 64).

In the 42 days before his death, Nusz submitted five HNRs reporting severe pain, difficulty breathing, bleeding, an inability to eat, an inability to use the restroom, and an inability to walk. In a HNR dated March 24, 2021, Nusz complained he was in significant pain. He stated "This isn't a joke. It's my life [you're] playing with ... Hurry before I die?? Hurry please." (Doc. 4138 ¶ 65). While Nusz recognized his condition was significantly worsening, all the health providers did not. Nusz was seen by at least six different registered nurses, who noted Nusz's troubling symptoms, but none referred him to a provider, and none identified the high dose of indomethacin was causing his symptoms. (Doc. 4138 ¶¶ 67-71). Nusz died on April 23, 2021, one month after the indomethacin was tripled, from a massive gastrointestinal bleed exacerbated by inappropriate indomethacin dosing in a patient with multiple contraindications. (Doc. 4138 ¶ 71).

Clearly Nusz's providers failed to monitor his fibrosis, failed to prescribe preventative medications to minimize the chance of an esophageal bleed, failed to perform preventive upper endoscopies to identify and treat dilated blood vessels in the esophagus, failed to perform blood tests to monitor his coagulation state to see that his ability to clot was significantly impaired, worsened Nusz's conditions by prescribing him a contraindicated does of indomethacin, failed to notice the change in Nusz's lab results (his blood ammonia level was 350 (normal is 15-45) and his ability to clot was 2.06, (normal is less than 1.1)), and did not take action when Nusz repeatedly reported significant changes in his condition. (Doc. 4138 ¶¶ 72-74). Finally, despite the fact that Nusz's care was plainly outside a nurse practitioner's scope of practice, there was no physician level oversight. (Doc. 4138 ¶ 74).

The mortality review could not determine if Nusz's death was avoidable. In Dr. Wilcox's credible opinion, that was "absurd." Medical providers gave Nusz a medication "absolutely contraindicated for people with serious liver disease." Then, when that medication caused significant problems, they increased his dose until his completely predictable death resulted. (Doc. 4138 ¶ 75). Again, this is not a situation where a prisoner had a single interaction with a single provider who did not accurately assess the situation. Instead, Nusz had months of contact, with numerous providers, all of whom failed to identify an obvious problem. The lack of meaningful attempts to diagnose Nusz's ongoing complaints led to his death. Even a marginally functioning health system would not have caused so many providers to get so much so wrong.

E. Timothy Ashing

Ashing, a 30 year-old, submitted an HNR on August 5, 2020, complaining of a "super sensitive" lump on his left testicle that had been there for two months. The nurse who evaluated him used the "musculoskeletal NET assessment" which, obviously, "has nothing to do with the genitourinary system." (Doc. 4138 ¶ 76). Ashing complained of testicular pain, but the nurse engaged only in assessing his handgrip, posture, and whether his gait was symmetrical. She told Ashing to "go easy" on workouts and report if his symptoms worsened. The notes inconsistently indicated that no follow-up was needed and that he would be referred to the provider line. He was not referred. (Doc. 4138 ¶¶ 76-78).

On October 3, 2020, Ashing submitted a HNR stating his left testicle was swollen and he reported severe abdominal pain. He was seen by a nurse a few days later who referred him to Dr. Hines. Upon examination, Dr. Hines documented a 3 cm × 2 cm mass on Ashing's left testicle with tenderness to palpation and swelling. The doctor entered an order for an ultrasound but delayed an entire week to enter the lab orders. (Doc. 4138 ¶ 79). On October 23, 2020, an ultrasound was completed. The notes for that indicated a 2.6 cm left intratesticular mass "highly concerning for testicular malignancy. Recommend an urgent urology consultation." (Doc. 4138 ¶ 80). Dr. Hines reviewed the ultrasound report but noted she would discuss the issue with Ashing "at his next chronic care visit." (Doc. 4138 ¶ 81). Dr. Hines’ lack of urgency is "astound[ing]" and "unethical." (Doc. 4138 ¶ 81). Dr. Hines only ordered a urology consult on an urgent basis. (Doc. 4138 ¶ 82).

Ashing saw a urologist via telemedicine on November 25, 2020, who recommended a radical orchiectomy STAT. Dr. Hines then waited three weeks, until December 13, 2020, to order the orchiectomy on a routine basis. That order was cancelled on January 22, 2021, because the urologist was no longer available. (Doc. 4138 ¶ 82). On February 16, 2021, Ashing saw another urologist. That "visit was essentially worthless" because Centurion did not provide the urologist with the previous ultrasound results or labs. Thus, the new urologist recommended the tests be repeated. (Doc. 4138 ¶ 83).

After some delay, additional labs and another ultrasound were completed on March 16, 2021. A nurse practitioner reviewed a lab result and noted Ashing's Beta HCG level of 81065 mIU/ml was "normal." In truth, the normal range for a male is < 6.51 mIU/ml. Thus, the test actually showed Ashing's result was "astronomically elevated" and "massively abnormal." (Doc. 4138 ¶ 85). In Dr. Wilcox's view, the nurse practitioner's determination the result was "normal" establishes "she did not understand anything about the patient or the disease she was managing."¹⁹ (Doc. 4138 ¶ 85). The result suggested advanced testicular tumor burden. The nurse practitioner, again showing she did not understand, submitted a request for another "possible" urology consult. (Doc. 4138 ¶ 85).

19 Dr. Wilcox also described the nurse's "determination that this is resolved is absolutely incorrect." (Doc. 4271 at 47-48).

The consult request was cancelled on March 30, 2021, when the urologist, who had now obtained the test results, told Centurion that Ashing needed urgent surgery. Ashing had a radical orchiectomy on April 8, 2021, but Centurion did not send Ashing back to the surgeon for post-surgical follow-up. (Doc. 4138 ¶¶ 86-87). In May 2021, Ashing was seen with obvious signs of gastrointestinal bleeding. (Doc. 4138 ¶¶ 89-90). On May 26, he submitted a request saying his stomach was painful and his stools were "dark almost black." (Doc. 4138 ¶ 90). He was seen by a nurse who ordered an x-ray at some unspecified time in the future. (Doc. 4138 ¶ 90). On May 29, Ashing submitted another request stating he "puked up black bio 3 times so far today and my stools are liquid black." (Doc. 4138 ¶ 91). He was seen by a different nurse who documented black stools, dark brown or black vomit, abdominal pain, and elevated heart rate. The nurse noted Ashing looked "very pale, almost jaundiced" but the obvious signs of internal bleeding failed to prompt additional treatment. Rather, the nurse sent Ashing back to his housing with fecal occult blood tests and a container for vomit. (Doc. 4138 ¶ 91). Ashing returned the same day with the vomit container which had "watery vomit with what appeared to be coagulated blood in it." (Doc. 4138 ¶ 92). Given the obvious indications of internal bleeding, Ashing should have been sent to a hospital immediately. Instead, a follow-up nursing appointment was made for the next day. (Doc. 4138 ¶ 92).

The follow-up appointment did not happen on May 30. On May 31, 2021, an ICS was called. The nurse reported Ashing was vomiting, had watery stools "black in color" and Ashing looked "more yellow today than yesterday." (Doc. 4138 ¶ 93). The nurse called a provider, who ordered that Ashing be sent to the emergency room. (Doc. 4138 ¶ 93). At the hospital, Ashing was diagnosed with a widely metastatic tumor and metastasis to the stomach, causing gastrointestinal bleeding. Just a few months before trial, on June 10, 2021, Ashing died after internal bleeding could not be controlled. (Doc. 4138 ¶ 94).

The mortality review found Ashing's death was preventable, there were repeated failures to recognize symptoms, a delay in access to care, and failure to follow-up/identify abnormal test results. (Doc. 4138 ¶ 96). The repeated failures and delays in treating Ashing's cancer "cost him his life." (Doc. 4138 ¶ 96). As explained by Dr. Wilcox, when caught early, testicular cancer "is one of the most curable cancers" and has a success rate of 99% (Doc. 4138 ¶ 96; Doc. 4269 at 78:5-9). The repeated failure to diagnose the obvious would not happen in a functioning health care system.

F. Summary

Defendants believe the treatment and outcomes for Johnson, A.D., Boldrey, Nusz, and Ashing are simply isolated occurrences that do not establish a pattern or practice of providing deficient health care. But the overwhelming evidence shows these cases indicate the opposite. Given the number of encounters each of these individuals had with the medical system, including many different personnel, it is impossible to conclude their treatment represented isolated occurrences. Rather, these outcomes show that if a prisoner develops a serious health condition while in ADCRR custody, he or she is at substantial risk of grievous harm or death due to medical personnel's inability to accurately assess and diagnose such conditions.

A prisoner such as Johnson may present with ongoing and consistent complaints that need expertise to accurately assess and diagnose. Instead, the complaints may be ignored until they are too serious not to require intensive intervention, often when it is too late. Or a prisoner may present with recurrent complaints of pain that, if properly diagnosed, could be remedied with no long-term effects. But such a prisoner may end up like A.D., suffering from lifelong disabilities because multiple providers failed to diagnose him accurately. A prisoner facing death might be in obvious need of pain medication. Instead of providing relief, that prisoner may be treated like Boldrey, ignored and left to waste away in extended horrifying pain. A prisoner might obtain inappropriate medication to treat a condition, such as Nusz did. When that medication causes obvious, life-threatening side-effects, providers might increase the dosage, ensuring death. Or a young prisoner may present with complaints of testicular pain, as Ashing did. Staff might then inexplicably assess his grip strength and, even if some treatment is provided, they will ignore obvious signs of painful distress that merit emergency action. No legitimate humane system would operate in this manner.

V. Defendants’ Challenges to Dr. Wilcox's Opinions Fail

Defendants attempt to undermine Dr. Wilcox's conclusions regarding systemic failures by challenging his methodology. Defendants complain Dr. Wilcox did not perform a random sampling and review of care provided to prisoners.²⁰ As previously explained, Dr. Wilcox's methodology, while not random, was appropriate and consistent with the practice by other experts. Again, as Dr. Wilcox noted, healthy patients are not an indicator of a system's capabilities. "It is the complex patients who test the capacity of staff and systems alike." (Doc. 4167 ¶ 25). If the issue is whether prisoners are provided adequate medical care, it makes sense to look to prisoners with serious needs and, in particular, those who suffer adverse outcomes, not those prisoners who have no contact with the medical system.

20 During cross-examination, Dr. Wilcox was asked "how many examples of delays in obtaining specialty care do you need to see in a system of 28,000 patients for you to make the determination that it's systemic?" Dr. Wilcox responded "Well, I don't know that there's an actual magic answer for that, but you certainly see patterns with regards to those delays in the medical records that I reviewed." (Doc. 4271 at 27).

Defendants further attempted to counter Dr. Wilcox's opinions through their own expert, Dr. Murray. To assess the quality of care provided to all prisoners, Dr. Murray relied on Healthcare Effectiveness Data and Information Set ("HEDIS") metrics, which were developed by the National Committee for Quality Assurance ("NCQA"). Dr. Murray reviewed data from eOMIS utilizing the HEDIS parameters so he could evaluate the quality of health care across the ADCRR system and make comparisons between ADCRR and other health care systems. (Doc. 4206 ¶ 998). Dr. Murray selected the HEDIS metrics regarding only blood pressure control and diabetic care for review because, in his view, they represent two of the largest patient populations in both correctional and non-correctional settings. (Doc. 4286 at 31:6-32:13, Doc. 4286 at 123:21-124:4). Dr. Murray determined ADCRR's performance on HEDIS measurements for diabetes and hypertension are higher than those provided by private health plans. (Doc. 4206 ¶ 1077). Dr. Murray concludes these scores are "reflective of a health system operating efficiently and focused on best care for the patients." (Doc. 4206 ¶ 1077). In other words, Dr. Murray concludes allegedly adequate testing and treatment of diabetes and hypertension, once those conditions have been accurately diagnosed, establish the entire prison health care system is operating effectively.

There are two problems with Dr. Murray's method and conclusion. First there is no reliable evidence that selecting only these two criteria—treatment of hypertension and diabetes—establishes the overall capabilities of a health care system like ADCRR's. The second is more fundamental. Dr. Murray's assessment assumes ADCRR adequately diagnoses and tracks the prisoners who suffer from hypertension and diabetes. And Defendants offered no evidence to confirm this unique theory. Dr. Murray necessarily only looks to data on prisoners who have previously been diagnosed with diabetes and hypertension. There was no evidence showing this data represents the entire population of prisoners actually suffering from these ailments and of course there could be prisoners with undiagnosed hypertension or diabetes. Certainly Defendants could have attempted to prove they had identified all such prisoners. But Defendants did not do so. In fact, given the level of care provided to prisoners discussed previously, it is overwhelmingly likely Dr. Murray's data does not include the entire relevant population. Dr. Wilcox specifically addressed this possibility during his testimony, indicating

My concern with the chronic care list and how they're derived and managed is that I found in my review of the medical records that there frequently were ... critical diagnoses missing from the master problem list, which is most likely how they're generating the reports for the chronic care list.

(Doc. 4272 at 83:19-24). Further, every witness who testified, including Dr. Murray, agrees ADCRR's electronic health record system, eOMIS, is fatally limited and flawed. The inadequacy of eOMIS is yet another reason to conclude ADCRR's records do not accurately represent all relevant prisoners or accurately represent prisoner health care provided.

If that were not enough, Dr. Murray's own opinions about care provided to prisoners establish HEDIS scores are not a valid basis for assessing ADCRR's system. In an attempt to "validate" his HEDIS findings, Dr. Murray performed a random review of medical records of ten prisoners with at least two chronic condition diagnoses at all complexes except for Phoenix and Florence. Prisoners’ charts were evaluated for quality of documentation, quality of chronic care, and quality of episodic care. A consultant reviewed the charts using the scale set forth in Appendix 2.²¹

21 Dr. Murray did not independently review the patient's medical records prior to preparing his expert report.

Dr. Murray created this unproven scale himself for evaluation of only ADCRR's health care system. Under this eccentric evaluation method, even "excellent" charts could reveal a lack of preventative health care and very good charts may show delayed treatment. Of greater concern are the charts considered "good, fair, or poor." In particular, the explanations for charts that were labeled "good" establish "good" was often profoundly insufficient. The following are several examples of chronic care considered "good" by Dr. Murray's consultant.

Patient 1 in Tucson is an 81-year-old male who suffers from multiple chronic conditions including hypertension, diabetes, coronary artery disease, peripheral arterial disease, and other serious conditions. The chart review reflects "chronic care visits sometimes addressed all of the patient's problems and sometimes did not." (Doc. 4206 ¶¶ 214-217). Further, his chart revealed a "notable lack of detailed foot exams ... with (later) documented severe peripheral arterial disease." (Doc. 4206 ¶ 221). There are no notes of an exam of the prisoner's feet, although there is a note that he "has pedal pulses [pulse found on top of a patient's foot]," which Dr. Murray describes as "interesting." (Doc. 4206 ¶ 221). The Court interprets the chart as documenting a finding of pedal pulses when the health care provider did not actually perform an exam. Dr. Murray muses that "one could wonder if earlier foot exams would have identified an issue" before Patient 1 developed gangrene and lost a toe to amputation. (Doc. 4206 ¶ 221). The summary of Patient 1's chronic care noted "a lag in treatment of some problems and a lapse in treatment of various issues." (Doc. 4206 ¶ 224). This is not "good" health care. Rather, this affirmatively documents health care that presents a substantial risk of serious harm to this patient. The failure to conduct foot exams for a patient such as this creates a possibility of a seriously adverse outcome.

Patient 4 in Yuma is a 78-year-old male with multiple chronic conditions. (Doc. 4206 ¶ 363). In 2015, the provider documented Patient 4 had been experiencing chest pains monthly. (Doc. 4206 ¶ 366). But despite an abnormal EKG, nothing shows follow-up on "a very great pre-test probability of having heart disease." (Doc. 4206 ¶ 366). Also in 2015, the provider noted Patient 4 needed colon cancer screening because of a positive fecal occult blood test and low hemoglobin readings but no follow-up ever occurred. (Doc. 4206 ¶ 367). This is not "good" health care because it overlooks obvious, potentially serious, conditions that might lead to death. In reviewing Dr. Murray's scale as applied to this prisoner, the failure to follow-up on potential heart disease or abnormal colon cancer screenings would "markedly increase" the risk of harm, indicating the score of "good" is inaccurate.

Patient 2 in Douglas is a 44-year-old man with hypertension and a prior history of lymphoma. (Doc. 4206 ¶ 459). Patient 2 was seen by a nurse practitioner in September 2020 to address persistent headaches following a head injury ; the nurse practitioner noted his blood pressure was 135/95 and prescribed a low-dose calcium channel blocker to address his blood pressure without lowering his already-low heart rate. The medication prescribed, however, was "inappropriate" and "should be avoided" if there is a concern about bradycardia. (Doc. 4206 ¶¶ 461, 464). Hypertension was officially added to Patient 2's problem list a few weeks later, but an EKG was not ordered for another year. (Doc. 4206 ¶ 461). This is not "good" health care but merely another example of care that creates a substantial risk of serious harm.

Patient 3 in Winslow is a 55-year-old male with multiple chronic conditions including cirrhosis and thrombocytopenia (low platelet count). (Doc. 4206 ¶ 619). While housed at Eyman, staff refrained from ordering NSAIDs, which are contraindicated for Patient 3. (Doc. 4206 ¶ 624). But when he was transferred to Winslow, Patient 3 was prescribed NSAIDs, despite the notations acknowledging cirrhosis, low platelets, and "needing to add [chronic kidney disease ] as a diagnosis." (Doc. 4206 ¶ 624). Patient 3 soon developed gastrointestinal bleeding, fecal tests were positive for bleeding, but no rectal exam was performed. (Doc. 4206 ¶ 625). Obviously this is not "good" health care. Patient 3 is at substantial risk of serious harm.

Patient 7 in Perryville is a 59-year-old female with diabetes, hypertension, and alcoholic cirrhosis. (Doc. 4206 ¶ 757). A July 2020 ultrasound showed "coarse echotexture" of her liver, splenomegaly, and a slightly dilated inferior vena cava. She had emergency nurse visits for abdominal pain in 2020. (Doc. 4206 ¶ 759). The consultant notes that Patient 7 has "an elevated risk of esophageal variceal bleeding," but "she has not been referred to a gastroenterologist." (Doc. 4206 ¶ 762). This is not "good" health care but an example of a health care system that is affirmatively ignoring conditions that present a substantial risk of serious harm.

Patient 4 in Safford is a 42-year-old male with several chronic conditions. He is Hepatitis C positive but there is no "management plan [ ] for HCV." (Doc. 4206 ¶ 824). He "should be considered to have significant, active liver disease with fibrosis, if not cirrhosis," (Doc. 4206 ¶ 824), but that was not noted as a problem, and he was prescribed contraindicated NSAIDs. (Doc. 4206 ¶ 824). This is not "good" health care but another prisoner who is at risk of serious harm.

Patient 4 in Eyman is an 81-year-old male with multiple chronic conditions. (Doc. 4206 ¶ 915). He was diagnosed with rheumatoid arthritis during his incarceration. (Doc. 4206 ¶ 918). The consultant states Patient 4 should be referred to a rheumatologist and concludes his "RA has not been well-managed." (Doc. 4206 ¶ 923). This is not "good" health care. Instead, the system has consigned Patient 4 to live with pain and limitations that could, with treatment, be alleviated.

These examples from 7 out of 8 complexes reviewed do not "validate" Dr. Murray's HEDIS findings or his conclusion that the chart reviews establish a functioning system. Rather, these examples show that "good" scores reflect patients in ill-health and receiving deficient chronic care treatment.

There are additional takeaways from these chart reviews. 54 of the 79 prisoners reviewed received scores of Good, Fair, or Poor—67% of the total.²² By Dr. Murray's own scoring system, a rating of Good presents highly questionable care, as outlined above. And scores of Fair or Poor present care involving obvious substantial risks of serious harm.

22 The Court also discovered several errors in Dr. Murray's chart reviews. Tucson only listed nine patients instead of ten (Doc. 4206 ¶¶ 214-315); Patient 5 in Tucson received a rating of Poor for chronic health care, but the score of 2 did not match the rating (Doc. 4206 ¶ 272); and Patient 10 in Winslow was determined to be not-applicable because he did not have any chronic care diagnoses, which is puzzling because the criteria for review of a patient was active chronic care diagnoses. (Doc. 4206 ¶¶ 700-701).

Analyzing Dr. Murray's chart reviews does not reveal a functioning health care system. It reveals a system with an electronic records system that everyone agrees is abysmal and must be replaced, a system that does not allow referral of patients who exhibit serious symptoms to providers, and a system that administers contraindicated and potentially deadly medications. Accepting Dr. Murray's belief that his sample was representative of ADCRR health care as a whole, the only possible conclusion is the system puts prisoners at a substantial risk of harm.

In addition to inadvertently establishing the risk of harm to prisoners through the chart review evaluation he chose to conduct, Dr. Murray did not evaluate the core deficiencies Dr. Wilcox identified, including those related to medication administration, specialty care, hospitalizations, discharge after hospitalizations, sick call, nursing care, and access to providers. Dr. Murray testified:

Question : So, Doctor, you didn't do a similar evaluation or study or what you can call -- what you -- what Dr. Baillargeon²³ called critically important analysis on a number of other components of the healthcare system, such as medication administration; correct?

Dr. Murray : Not a direct study, no.

Question : You didn't do one on specialty care either?

Dr. Murray : Other than what we came across in the medical records, but not a special -- not a special study on that, no.

Question : And you didn't do one on hospitalizations?

Dr. Murray : No.

Question : Or discharge after hospitalizations?

Dr. Murray : Not a special study, no.

Question : Sick call?

Dr. Murray : We did not do a special study on sick call.

Question : You didn't do a study about patients who didn't get to see a provider?

Dr. Murray : We didn't do a special study on that, no.

Question : Utilization review, you didn't do that either?

Dr. Murray : Not a special study, no.

Question : Language interpretation?

Dr. Murray : Not a special study, no.

Question : Emergency care?

Dr. Murray : Not a special study.

Question : Nursing care?

Dr. Murray : Not a -- not a specific study, no.

Question : Radiology?

Dr. Murray : No, not a specific study.

Question : Preventive care?

Dr. Murray : Not a specific study.

23 Dr. Jacques Baillargeon is an epidemiologist at the University of Texas Medical Branch with whom Dr. Murray works and consulted when developing his opinions for this case. (Doc. 4286 at 53).

(Doc. 4286 at 65:21-66:21). Dr. Murray use of HEDIS scores and his chart reviews have not established a well-functioning health care system at ADCRR. And of crucial importance, by offering and relying on Dr. Murray as their sole witness to establish a functioning health care system, Defendants effectively demonstrate satisfaction with the existing, completely highly inadequate system.

Dr. Murray's view that the health care system is operating appropriately is further undercut by the scores for Defendants’ Performance Measures on fundamental components of any functioning health care system. These scores were monitored and reported by Defendants and are therefore undisputed evidence of random samples drawn from the complexes during 2021. The following scores reflect the success percentage for the monitored files at each complex.

Performance Measure 50 required urgent specialty consultations and urgent specialty diagnostic services be scheduled and completed within 30 calendar days of the consultation being requested by the provider. (Doc. 1185-1, Ex. B at 11). The table below reflects failures across the system but establishes some locations where "urgent" specialty consultations repeatedly did not occur within 30 days. The highlighted numbers are those that were less than the 85% performance required under the Stipulation.

Performance Measure 50 Jan. Feb. March April May June July 2021 2021 2021 2021 2021 2021 2021 Douglas 85 100 100 100 100 100 100 Eyman 58 76 69 82 86 90 100 Florence 69 83 74 81 79 85 93 Lewis 65 70 64 88 77 87 88 Perryville 85 90 81 95 89 89 97 Phoenix 75 100 100 100 100 100 100 Safford 100 100 80 100 100 100 100 Tucson 59 93 89 92 97 95 100 Winslow 100 86 100 100 100 100 100 Yuma 57 88 84 77 77 97 58

(Ex. 1263).

Performance Measure 51 required routine specialty consultations be scheduled and completed within 60 calendar days of the consultation being requested by the provider. (Doc. 1185-1, Ex. B at 11). The table below establishes prisoners were waiting at least two months to obtain even routine consultations.

Performance Measure 51 Jan. Feb. March April May June July 2021 2021 2021 2021 2021 2021 2021 Douglas 86 97 95 100 100 100 100 Eyman 76 68 92 92 98 96 90 Florence 63 90 87 93 88 96 95 Lewis 60 82 83 84 83 92 93 Perryville 81 76 72 83 86 89 88 Phoenix 84 100 88 100 100 95 100 Safford 87 97 97 93 100 97 100 Tucson 80 83 85 90 94 95 93 Winslow 97 90 90 96 77 97 79 Yuma 72 86 86 80 88 88 86

(Ex. 1264).

Performance Measure 44 required prisoners returning from the hospital have the hospital's treatment recommendations reviewed and acted upon by a medical provider within 24 hours. (Doc. 1185-1, Ex. B at 11). The failures on this measure reflected below are inexplicable if the system was properly functioning. This Performance Measure involved some of the sickest prisoners in the system. As illustrated by Eyman, at times, there was only approximately a 50% chance a prisoner's discharge recommendations would even be reviewed, much less acted on, within 24 hours.

Performance Measure 44 Jan. Feb. March April May June July 2021 2021 2021 2021 2021 2021 2021 Douglas 100 86 100 100 75 100 100 Eyman 52 56 69 78 56 68 57 Florence 88 66 71 61 87 53 71 Lewis 81 88 66 72 71 76 76 Perryville 82 79 88 95 81 92 82 Phoenix 50 50 100 100 100 100 0 Safford 100 100 N/A 100 100 100 100 Tucson 74 60 59 67 65 34 66 Winslow 55 100 100 83 100 100 100 Yuma 40 80 38 80 85 65 70

(Ex. 1259).

Performance Measure 13 required chronic care and psychotropic medication renewals be completed such that there was no interruption or lapse in medication. (Doc. 1185-1, Ex. B at 8). The numbers reflected below establish Perryville, Tucson, and Yuma, in particular, were simply unable to provide medication refills in a timely manner.

Performance Measure 13 Jan. Feb. March April May June July 2021 2021 2021 2021 2021 2021 2021 Douglas 97 100 100 100 93 96 96 Eyman 86 84 84 80 92 88 92 Florence 88 65 77 86 88 88 75 Lewis 69 95 93 99 86 91 77 Perryville 84 82 86 79 74 73 86 Phoenix 94 91 91 97 98 95 100 Safford 90 90 100 90 100 97 100 Tucson 90 94 94 82 67 70 79 Winslow 100 100 100 100 93 100 100 Yuma 64 74 78 74 68 76 96

(Ex. 1256).

Performance Measure 37 required prisoners be seen by an RN within 24 hours after an HNR is received (or immediately if identified with an emergent need, or on the same day if identified as having an urgent need). (Doc. 1185-1, Ex. B at 8). The numbers below establish prisoners experienced significant delays in being seen by a nurse. Given that nurses are treated as a mandatory first step in obtaining care, any delay in seeing a nurse will necessarily result in a significant delay in seeing a higher-level provider.

Performance Measure 37 Jan. Feb. March April May June July 2021 2021 2021 2021 2021 2021 2021 Douglas 85 100 100 100 100 100 100 Eyman 58 76 69 87 86 90 100 Florence 69 83 74 81 79 85 93 Lewis 65 70 64 88 77 87 88 Perryville 85 90 81 95 89 89 97 Phoenix 75 100 100 100 100 100 100 Safford 100 100 80 100 100 100 100 Tucson 59 93 89 92 97 95 100 Winslow 100 86 100 100 100 100 100 Yuma 57 88 84 77 77 97 58

(Ex. 1258).

Dr. Murray did not address any of Defendants’ own Performance Measure scores or explain why they do not undermine his claims of a functioning health care system.

VI. Medical Care Conclusion

There is patently insufficient, and more often than not incompetent, health care staffing to adequately meet prisoners’ needs. The majority of medical care staff do not have necessary training or licensure to provide the type of care that is necessary to provide constitutionally adequate care. This is a completely ineffective and toxic combination. The patterns of delay and indifference are pervasive in ADCRR's own mortality reviews. Defendants have refused to remediate these practices to comply with their constitutional obligations.

MENTAL HEALTH CARE

In addition to serious medical care needs, it is undisputed Arizona's prisoners suffer from serious mental health needs. Defendants’ records reflect 8,548 prisoners on the "mental health caseload," meaning they have scores of MH-3 or above and require ongoing treatment for an active mental health diagnosis. (Ex. 3326). This large population undoubtedly is difficult to manage. However, Defendants’ failure to hire sufficient competent staff means prisoners with serious mental health needs are ignored.

As with medical care, the parties presented competing experts on the issue of mental health care. Again, Plaintiffs’ expert was far more credible. He testified "it's hard for me to adequately express how significantly ill the individuals that I encountered are. They were among the most mentally ill individuals that I have seen throughout my 40 years of being a psychiatrist." (Doc. 4260 at 36:2-6). The evidence at trial established Defendants’ system for the provision of mental health care creates a substantial risk of harm created by their refusal to hire adequate staff to provide necessary mental health care.

I. Overview of Expert Testimony

A. Dr. Stewart

Plaintiffs’ expert was Dr. Pablo Stewart. He earned a bachelor's degree in chemistry from the U.S. Naval Academy in 1973 and, after serving in the U.S. Marine Corps for five years, earned a medical degree from the University of California, San Francisco in 1982. (Doc. 4109 ¶ 2). After graduating from medical school, Dr. Stewart received the Mead-Johnson American Psychiatric Association Fellowship and became board-certified in psychiatry. (Doc. 4109 ¶ 2). Dr. Stewart currently serves as a clinical professor and psychiatrist at the Burns School of Medicine at the University of Hawai‘i. (Doc. 4109 ¶ 1). He serves as an attending psychiatrist at the Oahu Community Correctional Center and supervises psychiatry residents. (Doc. 4109 ¶ 1). Dr. Stewart has served as Plaintiffs’ mental health care expert in this case since 2013.

In preparing for his trial testimony, Dr. Stewart reviewed current ADCRR and Centurion policies, procedures, and practices; reviewed numerous documents and class members’ medical charts; and conducted on-site inspections and class member interviews in September 2021 at Eyman, Perryville, Phoenix, and Tucson. During those inspections, he visited housing units where people classified SMI are incarcerated, units designated for people with mental health needs (regardless of classification), mental health watch units, isolation units including maximum custody and detention units, and interviewed class members incarcerated in these units. (Doc. 4109 ¶ 8). Dr. Stewart also followed up with the people he originally evaluated in 2013. (Doc. 4260 at 16). All documents Dr. Stewart reviewed are listed in Exhibit 4 to his written declaration. (Id. ¶ 15; see Doc. 4109-1, Ex. 4).

B. Dr. Penn

Dr. Joseph V. Penn testified for Defendants as their mental health care expert witness. He earned a bachelor's degree in biology from the University of the Incarnate Word in San Antonio, Texas in 1987. (Doc. 4172-1, Ex. 1). He earned his medical degree from the University of Texas Medical Branch in 1992. (Doc. 4172 ¶ 9). After graduating from medical school, Dr. Penn completed a fellowship in Forensic Psychiatry from Yale University. (Doc. 4172 ¶ 9). Since 2008, Dr. Penn has served as the Director of Mental Health Services for UTMB Correctional Managed Care. (Doc. 4172 ¶ 12). Dr. Penn has served as Defendants’ mental health care expert in this case since 2013. (Doc. 4172 ¶ 2).

Dr. Penn did not interview any class members.²⁴ Instead, he generated a random selection of charts from prisoners at ADCRR's facilities that house prisoners with a MH score of MH-3 or higher: Eyman (15 prisoners); Florence (15 prisoners); Lewis (15 prisoners); Perryville (30 prisoners); Phoenix (15 prisoners); Tucson (15 prisoners); and Yuma (15 prisoners). (Doc. 4172 ¶ 48; see Doc. 4174, Ex. 4). Dr. Penn also conducted tours of facilities that house prisoners on the mental health caseload: Phoenix, Eyman, Lewis, Perryville, Yuma, and Tucson. Dr. Penn met with each facility's health care leadership. (Doc. 4172 ¶ 52). All documents Dr. Penn reviewed are listed in Exhibits 2 and 3 to his written declaration. (Id. ¶ 47; see Doc. 4174-1, Exs. 2-3).

24 Dr. Penn stated he "was not allowed to actually interview or evaluate any inmate[.]" (Doc. 4289 at 28:1-2). This is untrue. Upon cross-examination Dr. Penn explained Defendants’ counsel told him he "was prohibited from speaking to inmates because they said the plaintiffs didn't want me talking to inmates unless they were present." (Doc. 4283 at 11:21-24). Dr. Penn chose not to speak to the prisoners in the presence of their counsel.

II. Staffing is Chronically Insufficient

As with medical care, Plaintiffs’ primary claim regarding mental health care is inadequate staffing. Therefore, the Court evaluated whether sufficient mental health care staff are employed to address Arizona's prisoners’ serious mental health needs. As stated, the contract between ADCRR and Centurion requires 1052.75 FTEs. (Doc. 4287 at 17:11-12). This overall staffing allocation includes health care, mental health care, dental care, and administrative staff. Plaintiffs contend the mental health care staffing reflected in the contract is too low and, because of current vacancies, many complexes are substantially understaffed. Defendants contend current staffing levels, even though they are markedly beneath the levels contemplated by the contract, are sufficient to meet their constitutional obligations.²⁵

25 Defendants concede the staffing levels are too low and below contract requirements, but they argue they cover the deficiencies with overtime.

Centurion provides the following mental health care staff positions: Behavioral Health Technician; Behavioral Specialist; Mental Health Lead; Mental Health Clerk; Mental Health Midlevel (NP/PA); Mental Health Registered Nurse; Psychiatrist; Psychologist; and Psychology Associate. The tables at Appendix 3 represent the number of mental health care staff required by the contract, the number of hired staff, and the difference between those two numbers. These figures are as of August 2021. (Ex. 2167).

As of August 2021, only 149.55 mental health care staff were hired in contrast with 199.00 required under the contract with Centurion, reflecting only 75% of positions are filled. Every location is understaffed except for Winslow and Safford, which each have only one psychology associate. Similar to medical care, Defendants have structured the system to have the vast majority of care provided by lower-level individuals. Thus, the majority of individuals providing mental health care are behavioral health technicians, nurses, or psychological associates.

Based on the contract staffing level, at some point in time Defendants decided 199.00 mental health positions were actually necessary to provide adequate care to the prisoner population. That staffing level has never been achieved. (Doc. 4264 at 99:17-24). But the critical issue is not the contract staffing level. Rather, the issue is whether the actual staffing numbers, which are far below the contract staffing level, present a substantial risk of serious harm to prisoners. Dr. Stewart testified the number of filled positions "is abysmally low." (Doc. 4109 ¶ 20). In contrast, Dr. Penn concludes "ADCRR's contracted mental health services vendor, Centurion, has adequate numbers and types of mental health staffing." (Doc. 4172 ¶ 66). The Court first analyzes Dr. Penn's position.

Dr. Penn testified there is adequate staffing to meet the prisoner population's needs. But his opinion is not based on an accurate and careful analysis of the population's needs. Instead, Dr. Penn claimed it would be inappropriate to require a specific staffing formula because doing so would deprive the system of "flexibility." (Doc. 4289 at 40:16-22). He claims "staffing levels should be assessed for a particular system based upon a clinical determination that adequate mental health services may be, or are, provided by a particular number of various types of professional staff." (Doc. 4172 ¶ 65). In Dr. Penn's opinion, NCCHC's allegedly straightforward staffing standard makes "perfect sense"—the health care system or facility must merely have adequate staff to provide the services required or needed. (Doc. 4289 at 41:9-17). This theory is simplistic. When applied to ADCRR, it makes no sense.

According to Dr. Penn, because ADCRR created a specific staffing requirement of approximately 199 FTEs, that number should automatically be accepted as adequate without challenge. That is circular logic. Dr. Penn never testified to explain how ADCRR developed its staffing allocation or why that arbitrary number was deemed sufficient to "provide the services required or needed." Based on the evidence, it is nothing more than a number pulled out of thin air. This conclusion again encapsulates in one stroke the absurdity of ADCRR's reliance on NCCHC certification to establish the system is functioning well. Under Dr. Penn's theory, if ADCRR decided that one mental health staff position was needed for the entire prisoner population, then that decision could not by challenged. But ADCRR's unexplained, unsupported, and unreliable opinion regarding the adequacy of staffing for the entire prison mental health care system is contrary to the evidence.

ADCRR failed to present evidence demonstrating that its contract staffing allocation was based on an actual reliable assessment of the need for services. In fact, many witnesses, including those from ADCRR, testified that more mental health care staff beyond the contract staffing level are needed to provide adequate mental health care. Dr. Stefanie Platt, Centurion's former regional mental health director for ADCRR, testified she raised this concern to her superiors.

Question: And if even 100 percent of the contracted positions were filled, was there still more patient need than the staffing plan could meet?

Dr. Platt: In my opinion, yes.

Question: Did you raise your concerns about this, about the adequacy of the contract, with anybody?

Dr. Platt : Yes.

Question: With whom?

Dr. Platt: Dr. Carr, Mr. Dolan.

(Doc. 4264 at 100:2-10). Dr. Antonio Carr is Centurion's Statewide Psychiatric Director and Tom Dolan is Centurion's Vice-President of Operations in Arizona. Dr. Platt further testified mental health staff at the "vast majority" of complexes, including Eyman, Phoenix, Lewis, Florence, Perryville, and Tucson, told her of their concerns about caseloads being too high. (Doc. 4264 at 105:4-96). During his Rule 30(b)(6) deposition Dr. Carr confirmed staff repeatedly told him additional staff were needed. (Doc. 4146-1 at 47:16-19) (testifying that "it's not uncommon for a mental health nurse or psychology associate to say, ‘Hey, Carr, I think we need more staff’ "). This was corroborated by Dr. Pennington-Stallcup, ADCRR's Mental Health Program Director.

During Dr. Pennington-Stallcup's cross-examination, Plaintiffs introduced an August 27, 2020 email Dr. Pennington-Stallcup sent to Defendant Gann, wherein she expressed concerns about vacancies in mental health care provider positions. The staffing numbers reflected in Appendix 4 were referenced in the email.

Based on these numbers, Dr. Pennington-Stallcup noted there are "significant vacancies in clinical mental health staff who are primarily responsible for providing [ ] behavioral health contacts." (Doc. 4276 at 93:25-94:3). Later in another email to Dr. Platt on November 9, 2020, Dr. Pennington-Stallcup wrote, "I'm really concerned about the vacancies at Florence and Eyman given the high-need and high-risk patient populations." (Doc. 4276 at 95:15-19).

During her direct testimony, however, Dr. Pennington-Stallcup retreated from her original concerns about staffing, stating she was no longer concerned. (Doc. 4276 at 90:2-8). She attempted to explain by testifying that she performed a "new analysis" based on allocated staffing. Without explanation, she claimed during trial if all the positions were filled, mental health caseloads would be manageable. (Doc. 4276 at 100:21-24). This "new analysis" is incredible and nonsensical.

Comparing clinical mental health staffing levels from August 2020 until August 2021 shows the "significant vacancies" that originally concerned Dr. Pennington-Stallcup have not meaningfully changed, and some are worse. More importantly, Dr. Pennington-Stallcup never explained why it would be reasonable to rely on staffing levels that have never been achieved. Accordingly, Dr. Pennington-Stallcup's new opinion is unreliable and not supported by evidence. In contrast, Dr. Stewart's opinion which is based on analysis of the substantial evidence before the Court that "systemic deficiencies in the provision of mental health care are rooted, in whole or in part, in ADCRR's chronically inadequate health care staffing," is credible. (Doc. 4109 ¶ 32).

III. Inadequate Staffing and Failure to Provide Adequate Mental Health Care

The lack of mental health staff negatively impacts the ability to provide adequate mental health care. The staffing allocation reflects only seven allocated psychiatrist positions for the entire ADCRR population, and the evidence reflects only two of them physically work on-site. (Ex. 1531 May 2021 Staffing Report, Native Format at "All Staff" tab showing "Psychiatrist TH" [Tele-Health] at Eyman, Florence, Lewis, Perryville, Tucson, and Yuma, and two "Psychiatrists" (without the TH qualifier) at Phoenix and Yuma complexes). Dr. Stewart specifically referenced Eyman's two psychologists and stated he does not "see how one or two psychologists could even begin to address the needs at that [facility]." (Doc. 4260 at 21). The number of psychologists on staff varies from 0 (Douglas, Safford, Winslow, and Yuma), to 0.9 (Florence), to 1.75 (Perryville), to 2 (Eyman, Lewis), to 2.5 (Phoenix), to 3.9 (Tucson).

In addition to midlevel providers, psychology associates act as providers. But there are 14 psychology associates (out of 59) who are listed as not being licensed, including 4 at Eyman, 2 at Florence, 3 at Lewis, 1 at Perryville, 2 at Phoenix, and 2 at Yuma. Dr. Stewart explains that he observed a "disturbing pattern" of "mental health staff without pharmacological training serv[ing] as de facto gatekeepers of patients’ access to psychiatric prescribers." (Doc. 4109 ¶ 22). Without sufficient mental health providers with appropriate training, the outcome is the inability to provide varying levels of support for mental health conditions.

A. Inpatient Treatment Units and Inadequate Treatment at Existing Units

Defendants have inpatient units for prisoners with the most serious and intensive mental health needs. Dr. Stewart's opinion is the current system reflects a lack of inpatient intensive mental health care. For example, ADCRR provides inpatient psychiatric treatment for prisoners with severe mental illness at the Phoenix complex. A review in September 2021 just before trial reflected only one-third of the beds at that complex were in use. (Doc. 4109 ¶ 46). Dr. Pennington-Stallcup testified the inpatient treatment units are operating below capacity because, allegedly, there are not enough "patients in the system who need those beds." (Doc. 4277 at 34:15-20). During trial, however, the evidence showed prisoners needing significant inpatient treatment who were denied such treatment.

Dr. Pennington-Stallcup was specifically asked about a prisoner housed in Eyman, a non-mental health treatment unit. On October 15, 2020, a mental health provider indicated that the prisoner "appears to need more mental health resources than are available at this location. He is a good candidate for a referral to the [residential treatment unit]." (Doc. 4277 at 34:14-35:3). (Ex. 2401). While not entirely clear, it does not appear he was ever transferred to an inpatient facility. It was clear, however, that approximately one year later in 2021 this prisoner was housed in the same unit he had been housed in as of October 2020. And Dr. Stewart identified other prisoners that needed inpatient level care but were not being housed in such facilities or their transfer to an inpatient facility had been delayed. The explanation that follows of those prisoners’ diagnoses as well as Dr. Stewart's opinion illustrates that Defendants have failed to provide inpatient treatment for those who seriously needed it.

T.A. is designated as seriously mentally ill and has a history of self-harm. Dr. Stewart interviewed him when conducting a tour of the Eyman complex in September 2021. T.A. was on continuous watch and was laying on the floor. Dr. Stewart noted he was "extremely disheveled and reported experiencing command hallucinations" telling him to hurt himself. In June and July 2021, T.A. had repeatedly cut or otherwise harmed himself with ICS responses being called on a regular basis. On one day, July 18, three ICS were called for self-harm. On July 26, 2021, T.A. was found hanging with both feet off the ground. After he was discovered, he was placed on suicide watch. The assessment and diagnosis sections of the notes were blank. (Doc. 4109 ¶ 51). Dr. Stewart reliably testified T.A. needed to be "urgently transferred to an inpatient hospital setting." (Doc. 4109 at ¶ 51).

D.V. was on suicide watch at the Eyman complex on September 8, 2021, when Dr. Stewart encountered him. D.V. informed Dr. Stewart he previously took the mood stabilizer Trileptal, but it was discontinued because staff told him "inmates abuse it." (Doc. 4162 at ¶ 52). D.V. has a history of cutting himself, with one incident requiring he be sent to an outside hospital because he lost a significant amount of blood. Dr. Stewart was so alarmed by D.V. that he requested Plaintiffs’ counsel notify ADCRR and Centurion immediately that D.V. required transfer to an inpatient mental health care. (Doc. 4109 ¶ 52). Dr. Stewart's urgent recommendation was ignored and the following week D.V. cut himself again. A psych associate prepared a treatment plan the next day that failed to address possible medication modification, or transfer to inpatient mental health care. (Doc. 4109 ¶ 53). Dr. Stewart credibly concluded the complexity of D.V.’s case exceeds the skills of a psych associate and D.V. needed to be transferred to a higher level of care.

M.S. is a 32-year-old prisoner at Eyman who Dr. Stewart interviewed. He has repeatedly hurt himself while on suicide watch. M.S. has had numerous hospitalizations for self-harming behaviors, including "cutting open his abdominal cavity" and "attempting to cut his own throat." His current diagnosis is Schizoaffective Disorder, bipolar type. His most recent hospitalization was "on September 4, 2021 [two months before trial], after cutting a 7-inch long by one-inch-deep laceration to his right arm, swallowing three razor blades, and inserting three spork handles into his abdominal scar – all while on continuous watch at Kasson ." (Doc. 4109 ¶ 54) (emphasis in original). The psychiatry midlevel staff member determined M.S. "does not respond to medications and continues to self-harm regardless of therapy and medication interventions." (Doc. 4109 ¶ 54). Absent from the record is any indication he was evaluated for inpatient care. Dr. Stewart credibly concluded M.S. needed inpatient care.

Finally, Dr. Stewart encountered K.C., a prisoner at Perryville on mental health watch. K.C. reported hearing voices instructing her to kill herself, and her medical record reflected multiple past diagnoses. ICS was initiated five times between September 3, 2021 and September 10, 2021 for bizarre behavior, including defecating and urinating on the floor. She was eventually transferred to inpatient treatment just before trial on September 16, 2021 but follow-up with a psychiatric provider was not scheduled for 30 days. (Doc. 4109-1 at 46). Dr. Stewart notes her persistent psychotic and mood symptoms were wrongly managed.

It is clear from the evidence these prisoners should have been placed in inpatient units or, for K.C., placed much earlier. There is no basis to conclude the inpatient units are at only 36% capacity because there are not enough seriously mentally ill prisoners. What is more, though inpatient facilities are meant for the most challenging prisoners, those facilities also suffer from lack of sufficient staff, as illustrated by a prisoner named I.C. The following email to former ADCRR Interim Division Director of Health Services Richard Pratt and ADCRR Monitoring Bureau employee Vanessa Headstream on February 12, 2020, documents an appalling situation where ADCRR's own staff recognized urgent attention was needed for I.C. but the only trained mental health staff was "out of town."

This mental health patient at Phoenix has been self harming by banging his head for the past several days resulting in multiple ICS events and the use of OC spray. Mental health appears to be at a loss on how to deal with this inmate.

In an email sent today the Regional Director of Mental Health basically said to continue using OC spray as needed while the on site mental health team comes up with a treatment plan. We are told that Dr. Carr [the Regional Director of Psychiatry] has been consulted by phone but there is minimal documentation in the medical record to support any significant involvement by a psychiatrist. This inmate now has wounds on the back of his head and on his forehead from the head banging. There are staples holding the wound edges together on the back of his head but the forehead wound remains open as the two previous attempts to staple his frontal wound have failed because of the continuous head banging.

We just received a copy of an I/R [incident report] completed by security staff from last evening indicating that the mental health RN was encouraging the inmate to bang his head so that the restraint chair could be used. At the time of this nurse/patient encounter, the patient was NOT participating in head banging but began banging his head after the nurse told him to do so .... which resulted in a Use of Force event. This entire event was captured on video.

The FHA [Facility Health Administrator] and mental health apparently had a meeting this morning about this patient without having any input from Complex Operations. The FHA reported at the Warden Tracker meeting this afternoon that:

• the patient has allegedly lost 30 pounds since December

• Mental health staff and nursing staff are verbally reporting that the condition of this patient "is deteriorating" from his normal baseline standards

• When asked at the Tracker meeting this afternoon why this situation has not (apparently) been escalated to a psychiatric emergency with a Psychiatrist coming to Phoenix to complete a comprehensive examination and evaluation of this patient, the FHA responded that Dr. Carr would be coming on February 24 to assess the patient.

Apparently Dr. Carr is out of town. When the Warden asked the FHA if there is another Psychiatrist in the system who can come to Phoenix to assess the patient, she did not know.

The Warden and I share the concern that this particular issue is a true psychiatric emergency and that the response from the Centurion mental health leaders at the Regional level is inadequate. Warden Weiss is escalating this concern through his chain of command and I am doing likewise.

(Doc. 4109 ¶ 191). Dr. Stewart expressed understandable outrage at I.C.’s treatment at the Phoenix complex, ADCRR's inpatient treatment facility for the most acutely mentally ill prisoners. I.C. needed emergency care but a psychiatrist was "out of town" until twelve days later. Dr. Carr's response to this "true psychiatric emergency" was revelatory in its confirmation of systemic deficiencies.

This unfortunate situation is a symptom of a larger problem.

Numerous elements impede timely intervention, quality of care, implementation of a comprehensive treatment plan and psychiatric stability.

Our inpatient unit needs a larger investment from Psychiatry, Nursing, Mental Health, ADC and Medical.

As you know our inpatient unit is licensed by DHS. It is imperative we model our clinical program according to clinical guidelines and license rules/regulations.

[I.C.] is the immediate focus. However, barriers to care, collaboration, education, training and communication need to be addressed in order to implement a solid care plan.

(Doc. 4109 ¶ 195).

Dr. Penn did not address I.C.’s treatment. Rather he concluded "I found no evidence that ADCRR or Centurion engaged in a planned or purposeful effort to effect inappropriate uses of force on any of the above individuals identified as having mental illness." (Doc. 4172 ¶ 275). This means either Dr. Penn did not have knowledge of I.C.’s treatment, including the email, or if he was aware of it, his opinion reflects a profound lack of concern. And Defendants have not addressed the email or Dr. Carr's assessment of inadequate mental health care.²⁶

26 At times, Defendants attempted to tout telepsychiatry as an augment to the apparent inadequate number of psychiatrists on-site. But telepsychiatry was, for unknown reasons, not available for I.C.’s situation. If telepsychiatry could not help a situation as dire as this, there is no evidence that telepsychiatry is a viable alternative to on-site personnel.

The inadequate mental health staffing has led to prisoners who require inpatient levels of care not receiving such care. As Dr. Stewart testified, many prisoners in obvious need of competent intensive treatment are kept in other units that are ill-equipped to treat their conditions. The inadequate staffing of inpatient facilities creates a substantial risk of serious harm to prisoners.

B. Inadequate Treatment for Suicidal Prisoners

Inadequate mental health care staffing is demonstrated in the records of prisoners who died by suicide in recent years. Dr. Stewart provided a review of some of the twenty-three prisoners who died by suicide from 2019 until the close of discovery for trial in September 2021. What follows are some of them.

1. Reuben Neal

Neal died by suicide on August 27, 2020 at the age of 29. He had five visits with mental health staff in the months leading up to his death. On July 8, 2020, he had an eight-minute encounter with a midlevel provider where he was prescribed subtherapeutic levels of two medications to address his anxiety and insomnia. (Doc. 4109-1 at 103). He had another encounter with a midlevel provider on July 16, 2020. That encounter lasted ten minutes. Neal expressed his anxiety was causing a "racing heart" and he felt "backed up against a wall." He also reported feeling all alone. His medication was not adjusted, despite his obvious and urgent need. His follow-up was kept to every 90 days or as needed.

Neal had a visit with a psych associate on July 22, 2020. He reported having a lot of issues, but they were not addressed, and no follow-up plan was documented. His visit on August 17, 2020 lasted five minutes. While the clinician inquired about suicidal ideation, no therapy occurred. His final visit, on August 19, 2020, lasted ten minutes. The notes reflect it was a superficial visit where nothing of substance was discussed. (Doc. 4109-1 at 104). Neal died by suicide eight days later.

The psychological autopsy noted Neal had been transferred to a higher custody level unit eight days before his suicide. That transfer increased his anxiety level and negatively affected his sleep and concentration. Dr. Stewart concludes Neal received "exceedingly poor counseling care and psychiatric care in the weeks leading up to this death." (Doc. 4109-1 at 103). In particular, Neal had encounters with staff who were not qualified to accurately assess his needs. Neal needed "more frequent and thorough assessments" by higher level staff. (Doc. 4109-1 at 103).

2. Eric Haag

Haag died by suicide on September 13, 2020 at the age of 25. (Doc. 4109-1 at 105). The records from Haag's initial psychiatric evaluation when he entered prison in 2016 establish that evaluation was effectively useless given how carelessly it was conducted. The history indicated Haag admitted he had taken medications for paranoid schizophrenia, but he attributed the voices he heard not to schizophrenia but to "drugs." But the initial evaluation stated he had no history of drug abuse and no history of psychiatric hospitalization. Later records establish Haag began abusing drugs at age nine and had a history of psychiatric hospitalization.

The records immediately prior to Haag's death establish he was not on medication for schizophrenia. On July 11, 2020, Haag had a five-minute visit with a psych associate and reported being under a lot of stress at his prison job and requested to have his medication restarted. (Id. ). It appears that was ignored. On July 20, 2020, a midlevel prescribing provider reported that Haag was experiencing irritability as well as depressive and anxiety symptoms. There is no indication this provider knew of Haag's earlier request to restart medication. On August 12, 2020, Haag had his last mental health visit prior to his suicide, which consisted of a three-minute visit with a psych associate who noted that Haag allegedly was not willing to engage. Assuming that was correct, the psych associate stated Haag was not at risk for self-harm. There is no explanation how that assessment was possible due to Haag's alleged unwillingness to engage.

Haag had a complicated mental health history and due to the abbreviated and the brief nature of the mental health visits attributable to staffing issues, Haag was not provided adequate care. Those brief visits combined with the lack of coordination between the counselor and midlevel provider, and incomplete initial psychiatric assessment and inconsistent medical records, contributed to Haag's suicide.

3. Tracie Otero

Otero died by suicide on September 22, 2020 at the age of 47. (Doc. 4109-1 at 107). As of July 2020, Otero was in severe pain, which staff attributed to fibromyalgia. On July 31, 2020, Otero had an encounter with an unlicensed psych associate. That encounter confirmed Otero was suicidal based on the amount of pain she was in. On August 6, 2020, an unlicensed psych associate documented a flat affect, low appetite, and that Otero reported she was "tired and in pain." Otero ended the session due to pain. On August 21, 2020, an unlicensed psych associate documented Otero was refusing medications and struggling with anxiety. Otero again ended the session due to pain. On September 14, 2020, Otero had her last visit with any mental health staff prior to her suicide. That visit with an unlicensed psych associate lasted three minutes before Otero ended the session due to pain. The psych associate said Otero "does not appear to be a danger to self or others at this time." But no explanation was given on how the psych associate, after a three-minute visit that was terminated due to extreme pain, came to that conclusion. In Dr. Stewart's view, a more experienced clinician would have recognized the extent of Otero's pain put her at risk for self-harm.

Otero was extremely undermedicated for her pain leading up to her death, leaving her in unremitting pain. Otero's risk of suicide due to pain was missed because of the inexperience of the unlicensed and unsupervised clinicians as well as the abbreviated nature of the visits. Sufficiently qualified staff able to spend the requisite amount of time with Otero would have recognized the gravity of her situation.

4. Austin Georgatos

Georgatos died by suicide on January 28, 2021 at the age of 20. (Doc. 4162-1 at 110). Similar to Haag, Georgatos also experienced an initial incompetent assessment upon arriving at prison. On January 14, 2021, Georgatos arrived from jail and had an initial mental health assessment. That assessment lasted five minutes and determined Georgatos had been on medications for anxiety and depression in jail, but those medications, without explanation, had been discontinued two weeks prior. The assessment also documented Georgatos’ history of methamphetamine use and both sexual and physical abuse in childhood. He was wrongly assessed to have "no emergent [mental health] issues" and no subsequent mental health appointments were scheduled. (Doc. 4109-1 at 110).

On January 25, 2021, he submitted an HNR stating "I need to see psych doctor about the voices I am hearing in my head. They returned since I stopped taking medications." But Georgatos was never seen by health care staff. Instead, he committed suicide three days later. The mortality review determined the death was possibly avoidable, with a failure to recognize symptoms or signs and delay in access to care as contributing causes of the suicide. Dr. Stewart opined Georgatos suicide was preventable. The inadequate intake screening was far too brief, likely as a result of inadequate staffing. In addition, inadequate follow-up based on the reports at intake and the significant delay in care after reporting severe psychiatric symptoms contributed to Georgatos’ death.

5. Jason Rothlisberger

Rothlisberger died by suicide on April 15, 2021 at the age of 45. His final week of life demonstrated he was at grave risk of self-harm. (Doc. 4109-1 at 113). On April 7, 2021, Rothlisberger attended a "sick call" where he reported depression, anxiety, and not sleeping for days. He also reported thoughts of harming others as well as concerns about his own safety. That same day an ICS was initiated after Rothlisberger's sister called and reported concerns that he may end his life. When asked, Rothlisberger admitted he had placed a rope around his neck to get the attention of staff due to fear for his life and said his anxiety was "10/10" as well as suicidal and homicidal ideations.

Between April 8, 2021, and April 12, 2021, Rothlisberger was placed on 10-minute and 30-minute watches. He continued to report concerns regarding his own safety but also started reporting auditory hallucinations. At one point, Rothlisberger told the "mental health team" he was "genuinely terrified" of remaining in the suicide watch pod because "they were raping, torturing, and killing sex offenders at 3 or 4 in the morning." Inconsistently, the very next day, April 12, 2021, Rothlisberger allegedly "did not endorse significant psychiatric symptoms to the mental health worker." At some point between April 12 and April 15, Rothlisberger was discharged from crisis watch and there was no record this occurred or why.

The mortality review determined contributing factors to Rothlisberger's suicide included failure to recognize symptoms or signs and failure to follow clinical guidelines. In addition, there were no suicide risk assessments conducted upon placement or removal from crisis watch and no crisis treatment plan developed within one day of placement into crisis watch. In fact, the records established Rothlisberger's "mental status was significantly worse at the time of discontinuing [the] watch than when it was started." Rothlisberger did not receive an appropriate referral to a prescribing psychiatrist. In Dr. Stewart's opinion, because of "significant red flag symptoms" of psychosis, this suicide was avoidable.

6. Failures Discovered in Psychological Autopsies

Dr. Stewart expressed significant concerns regarding the care leading up to these suicides but his concerns were unrebutted by Defendants at trial. The Court asked Dr. Penn about whether he had examined the mortality reviews of prisoners who died by suicide and whether he had opinions about those prisoners’ treatment. (Doc. 4283 at 46:5-17). He said he had reviewed some of them. (Id. ). But Dr. Penn only pointed to his declaration where he found that ADCRR performs mortality reviews, which he concluded establishes an effective mental health care delivery system. He testified, "I noted, when rare suicides occur, psychological autopsies and administrative suicide reviews are conducted, morbidity and mortality reviews are completed ... and I think that's the important take home point, is that all of the ADCRR facilities were found to be 100 percent compliant with the essential standard of NCCHC suicide prevention and intervention." (Doc. 4283 at 47:7-10, 47:19-23). Again, Dr. Penn, without credible explanation, appears to be relying entirely on the existence of certain ADCRR policies and the NCCHC policies. The NCCHC standard determines only whether ADCRR performs mortality reviews or has some suicide prevention policies. It is undisputed ADCRR performs reviews and has prevention policies. But the existence of those reviews and policies does not establish the reviews or policies are effective, thorough, or competent. And the evidence establishes they are not.

When the Court asked Dr. Penn about his evaluations of the psychological autopsies he said, "I definitely remember reviewing all of the morbidity and mortality reviews, but as far as writing a specific analysis of each and every single adverse patient outcome, I don't recall writing something separate." (Doc. 4283 at 48:22-49). What is more, Dr. Penn failed to take a single note during his chart and record reviews. Rather, he kept all the information "in [his] head." (Doc. 4283 at 45:15-18). It was obvious from his testimony that Dr. Penn did not retain sufficient information or conclusions about particular prisoners to discuss them in helpful detail. Rather, Dr. Penn simply concluded all treatment was appropriate. In effect, the Court has only Dr. Stewart's reliable testimony where he concluded the care was inadequate.

C. Overburdened Staff Have Not Met the Demand Required

1. Prisoners Receive Either No Treatment Plans or Inadequate Ones and Experience Long Term Symptoms Due to Insufficient Treatment

The overwhelming evidence demonstrates the lack of mental health staffing—both in terms of overall staffing and insufficient higher-level staff—demonstrates inadequate care. This results in prisoners receiving ineffective treatment leaving prisoners to remain potentially, and unnecessarily, profoundly symptomatic for extended periods of time. (Doc. 4260 at 54). This is established through instances of prisoners who remain symptomatic despite the availability of other treatment options. Thus, prisoners often have treatment options worthy of exploration but, due to staffing, they cannot be explored and implemented. Dr. Stewart highlighted a few of these prisoners.

A.C. is presently housed at the inpatient facility in the Phoenix complex. Dr. Stewart interviewed A.C. and assessed him as "extremely manic and agitated ... yelling at the custody staff and running around his cell naked." (Doc. 4109 ¶ 115). A.C. had bruises on the left side of his torso where custody staff had fired a paint ball gun at him during a psychotic episode. He is prescribed Risperdal Consta, 37.5 mg every two weeks. Dr. Stewart noted it "takes up to ten weeks for this long-acting antipsychotic medication to reach therapeutic levels." Dr. Stewart explained it is not appropriate to "attempt to stabilize a highly agitated and psychotic patient using long-acting injectable antipsychotic medication." As a result, A.C. remains out of control and "is not receiving anything close to inpatient level of care." (Doc. 4109 ¶ 117).

D.W. is housed at Eyman. D.W. reported "persistent auditory hallucinations ... command [ing] him to hurt himself and others." (Doc. 4109 ¶ 118). His medical records reflected he has been prescribed the same 15 mg dose of Abilify for sixteen months despite his ongoing auditory hallucinations commanding him to hurt himself and others. D.W. has not been referred to a psychiatrist for medication adjustment or any other treatment. Dr. Stewart noted 15 mg of Abilify is a "low dose" and a higher dose likely would be effective.

J.R. is housed at Eyman. When Dr. Stewart visited J.R., he was "overtly psychotic with very loose associations, disorganized and rambling speech, responding to internal stimuli and thought blocking." (Doc. 4109 ¶ 114). J.R. was not currently prescribed or taking a single medication despite a diagnosis of schizophrenia. In Dr. Stewart's opinion, J.R. represents "the ultimate example of deliberate indifference. That is, the staff acknowledge that [J.R.] is psychotic but are not doing anything to address it." (Doc. 4109 ¶ 114).

The appalling evidence regarding Rahim Muhammad's treatment is the strongest example of how prisoners receive inadequate treatment and, as a result, remain profoundly symptomatic for extended periods of time. Muhammad has been diagnosed with schizophrenia and schizoaffective disorder and has been placed on suicide and self-harm watch "dozens" of times. (Doc. 4109 ¶ 181). Dr. Stewart noted Muhammad was repeatedly on suicide watch and had chemical agents used against him on July 5, 2021 (twice), July 7, 2021, July 8, 2021, July 9, 2021, July 12, 2021, and July 13, 2021 (twice). (Doc. 4109 ¶ 181). The Court observed some of these uses of force against Muhammad, and they were nothing short of shocking. It is clear Muhammad was profoundly mentally ill and Defendants only solution to his self-harming was to use chemical agents against him.

Moreover, Deputy Warden Van Winkle testified that even though Muhammad was housed in Florence-Kasson—a mental health unit—he approached mental health staff and told them Muhammad needed "more care than what we're able to provide here." (Doc. 4279 at 34). Muhammad remained at Florence-Kasson for seven months while exhibiting profoundly dangerous and symptomatic behavior. And because of the severity of his illness, Dr. Stewart planned to interview him just prior to trial during his September 23, 2021 tour, but found that Muhammad had been transferred to a different facility "earlier that very same morning." (Doc. 4109 ¶ 181).

Dr. Penn did not meaningfully address these prisoners’ care. (Doc. 4172 ¶ 270). Dr. Penn offered only that Dr. Stewart's view of these prisoners was wrong because "psychotherapy and psychotropic medications require treatment compliance, initiative, time, and effort from the patient." (Doc. 4172 ¶ 270). But Dr. Penn offered no analysis of how his opinion applies to the unique illnesses for each of these prisoners. In short, Dr. Penn's opinion has no merit.

2. Drive-By Mental Health Encounters

Another consequence of insufficient staff is that mental health encounters are extremely short in duration, remarkably so. This issue was repeatedly litigated during the life of the Stipulation, without resolution. Defendants insisted extremely short encounters were sufficient while Plaintiffs claimed encounters were so short that they could not possibly have been efficacious. Eventually, because Defendants proved unreasonable, the Court imposed minimum durations for visits, subject to exceptions if a mental health clinician concluded visits of short duration were acceptable. (Doc. 3861 at 12). Defendants refused to comply with that Court Order, which required another Order clarifying that durations shorter than the presumptive minimum were required to be assessed by a psychiatrist. (Doc. 3861 at 13). Defendants failed again to comply and ironically admitted they lacked sufficient qualified staff to conduct reviews of short-duration visits. (Doc. 3907 at 6; Doc. 3921 at 16-17).

At trial, Dr. Stewart testified that "[i]t is simply not possible to assess a patient and determine their risk of self-harm or suicide in an encounter lasting five, three, or two minutes. Such an assessment requires more than literally ‘seeing’ a patient; it first requires establishing a therapeutic relationship." (Doc. 4109 ¶ 60).²⁷ Dr. Penn insisted a mere one-minute encounter could be sufficient to determine if a prisoner was at risk of self-harm or suicide.

27 Dr. Stewart also explained that when mental health care providers attempt to speak with prisoners at their cell fronts, the lack of confidentiality inhibited communication. And many prisoners refused to be moved to the cage where confidential encounters were held because "they were very psychotic and paranoid and they were afraid that if they were to be taken out of their cells, that they would be at risk of being attacked by others." (Doc. 4260 at 45).

Question: Could it be one minute?

Dr. Penn: Yes, possibly.

Question: To determine that the person was no longer a risk of harm to themselves or others?

Dr. Penn: Yes, certainly, if the mental health staff member knows the inmate, has reviewed the chart, has spoken to staff, and is a qualified mental health professional, yes.

(Doc. 4283 at 100:12-18).

Dr. Penn went further, claiming a prohibition on extremely short visits would be counterproductive. In his view, "setting minimum required durations for mental health care encounters can hamper the clinician's ability to provide mental health care." (Doc. 4172 ¶ 139). Dr. Penn, however, spoke entirely in generalities and did not point to a single actual encounter and then explain why it nonetheless was enough. Moreover, the history of this issue involves repeated Court Orders expressing hesitation at imposing minimums and always including a simple uncomplicated procedure whereby a qualified individual could assess if a short encounter was appropriate. Finally, Dr. Penn's opinion that the extremely short encounters were sufficient is not credible for the simple reason that Dr. Pennington-Stallcup conceded such encounters were never sufficient in any circumstance.

Question: But a one-minute encounter would never be sufficient to determine that a patient is not at risk of self-harm; correct?

Dr. Pennington-Stallcup: Correct.

(Doc. 4276 at 125: 9-11).

The current prevalence of conducting extremely short duration mental health encounters is established by the numerous performance measures while the Stipulation was in effect. These performance measures required prisoners be "seen" at identified intervals. For example, Performance Measure 80 required a prisoner with a mental health score of MH-3A "be seen a minimum of every 30 days by a mental health clinician." As stated, "seen" was finally interpreted to require an encounter meeting minimum duration requirements, subject to the exception for short encounters later deemed clinically appropriate by a qualified individual. The table below reflects Defendants were consistently unable to meet the duration requirements for April through July 2021.

Performance Measures for Mental Health Durational Requirements Performance Complex April May June July Measure 2021 2021 2021 2021 80 Lewis 88% 86% 24% 43% 80 Tucson 99% 87% 17% 42% 85 Eyman 95% 97% 25% 73% 85 Florence 83% 100% 50% 100% 85 Lewis 97% 97% 24% 69% 85 Perryville 92% 83% 0% 68% 85 Tucson 98% 100% 6% 69% 85 Yuma 97% 100% 5% 85% 91 Phoenix 100% 100% 80% 100% 94 Eyman 94% 100% 82% 91% 94 Florence 93% 100% 80% 93% 94 Tucson 100% 96% 80% 87%

(Exs. 1729-1732).

3. Inconsistent Distribution of Medications

It is undisputed that psychotropic medications require a sophisticated medication distribution system. Many medications require timed distribution or must be taken with meals. Dr. Stewart testified the "failure to provide timely and consistent delivery and administration of psychotropic medication places patients at substantial risk of serious harm, including pain and suffering, withdrawal symptoms, or deterioration." (Doc. 4109 ¶ 136). Defendants’ own evidence consistently demonstrates their staffing failures result in prisoners not receiving timely and consistent delivery of psychotropic medications.

Performance Measure 13 required chronic care and psychotropic medication renewals be completed such that there was no interruption or lapse in medication. (Doc. 1185-1, Ex. B at 8). The performance numbers found below show that, for most of 2021, Perryville and Yuma in particular were unable to provide medication refills in a timely manner.

Performance Measure 13 Jan. Feb. March April May June July 2021 2021 2021 2021 2021 2021 2021 Douglas 97 100 100 100 93 96 96 Eyman 86 84 84 80 92 88 92 Florence 88 65 77 86 88 88 75 Lewis 69 95 93 99 86 91 77 Perryville 84 82 86 79 74 73 86 Phoenix 94 91 91 97 98 95 100 Safford 90 90 100 90 100 97 100 Tucson 90 94 94 82 67 70 79 Winslow 100 100 100 100 93 100 100 Yuma 64 74 78 74 68 76 96

(Ex. 1256).

Of significance, the failure to deliver medication when needed was well-known throughout the system, yet Defendants failed to prevent it. The Continuous Quality Improvement meeting minutes across multiple facilities document interruptions and delays in delivering medications to patients. And Dr. Stewart noted meeting minutes showed delays at Eyman, Perryville, and Tucson in March 2021. (Doc. 4109 ¶ 138). Similar issues also arose at the Phoenix inpatient complex in January and June 2021. (Ex. 905 at 13395 (Phoenix Complex January 2021: prisoners were erroneously given their medications at 4:00 p.m. and again at 8:00 p.m. "due to several breakdowns in communications ....")); (Ex. 831 at 61889) (Phoenix Complex June 2021: "Medication errors were noted and discussed with individual staff."). For many years, the lack of staffing has meant prisoners do not receive medication when they need them to achieve or maintain mental health.

IV. Defendants’ Challenges to Dr. Stewart's Methodology and Opinions Fail

Dr. Stewart's opinions focused on the most seriously mentally ill prisoners. Dr. Stewart explained he "wanted to focus [his] attention on the more seriously mentally ill ... [because] one of the best measures of [ ] evaluating a system's ability to care for mentally ill inmates is to look at the more serious mentally ill cases." (Doc. 4260 at 17). Dr. Penn claimed this approach was unscientific and argued a random approach was preferable. Dr. Penn purported to conduct a "rigorous and representative analysis" but in essence his analysis supports Plaintiffs’ position.²⁸

28 It is difficult to overstate Dr. Penn's lack of credibility. He was evasive when asked direct, simple questions. He failed to take a single note during his medical records review. (Doc. 4283 at 21:16-19). His ignorance of fundamental aspects of Defendants’ health care system was obvious and his testimony contradicted the undisputed evidence at trial.

Dr. Penn testified he reviewed, with the assistance of four practicing correctional psychiatrists, 120 randomly selected charts in addition to the 155 charts Dr. Stewart reviewed. Dr. Penn instructed each consultant to determine whether "access to care" was provided for each of the prisoners reviewed.²⁹ Dr. Penn did not provide an entirely coherent explanation of "access to care." At one point, Dr. Penn said "access to care" required consultants determine if the prisoner was given "timely, appropriate" care that "met the standard of care." (Doc. 4289 at 33:18-21). Later, Dr. Penn redefined it such that a prisoner who had missed medications may or may not be labeled as having "access to care." (Doc. 4283 at 29:20-22). This method for defining "access to care" is far from objective, and much less scientific. Dr. Penn's methodology is further flawed because his "access to care" standard does not carefully analyze if the care was medically acceptable.³⁰ In essence, it was ambiguous, inconsistent and of no value.³¹

29 Dr. Penn's consultants were not, however, provided copies of mortality reviews for prisoners who died by suicide. (Doc. 4283 at 30).

30 For example, one prisoner who was placed on suicide watch for two weeks was designated as having "access to care" despite having never been seen by a psychiatric clinician during that time. Another patient's care was reviewed and the consultant noted "concern patient is not on meds and has diagnosis of schizophrenia or bipolar." (Doc. 4283 at 65). Dr. Penn did not share his consultant's "concern." The Court asked Dr. Penn, "But I am wondering in your report is there anything to indicate how you came about deciding that despite what the concerns were from your reviewer the treatment that this patient received was appropriate?" Dr. Penn only responded that because the patient continued to be seen, despite receiving no treatment, his care was acceptable. The table of various patients found at Appendix 5 shows those classified as not having access to care and those marked as having access but the accompanying notes for those patients demonstrate otherwise.

31 During trial, Defendants attempted to address a patient, M.L., that Dr. Stewart highlighted in his declaration. (Doc. 4109 ¶ 35). M.L. told Dr. Stewart he experienced severe anxiety, depression, and auditory hallucinations, submitted multiple HNRs about his symptoms, but did not receive a response. (Doc. 4109 ¶ 35). Dr. Stewart reviewed his medical records and noted M.L. submitted an HNR on August 28, 2021, requesting to see someone about getting his medication changed, but did not see a midlevel provider until September 8, 2021. (Doc. 4109 ¶ 35). During trial, Defendants sought admission of Exhibit 5315(a) for the purpose of demonstrating that M.L. was seen by a behavioral health technician on August 30, 2021. (Doc. 4261 at 65-70). But as Plaintiffs noted, "There is absolutely no reason why a patient would tell a behavioral health tech, who has no training and no ability to help him, that he had put in an HNR the day earlier asking to see a psychiatrist about his medications." (Doc. 4261 at 69-70).

Altogether, Dr. Penn and his consultants reviewed a total of 275 prisoner records, consisting of 120 randomly selected prisoners and all of the 155 records reviewed by Dr. Stewart. Defendants argue that of the 120 randomly selected prisoners, only 5% did not have "access to care." And because Defendants adopted Dr. Penn's random selection, they argue the Court should extrapolate that 5% to the entire prisoner population. Thus, they contend 5% of the mentally ill prisoner population without access to care is adequate.

This conclusion is erroneous. Even if only 5% of mentally ill prisoners do not have access to care, that is unconstitutional. But Dr. Penn's methodology is deeply flawed. In addition to the prisoners who were deemed not to have access to care at all, six additional prisoners’ charts were reviewed and, despite being characterized as having "access to care," they died by suicide (Patients 12, 13, 52, 115, 120, and 228). Others went weeks or many months without seeing a provider (Patients 86, 95, and 272). And of significance, combining Dr. Penn's random selection with the charts reviewed by Dr. Stewart results in a total of 275 prisoners. Of those, 38—13.8%—did not have access to care as determined by Dr. Penn's consultants. For another 26 prisoners the consultants determined they had "access to care," but also found deficient treatment, sometimes with fatal consequences. Ultimately, 64 out of 275—23.3%—of the records reflect either an outright denial of access to care or serious deficiencies in mental health care administration.

A more fundamental flaw with Dr. Penn's analysis is that, in his opinion, the only outcome that should qualify as "bad" is "death."³² And with death as the relevant threshold, Dr. Penn identified only one case where he had concerns about the care, where a prisoner submitted a request for mental health treatment, was not seen, and then committed suicide. Even there, however, Dr. Penn ultimately concluded the prisoner "was afforded appropriate mental health care." (Doc. 4285 at 61). As for the other 275 records reviewed, Dr. Penn saw no cause for concern:

32 Dr. Penn testified "I reviewed all of these charts, I didn't find any of these patients to be -- to have death or morbidity or mortality resulting, so there wasn't a bad outcome with these patients." (Doc. 4283 at 61:11-14). Thus, Dr. Penn believes death is the appropriate metric for a "bad outcome." This establishes a callous indifference to the prisoners whose charts did reflect inadequate mental health treatment but did not die.

Question: Other than this particular patient [who committed suicide], did you find any issues with respect to any of the other 275 records that you reviewed?

Dr. Penn: No, I did not find any gross deviations in standard of care or any access to care or continuity of care that fell below the standard of care.

The Court: So when you -- I don't have the testimony in front of me, but you said there was only one patient where you found that the standard of care had not been up to what it should have been; is this the one patient?

Dr. Penn: Yes, Your Honor.

(Doc. 4285 at 63:23-64:7).

Dr. Penn's opinion that only a single prisoner received possibly deficient mental health care is appalling and overwhelmingly contradicted by the evidence at trial. Again, Dr. Penn's handpicked consultants reviewed 275 files. (Doc. 4283 at 75) (stating Dr. Penn chose the consultants). When testifying about the consultants’ findings, Dr. Penn stated:

I think the question was, was access to care provided, and [the consultants] opined no based on the document review. And you are right, it was 38 out of the 275 charts.

(Doc. 4238 at 73). Thus, 13.8% of the prisoners reviewed by Dr. Penn's consultants did not have access to care. Dr. Penn then attempted to claim that lack of access to care was of no concern because only one prisoner suffered the "bad" outcome of "death." It is unclear whether Dr. Penn truly believed that not a single prisoner other than the one he discussed received deficient care or whether Dr. Penn believed he was hired to convey that unsupported opinion. But, either way, that obstinance—and indifference—was devastating to his credibility and renders his opinions unworthy of any weight.

V. Mental Health Care Conclusion

Excessively inadequate staffing pervades Defendants’ lack of adequate mental health care treatment. Every single complex is understaffed, except for two complexes which have only a single individual to address prisoners’ needs. Again, the consequences of this understaffing are dire, as reflected in the mortality reviews and psychological autopsies of prisoners who died by suicide. In Dr. Stewart's opinion,

the chronic shortage of mental health staff, delays in providing or the outright failure to provide mental health treatment, the inadequacies in the provision of psychiatric medications, and the other deficiencies identified in this report are systemic problems, and incarcerated people who need mental health care have already experienced, and will experience, a serious risk of injury to their health if these problems are not addressed.

(Doc. 4260 at 18-19). And Defendants own evidence confirmed Dr. Stewart's opinion.

SYSTEMIC PRACTICES THAT IMPEDE HEALTH CARE

There are two overarching aspects of ADCRR's provision of medical and mental health care that merit separate discussion. First, ADCRR's health record system is difficult to the point of being impossible to use. As such, it is a significant impediment to constitutionally adequate care. Second, ADCRR's failure to provide reliable language interpretation services is another significant impediment to care.

I. eOMIS Prevents Adequate Care

It is undisputed that Defendants’ electronic health record system ("eOMIS") is very inefficient. Defendants’ expert, Dr. Murray, testified eOMIS "does not have the work flow efficiencies, logistical safeguards, and data reporting capabilities necessary to produce ADCRR's best care. Without exception, the facility management teams at every complex indicated that a new EHR or modernized eOMIS would drastically increase their ability to see more patients." (Doc. 4206 ¶ 1080).

Dr. Wilcox testified in detail how eOMIS’ limitations prevents staff from seeing data in any usable format. eOMIS does not allow staff to view testing results or encounter notes over time. Rather, providers must manually review each lab result and provider note. Dr. Wilcox further testified lab results are reported alphabetically, but not by category. Similar inefficiencies exist for medication administration and x-rays. As a result, staff are not able to efficiently review a patient's history, creating a significant barrier to appropriate care.

This was demonstrated during trial. During Dr. Wilcox's cross-examination, counsel for Defendants struggled to maneuver through a patient's medical record to find the information they were seeking. To expect this from physicians who are expected to see 18-20 patients per day is unreasonable.

Dr. Murray elaborated about the problems when treating complex patients. "[T]he more complex the patient, the more fragmented the documentation in eOMIS and the more challenging to follow their care plan. At times, it was difficult to discern what happened in that patient's care, and what was supposed to have happened." (Doc. 4206 ¶ 990). He continued, "[t]his fragmentation issue is due to using an EHR that does not support the flow of patient care, does not push desired behavior and care, does not contribute to ideal decision-making, and does not allow a rapid review of a patient's history and health care needs. In short, this EHR does not allow for the care of complex patients in a comprehensive way and has become an impediment to the timely and efficient care of these patients." (Doc. 4206 at ¶ 990).

Thus, it is undisputed the very foundation of the health care delivery system—the medical records—"does not allow" for "comprehensive" care. In terms of the relevant issue for this case, eOMIS substantially contributes to a substantial risk of serious harm.

II. Defendants Do Not Provide Adequate Language Interpretation Services

Plaintiffs allege Defendants have a policy or practice of failing to provide adequate language interpretation services to prisoners receiving medical or mental health care. Plaintiffs’ experts stated "[c]ommunication with patients is essential to providing adequate medical care." (Doc. 4138 ¶ 442). And "[e]ffective communication is a fundamental component of providing therapeutic [mental health] care" and mental health care encounters without effective interpreter services are "largely meaningless and superficial." (Doc. 4109 ¶ 89). Dr. Wilcox explained that, ideally, both medical and mental health care staff "who have been evaluated and determined to be qualified to conduct healthcare encounters in the patient's language would conduct the healthcare encounter in that language." (Doc. 4138 ¶ 443). Another option would be the use of "qualified interpreters with experience with medical terminology." (Doc. 4138 ¶ 443). It is not appropriate, however, to use other prisoners as interpreters because the prisoner patient "may not be comfortable disclosing sensitive, potentially embarrassing medical information" to the prisoner acting as a translator. (Doc. 4138 ¶ 449). It is also inappropriate to use custody staff as interpreters because prisoners may "self-censor or alter communications with the provider" to prevent custody staff from learning sensitive information. (Doc. 4109 ¶ 91).

Defendants do not dispute that language interpretation services, in general, are a crucial aspect of minimally sufficient medical and mental health care. Dr. Pennington-Stallcup, agreed that "effective mental health diagnosis and treatment" requires "the clinician and the patient ... be able to communicate." (Doc. 4277 at 21:1-5). But Defendants claim the "various policies and practices" already in place are sufficient. (Doc. 4315 at 50). But the evidence establishes those policies and practices are far from sufficient. Prisoners are routinely denied language interpretation services, necessarily resulting in the deprivation of appropriate medical and mental health care. The absence of interpretation services creates a substantial risk of serious harm.

The main policies Defendants cite as establishing adequate interpretation services are found in the MSTM. The MSTM requires, during the initial intake process, every prisoner be provided a handout, in both English and Spanish, that "outlines how they are to access health care." (Ex. 1305 at 122). Similar handouts must be provided whenever a prisoner is transferred to a new complex. In addition, the MSTM requires the posting of signs in the intake area of each complex explaining how to access health care. (Ex. 1305 at 122, 145). Per this policy, the information need only be in Spanish and English, meaning the information would be useless to prisoners who speak other languages. (Ex. 1305 at 123).³³

33 There was no evidence presented whether the handout and signage policies are followed in practice. The Court will assume they are.

In addition to those requirements, during the intake process each prisoner meets "with nursing providers, dental, [and] mental health." (Doc. 4231-1 at 97:2-3). During those meetings, the prisoner's primary language is identified and entered into the electronic medical record. (Doc. 4231-1 at 97:3-7). The medical record system is programmed so that, in the future, staff must always answer a question regarding the prisoner's language abilities before inputting any new information. The language question is a "hard stop," meaning the staff inputting information "cannot proceed any further until they answer the question of whether a patient needs interpreter services or not." (Doc. 4231-1 at 97:5-12). The initial evaluation of language ability completed at intake is not governed by any written policies or procedures. (Ex. 1976 at 2-3 No. 3 ). Rather, determining the need for interpreter services apparently is left to the discretion of the intake personnel.

When a prisoner who is not fluent in English is seen by medical or mental health care staff, interpretation services are supposed to be available through a "telephone or any platform that has audio including a computer screen if [there is a] need [for] sign language interpretation and a visual communication cue." (Doc. 4231-1 at 100:6-10). This translation service is referred to as the "LanguageLine." Centurion "prefer[s]" that its staff use the LanguageLine even when the staff speaks the same language as the prisoner. However, that is not required, and staff may proceed without using the LanguageLine for any language the staff are "comfortable" using. It is very common for staff to do so. (Doc. 4231 at 101:21-102:3; Doc. 4277 at 68). Centurion does not have policies regarding assessing the language proficiency of its staff. (Doc. 4231-1 at 102:24-103:6). Centurion does not document or know if a staff member can, in fact, converse sufficiently in the language the staff member purports to speak and understand. (Doc. 4231-1 at 103:5-7).

In practice, the current policies and practices are ineffective and allow insufficient communication between prisoners and health care staff. First, the absence of any formal assessment process regarding prisoners’ interpretation needs creates a situation where staff have no way of knowing "whether and how much a patient actually understands" when being told information in English. (Doc. 4138 ¶ 454). This may lead to staff concluding a prisoner understands when, in fact, he does not. An unknown language barrier may also lead to "misdiagnosis [and] failure to follow treatment plans." (Doc. 4138 ¶ 454). If, during intake, a prisoner is not assessed as needing interpretation services, no notation is made in the electronic medical record and future encounters likely will occur in English, whether or not the prisoner can understand.

The evidence establishes the lack of any assessment leads to prisoners being incorrectly identified as not needing an interpreter. Dr. Wilcox spoke with two monolingual Spanish speakers at Tucson who were labeled in their health records as not needing interpretation services. Based on his discussion with them, it was "clear" these two prisoners could not have any "meaningful healthcare encounter in English." (Doc. 4138 ¶ 448). In fact, other portions of those prisoners’ medical records made clear they needed interpretation services. (Doc. 4138 ¶ 448). According to these prisoners, they have not been provided interpreters for medical services and one of the prisoners stated a fellow prisoner had to act as an interpreter. (Doc. 4138 ¶ 449).

Even when the need for interpreter services is known, Defendants and Centurion may not provide one. A medical record for one prisoner showed staff had to use "google translation for interpretation of questions and responses." (Ex. 928 at 1). In addition, Dr. Stewart, Plaintiffs’ expert on mental health care discussed in detail later, reviewed medical records of numerous hearing-impaired prisoners and concluded "they had gone months – if not years – unable to communicate meaningfully with health care staff." (Doc. 4109 ¶ 100).

The language difficulties for deaf prisoners are especially pronounced. Dr. Stewart referenced a "written exchange" between a psych associate and a deaf prisoner when that prisoner was placed on suicide watch after the death of his brother. Interpretation services should have been provided. Instead, the written exchange consisted of simple questions by the psych associate and responses by the prisoner. (The handwritten document makes clear who wrote which portion. For ease of reading, the prisoner's responses are bolded).

Homicidal? Kill Others? No

Hallucination No

Eating ok ok

Sleeping ok ok

Feeling anxious no

Depressed no

(Doc. 4109 ¶ 101). In Dr. Stewart's opinion, this "is an awkward, stilted, and slow way to communicate, and does not provide an appropriate or adequate medium to engage the patient in discussion of sensitive and important mental health matters." (Doc. 4109 ¶ 102). That is an understatement. It is hard to believe Defendants think it was appropriate to determine the mental status of a prisoner on suicide watch by passing him a note asking if he felt "Homicidal? Kill Others?" and then accepting his monosyllabic answer.

Dr. Stewart also spoke with several monolingual Spanish speakers who had been classified as seriously mentally ill. Those prisoners recounted significant difficulties in communicating with mental health providers, including complaints about being unable to convey difficulties with their medications. (Doc. 4109 ¶¶ 105-106). One prisoner did not respond to Dr. Stewart's questioning in English but, when Dr. Stewart switched to Spanish, the prisoner "lifted his head off his bunk and look[ed]" at Dr. Stewart. (Doc. 4109 ¶ 109). That prisoner's medical records indicated he does not need interpretation services but there is a plausible basis to conclude that is inaccurate. (Doc. 4109 ¶ 111).

Reliance on LanguageLine may, in theory, suffice. However, the actual implementation of LanguageLine establishes it does not solve the interpretation problems. LanguageLine is not available in every location and, at times, there are no interpreters available even when staff attempt to use LanguageLine. For example, when Dr. Wilcox visited the IPC at Tucson, he attempted to speak with two monolingual Spanish speakers. Dr. Wilcox asked to use the LanguageLine, but it could not be used where the patients were located. (Doc. 4138 ¶ 462). The only location where it was available was in an office, away from the prisoners. That office did not have an exam table and "would not have been appropriate for a healthcare encounter." (Doc. 4138 ¶ 462). Medical records also establish instances of the LanguageLine being ineffective because of "poor reception" or no interpreter being available when requested. (Doc. 4138 ¶ 466). While Defendants seem to concede the need for interpretation services, they presented opinions from Dr. Penn, Defendants’ expert on mental health care discussed in detail later. Dr. Penn suggested the status quo is sufficient. But Dr. Penn's testimony was so flawed on this topic that the Court cannot credit any of his opinions. At the outset, Dr. Penn began by misidentifying the issue. According to Dr. Penn, the "requisite standard of care within a correctional setting" does not mandate "use of certified translators for all healthcare interactions and/or for group psychotherapy." (Doc. 4172 ¶ 171). Dr. Penn did not explain, however, why he was evaluating whether "certified translators" were necessary. Neither of Plaintiffs’ experts opined "certified translators" were necessary or even possible. Rather, the issue is whether the current practice of having no interpreter services when needed is appropriate. Dr. Penn's initial opinion regarding the standard of care, even if accepted, is meaningless.

Next, Dr. Penn testified the correct "standard of care in the community and in correctional health care is for staff to assess their own level of comfort and proficiency before determining whether a separate translator/interpreter or translator service is required." (Doc. 4172 ¶ 174). This opinion makes no sense. Whether a staff member can communicate effectively in another language depends on an assessment by the prisoner, not the staff member. That is, a staff member may believe he has native speaker fluency while true native speakers cannot understand a word he says. The standard of care cannot plausibly be based on the staff's own subjective evaluation of their language skills.³⁴

34 Dr. Penn also testified it would be "unreasonable" to require interpreters because "in emergency situations, time is simply not available to have translators/interpreters and other services available." (Doc. 4172 ¶ 176). Plaintiffs’ experts did not state the standard of care requires interpreters in "emergency situations." No one is faulting Defendants for not having interpreters available, at all hours, for every emergency. Dr. Penn's testimony that adequate interpreters need not be provided whenever possible because some emergencies might occur when they are not available defies logic.

Dr. Penn further attempts to explain away the absence of effective interpretation policies and practices by opining the standard of care is that "use of an interpreter is dependent upon the nature and extent of the encounter." (Doc. 4172 ¶ 175). He gives the example that a blood pressure measurement may not require an interpreter but "[a]n interpreter would be necessary ... where there are discussions regarding advance directives." (Doc. 4172 ¶ 175). At trial, Dr. Penn was asked to identify other types of encounters when an interpreter "would be necessary." Dr. Penn refused to do so:

Question : And so what about individual health care counseling?

Dr. Penn : I would say it would depend.

Question : Mental health groups?

Dr. Penn : Again, it would depend.

Question : Suicide watch checks?

Dr. Penn : Again, it would depend.

Question : Chronic care encounters?

Dr. Penn : Same answer, it would depend.

Question : Appointments with health care or mental health providers?

Dr. Penn : It would depend.

(Doc. 4283 at 108:19-109:3). Dr. Penn explained he could not provide definitive answers whether an interpreter would be needed for these encounters because "[s]ome individuals are more fluent." (Doc. 4283 at 110:5-6). But Dr. Penn's testimony is nonsensical considering he testified an interpreter "would be necessary" for discussing advance directives without then stating that did not apply if the prisoner was fluent in English. The reason his testimony did not include an exception for prisoners fluent in English is because the relevant prisoner is one who is not fluent in English. Dr. Penn's invocation of the various degrees of fluency to avoid determining when interpreters should be provided was a way for him to avoid conceding the obvious: all but the simplest of health care encounters require an interpreter.

Finally, Dr. Penn testified that Dr. Stewart reported "he interviewed monolingual Spanish speakers" but Dr. Penn's review of their records "evidences [those prisoners’] ability to communicate in English." (Doc. 4172 ¶ 184). To reach this conclusion Dr. Penn explained:

I base this -- and I am not questioning Dr. Stewart or his Spanish proficiency. My first language was Spanish also. But I reviewed the Health Needs Request from inmates, and many of them were able to write in English.

(Doc. 4283 at 114:14-15). When pressed on this point at trial, Dr. Penn's testimony became absurd.

Dr. Penn was presented with a HNR written in Spanish and asked if that was written by the prisoner. Dr. Penn stated he is "not a handwriting expert" and, therefore, he could not tell if the prisoner identified in the HNR actually wrote that HNR. (Doc. 4283 at 115:17-19). Thus, Dr. Penn acknowledged his "opinion" based on reviewing HNRs may not reflect the named prisoners’ fluency in English because the HNRs may have been drafted by someone else. In addition, one of the prisoners identified by Dr. Stewart as a monolingual Spanish speaker submitted an HNR in Spanish that, when translated, stated in relevant part "If you do not understand Spanish, please tell me and I will find a way to send this in English." (Doc. 4283 at 118:1-3). After reviewing that HNR, Dr. Penn stated

So the patient is saying, I can find a way to get it translated to English. So as we sit here today, without reviewing the chart and without talking to the patient, I wouldn't be able to answer if the patient is fluent in English or not.

(Doc. 4283 at 118).

When pressed on how he had testified earlier that this patient could communicate in English, Dr. Penn stated

So again, I would need to review the patient's record in its entirety and perhaps interview the patient to be able to determine that.

(Doc. 4283 at 119). In other words, Dr. Penn contradicted his earlier testimony that this prisoner could communicate in English. Instead, Dr. Penn stated he would need to review the prisoner's record and interview him before determining whether the prisoner can communicate in English. And in any event, Dr. Penn's testimony that this prisoner was able to write in English cannot withstand the slightest scrutiny. The prisoner submitted the HNR in Spanish and included a statement that he would "find a way" to send it in English, if necessary. It is not plausible the prisoner would try to "find a way" to submit his HNR in English if his own language abilities meant he could have submitted in English.

In general, Dr. Penn's testimony on the topic of language interpretation was unreliable and incredible. Dr. Penn's opinion—the status quo of interpretation services is sufficient—is rejected. The only credible evidence before the Court demonstrates Defendants’ language interpretation policies and practices often lead to "largely meaningless and superficial" health care encounters. And inability to communicate one's medical and mental condition creates a substantial risk of serious harm.

LEGAL STANDARD FOR HEALTH CARE

Eighth Amendment claims involving medical or mental health care requires a prisoner prove an "objective prong" as well as a state-of-mind element, or "subjective prong." See Disability Rts. Montana, Inc. v. Batista , 930 F.3d 1090, 1098-1099 (9th Cir. 2019). For systemic challenges of policies and practices, "[t]he first, objective, prong requires that the plaintiff show that the conditions of the prison pose a substantial risk of serious harm." Id. If the prison conditions present such a risk, "[t]he second, subjective, prong requires that the plaintiff[s] show that a prison official was deliberately indifferent by being aware of the facts from which the inference could be drawn that a substantial risk of serious harm exists, and also draw[ing] the inference." Id.

The "objective prong" involving a "substantial risk of serious harm" requires prisoners establish they might suffer "further significant injury" or experience the "unnecessary and wanton infliction of pain." Id. The "subjective prong" requires prisoners establish prison officials "knowingly and unreasonably disregarded" the risk of harm. Farmer , 511 U.S. at 846, 114 S.Ct. 1970. Put differently, prison officials must know "about the risk to which prisoners were exposed" but then "deliberately choose to maintain the harmful policies" or practices. Batista , 930 F.3d at 1099. The subjective prong can be met when "prison officials deny, delay or intentionally interfere with medical treatment, or it may be shown by the way in which prison physicians provide medical care." Colwell v. Bannister , 763 F.3d 1060, 1066 (9th Cir. 2014) (quoting Hutchinson v. United States , 838 F.2d 390, 394 (9th Cir. 1988) ).

CONDITIONS IMPOSED ON THE CLASS

In its Order certifying the class, the Court identified the following six medical care practices that allegedly exposed Plaintiffs to a substantial risk of serious harm in violation of the Eighth Amendment:

(i) Insufficient health care staffing;

(ii) Failure to provide timely access to health care;

(iii) Failure to provide timely emergency treatment;

(iv) Failure to provide necessary medication and medical devices;

(v) Failure to provide for chronic diseases and protection from infectious disease; and

(vi) Failure to provide timely access to medically necessary specialty care.

The mental health care practices that allegedly subjected Plaintiffs to a substantial risk of serious harm in violation of the Eighth Amendment were:

(i) Failure to provide mentally ill prisoners medically necessary mental health treatment (i.e., psychotropic medication, therapy, and inpatient treatment) and;

(ii) Failure to provide suicidal and self-harming prisoners basic mental health care.

The challenged practices significantly overlap. For example, "insufficient health care staffing" relates to and affects "failure to provide timely access to health care." That is, insufficient health care staffing results in the failure to provide timely access to care. As set forth above, inadequate staffing has resulted in a cascade of failures including other discrete practices, including inadequate access to necessary medication and inadequate access to specialty care. Similarly, the mental health care practices of failing "to provide mentally ill prisoners medically necessary mental health treatment" overlaps with failing "to provide suicidal and self-harming prisoners basic mental health care." Accordingly, some of the practices will be considered together to demonstrate Plaintiffs are being subjected to a substantial risk of harm.

I. ADCRR does not have sufficient health care staffing to provide prisoners with timely access to health care, including emergency treatment, medication, treatment for chronic disease, and specialty care. ³⁵

35 This does not address some of the originally identified practices nor does it address particular health care conditions. For example, Plaintiffs did not establish Defendants failed to protect them from "infectious diseases" or that Defendants failed to provide access to "medical devices." And while Plaintiffs presented evidence regarding specific health care conditions like Hepatitis C or Medically Assisted Therapy, this case is not about particular diseases. Rather the focus must be on the overall provision of health and mental health care and whether there is a substantial risk of serious harm.

Long ago, the Ninth Circuit established the basic requirements for a constitutionally sufficient correctional health care system. It must "provide a system of ready access to adequate medical care." Hoptowit v. Ray , 682 F.2d 1237, 1253 (9th Cir. 1982) (citation omitted), overruled in part on other grounds by Sandin v. Conner , 515 U.S. 472, 115 S.Ct. 2293, 132 L.Ed.2d 418 (1995). Prisoners must be able "to make their medical problems known to the medical staff." Id. And that staff must be competent because "[a]ccess to the medical staff has no meaning if the medical staff is not competent to deal with the prisoners’ problems." Id. This requires staff be competent enough "to examine prisoners and diagnose illnesses" and then "treat medical problems or to refer prisoners to others who can." Id.

The evidence demonstrates Defendants operate a system lacking sufficient numbers of qualified medical staff to provide prisoners with "ready access to adequate medical care." Id. As discussed above, access to care necessarily requires access to competent health care. The standard of care in the community is higher than the constitutional minimum. But the "[a]ccepted standards of care and practice within the medical community are highly relevant in determining what care is medically acceptable and unacceptable." Edmo v. Corizon, Inc. , 935 F.3d 757, 786 (9th Cir. 2019). The overwhelming evidence establishes the health care available to prisoners is well below the constitutional minimum, let alone the community standards. The current levels of staffing, including the type of staff, are far from adequate to meet constitutional requirements. Defendants’ insufficient hiring and retention of qualified staff create a substantial risk of serious harm to all class members, which affects every aspect of health care delivery because there is simply not enough time to provide the level of care that is necessary. The broken electronic health record, the lack of time to employ differential diagnoses to properly evaluate prisoners’ conditions, and nurses’ failure to refer prisoners to providers are all predictable effects of a lack of staff.

II. ADCRR does not provide mentally ill prisoners with medically necessary mental health treatment and suicidal and self-harming prisoners with basic mental health care.

ADCRR and Centurion employees also confirmed insufficient staffing extends to the provision of mental health care. The lack of staff—specifically, again, licensed staff with the training to treat prisoners with serious mental health needs—impacts every aspect of the provision of mental health care treatment. The evidence reflects the prisoners with the highest acuity and most serious needs are not provided the medically appropriate level of treatment. Inpatient units remain unfilled, and Centurion's own officials confirmed the reason was a lack of staffing. Defendants’ mental health care expert's chart reviews confirmed a lack of access to medically acceptable mental health care. And, again, psychological autopsies reflect systemic deficiencies, but no follow up action was taken to prevent such harm from reoccurring.

III. Defendants Are Deliberately Indifferent

The testimony from Dr. Wilcox and Dr. Stewart was credible while the testimony from Dr. Murray and Dr. Penn was not. Therefore, as detailed above, the medical and mental care available at ADCRR complexes presents a substantial risk of serious harm to every prisoner. Therefore, the "objective prong" of the Eighth Amendment analysis is met. The only remaining question is whether Defendants have acted, and continue to act, with deliberate indifference to that risk.

The overwhelming evidence demonstrates Defendants Shinn and Gann are both aware of insufficient staffing and the other conditions that present a substantial risk of serious harm to prisoners. Defendants, however, have not taken effective actions to remedy those conditions. Despite years of knowledge, driven by this litigation and Defendants’ monitoring of their private healthcare contractors’ performance, Defendants have in fact made no significant attempts to substantively change the health care system and compel sufficient staffing.³⁶ Thus, Defendants are acting with deliberate indifference to Plaintiff's serious medical and mental health care needs. The testimony from Defendants Shinn and Gann provides compelling evidence of knowledge of the failures but a refusal to take meaningful measures to correct systemic flaws.

36 Defendant Shinn, as ADCRR Director, is legally responsible for providing a constitutionally adequate health care system. Ariz. Rev. Stat. § 31-201.01(D) ; West v. Atkins , 487 U.S. 42, 56, 108 S.Ct. 2250, 101 L.Ed.2d 40 (1988) ("Contracting out prison medical care does not relieve the State of its constitutional duty to provide adequate medical treatment to those in its custody, and it does not deprive the State's prisoners of the means to vindicate their Eighth Amendment rights."). As the Court noted in Doc. 3940, when Defendant Shinn believes his chosen vendor must provide health care in a different manner, he must enforce his rights under the vendor contract.

During his testimony Defendant Shinn was asked about a letter he sent to the Chief Executive Officer of Centurion on February 14, 2020. (Ex. 2165). That letter was prompted by a Court Order that had threatened Defendants with contempt if they did not perform their obligations under the Stipulation. Defendant Shinn's letter informed Centurion of that Order and he identified the particular performance measures that were the subject of possible contempt fines. Defendant Shinn explained ADCRR was facing fines of up to $14.6 million per month if Centurion did not improve. The letter then stated "ADCRR expects and demands Centurion to immediately commit all necessary resources to achieve and maintain compliance on every Performance Measure at every specific location addressed in [the Court Order]." The letter further demanded Centurion "take all reasonable steps to substantially comply" with the Court Order and it listed ten specific steps Centurion had to take. Those steps included the relocation of existing "Arizona healthcare personnel to locations that face challenges in achieving compliance," temporary transfer of Centurion personnel from other states, scheduling additional meetings, filling all vacant positions, and drafting "remedial plans for root causes that hinder compliance." Defendant Shinn was asked about his actions after sending this letter.

When asked at trial if he knew "how many health care personnel Centurion reallocated from one facility to another" following his letter, Defendant Shinn responded "No, we were focused on outcomes." (Doc. 4274 at 90). And he failed to elaborate on which specific outcomes were his focus and were achieved. He was then asked if he knew "from which Arizona facilities Centurion reallocated health care staff" or "to which facilities Centurion reallocated health care staff within Arizona." Defendant Shinn knew neither. In fact, Defendant Shinn did not know if any health care staff had been reallocated within Arizona. Defendant Shinn was asked if he knew of any health care staff transferred to Arizona from other states. Defendant Shinn did not know. When asked if he knew where Centurion had increased telemedicine availability, Defendant Shinn did not know. On this point, Defendant Shinn volunteered his "expectation was [Centurion] get to 100 percent staffing across the board. I wasn't concerned with individual pockets." (Doc. 4274 at 93). When asked if he knew what actions Centurion had taken to recruit additional staff, Defendant Shinn deflected:

What I am looking at is outcomes and are we moving the needle forward. That's exactly what I'm looking at, health care delivery. I'm not necessarily looking at, did you move one from this location to that location? I'm simply asking, provide the services that were asked for and contracted and the Court has demanded as well.

(Doc. 4274 at 94). When pressed, Defendant Shinn admitted he did not know of any efforts to recruit more staff, nor did he know how many vacancies Centurion filled as a result of his letter.

Based on Defendant Shinn's testimony, he somehow determined he had no obligation to follow-up with the demands he made in his letter. The only possible conclusion to draw is that Defendant Shinn had little interest in changing the underlying reality. Rather, his letter appears to have been nothing more than a half-hearted effort to generate a piece of paper he could cite to avoid contempt. Obviously, the Court expected Defendant Shinn to take more direct action than signing a letter. Defendant Shinn's disinterest in Centurion's response to his letter shows he is knowingly ignoring the reality of the fundamental problems with ADCRR's health care system.

Defendant Shinn also evidenced little curiosity about the conditions at issue in this case. Defendant Shinn agreed that if there were problems in ADCRR, he would want to know of them. (Doc. 4274 at 100). But Defendant Shinn admitted he had not read any of the expert declarations in this case. (Doc. 4274 at 100). There was no explanation why Defendant Shinn, if he was interested in potential problems, chose to make no effort to familiarize himself with Plaintiffs’ factual contentions.

Another crucial aspect of Defendant Shinn's testimony that proved his knowledge of ongoing systemic failures is found in the negotiation of an extension to Centurion's contract. In June 2021, ADCRR negotiated an extension with Centurion for another year. As part of that contract, ADCRR "agreed to cap the indemnification of Centurion at $2 million for court ordered sanction and fees related to [this case]." (Doc. 4274 at 97). That meant Centurion would be liable for no more than $2 million in Court-ordered fines and any amounts beyond that would be borne by ADCRR.

As explained by Defendant Gann, that indemnification cap was something Centurion "wanted in the contract." (Doc. 4275 at 144). Accordingly, it can only be assumed that it was clear to Defendant Shinn at the time of the contract renewal that Centurion had significant concerns regarding its performance. More specifically, Centurion apparently realized it would not be able to perform adequately and significant contempt fines were likely. To avoid catastrophic liability, Centurion ensured ADCRR would bear the brunt of nonperformance. Instead of looking into why Centurion was so concerned with additional sanctions, and taking action to address the potential for continued nonperformance, Defendant Shinn simply agreed to limit Centurion's liability and insulated it from meaningful consequences for its failures.

Defendant Shinn's testimony also made clear he has adopted a strategy of pretending the problems he knows about do not exist. According to Defendant Shinn, Centurion has "done some great work" and in the 24 months leading up to trial, Centurion "produced some extraordinary results." (Doc. 4274 at 101-02). When asked if "Centurion has done extraordinary work," Defendant Shinn responded "Absolutely." (Doc. 4274 at 102). And despite his February 2020 letter, and his ignorance regarding any responsive actions to that letter, Defendant Shinn was "overall" satisfied with Centurion's performance under the contract. (Doc. 4274 at 102). In other words, "overall," Defendant Shinn is satisfied with a system that presents a substantial risk of serious harm. That is almost a perfect illustration of "deliberate indifference."

The Court pressed Defendant Shinn on his strange nonchalance regarding his 2020 letter to Centurion and Centurion's response. The Court asked him if he thought the demands in the letter were "necessary in order to comply with the court order at that time." Defendant Shinn stated he believed the demands were necessary. (Doc. 4274 at 111). But the Court then asked him about his testimony that he was only concerned with outcomes, not performance of his demands. Defendant Shinn agreed he was primarily concerned with outcomes, but he stated Centurion had experienced some limited success:

[Centurion] experienced success in a number of -- in a number of performance measures at many locations throughout our system. Have they failed at certain locations? There's no question about that. That's not debatable. It's just math, and we are looking at that on a monthly basis. That's what I'm looking for, is for them to ultimately be 100 percent across the board with all measures. But we're looking in reductions of failure as well as moving the needle from a systemic perspective.

(Doc. 4274 at 112). That did not answer whether he continued to believe Centurion must perform the demands in his letter. Thus, the Court asked Defendant Shinn if he was still demanding Centurion perform. He responded "we are still asking for them to complete [the letter's demands]." When asked to clarify, Defendant Shinn admitted he "was less concerned about how they do this and more concerned about the results that they produce, in terms of delivery of health care." (Doc. 4274 at 113). In sum, Defendant Shinn believed the demands in his letter were important, except when that required him to follow up or take additional action. He further believed the demands were still in effect, except he actually only cared about "results." Defendant Shinn's actions are not the type of actions a concerned administrator would have taken if he was actually interested in ensuring the undisputed failures were being resolved.

Defendant Shinn's state of mind is further illustrated by testimony that can only be described as shocking.

Question : And you believe that the health care that's currently provided to the ADCRR prisoner population meets and often exceeds the community standard of care?

Defendant Shinn : In some aspects, yes.

Question : And you believe that has been the case continuously since your arrival at the Department two years ago?

Defendant Shinn : Yes.

Question : Okay. And you believe that the access to care for Arizona prisoners exceeds access to care for people in the community?

Defendant Shinn : In terms of timeliness and immediate access in person, I know that from a personal perspective with my provider. It is not even close as a private citizen.

Question : You believe that in an overall sense, access to care for Arizona prisoners exceeds access to care for people in the community, correct?

Defendant Shinn : I know it exceeds my personal experience in Arizona, yes, sir.

Question: You believe that in an overall sense, access to care for Arizona prisoners exceeds access to care for people in the community, correct?

Defendant Shinn: I know it exceeds my personal experience in Arizona, yes, sir.

(Doc. 4274 at 102-03).

The claim that prisoners’ access to care "exceeds" the access to care enjoyed by people in the community is completely detached from reality. Given the overwhelming evidence and repeated instances of insufficient care leading to suffering and death, Defendant Shinn could not possibly believe prisoners have the same access to care as people in the community.³⁷ Defendant Shinn is acting with deliberate indifference to the substantial risk of serious harm posed by ADCRR's health care system.

37 There is no evidence of what medical conditions, if any, Defendant Shinn had at the time he testified. Charitably, Defendant Shinn's opinion would have modest validity if his assessment of ADCRR medical care was based on Dr. Murray's simplistic and incredible theory that the alleged periodic testing at ADCRR of blood pressure and sugar levels is adequate medical care. But Defendant Shinn's opinion would have overlooked the overwhelming evidence of prisoners suffering from disabling psychosis, fatal cancers, fatal bleeding ulcers, multiple sclerosis, fatal Hepatitis C, and severe drug addiction. Because Defendant Shinn was aware of this evidence when he gave his opinion, he could not have plausibly believed Arizona prisoners’ access to care exceeds his own experience. And because he is responsible for ADCRR's grossly inadequate health care system, his answer was a blatant admission of his flagrant dereliction of responsibilities as the Director of the Arizona prison system.

Defendant Gann's testimony also contains overwhelming evidence he knows of ongoing systemic failures but refuses to take adequate remedial action. Defendant Gann testified he had "prepared numerous staffing plans in [his] career." (Doc. 4275 at 45). He stated those plans are "extremely complicated" and require detailed analysis of each facility, including physical layout, as well as determining the scope of services to be provided and the attributes of the prisoner population. (Doc. 4275 at 48-60). He further stated it is "impossible" to use a staffing analysis prepared for a particular facility and then "overlay that" analysis to another facility. (Doc. 4275 at 53). In brief, Defendant Gann testified that determining the health care needs of a particular prisoner population requires a tremendous effort, including application of specialized knowledge.

Despite knowing a valid staffing analysis is critical for the delivery of health care, Defendant Gann has not prepared any staffing analysis for the Arizona system as a whole or for any individual complex. (Doc. 4275 at 124). Instead, Defendant Gann and ADCRR have simply relied on Centurion to determine the appropriate staffing levels. (Doc. 4275 at 92). Defendant Gann admitted he did not know how Centurion "arrived at the current staffing levels that they needed." (Doc. 4275 at 92). Thus, there is no way Defendant Gann could have a meaningful understanding of how the staffing numbers were originally devised or why they were deemed sufficient. Despite his own knowledge that staffing analyses are crucial, Defendant Gann has been uninterested in assessing the adequacy of the current staffing levels.

Defendant Gann admitted ADCRR does not have any policies or procedures "regarding the number of needed healthcare staff at each prison." (Doc. 4275 at 92). Defendant Gann further admitted he knew Centurion was not paying its employees what they could command elsewhere. (Doc. 4275 at 97). But, at the same time, Defendant Gann professed he had "no idea" what Centurion was paying its employees. When asked if he had demanded Centurion increase employee pay, Defendant Gann responded "That's a no. That's their business." (Doc. 4275 at 97).

Defendant Gann testified insufficient staffing was not a "barrier" to complying with the Stipulation. (Doc. 4275 at 98). But he then agreed insufficient staffing often leads to cutting corners and inadequate documentation. He also agreed the complexes at Eyman, Florence, Lewis, Tucson, and Yuma all had "compliance problems" with the Stipulation. (Doc. 4275 at 101). Tucson in particular had canceled the nurse line many days, had a backlog of HNRs, and was not delivering insulin to diabetics in a timely fashion. Defendant Gann admitted these issues "seem to be related to staffing." (Doc. 4275 at 109). Defendant Gann also identified other failures, such as "pre-pouring medications."³⁸ According to Defendant Gann, pre-pouring medications could be due to insufficient staff, "laziness or poor culture." (Doc. 4275 at 103-04). Given his acknowledgment that inadequate staffing can lead to cutting corners, the lack of adequate staffing is the most likely explanation for the pre-pouring problem.

38 This refers to a practice of staff removing medications "from the blister pack" and putting them into envelopes for delivery to prisoners in the housing units. (Doc. 4275 at 103). Defendant Gann agreed this was not an acceptable method of medication delivery.

When asked about specific staffing levels reported by Centurion, Defendant Gann offered some revealing testimony. The Court asked Defendant Gann whether it was within Centurion's discretion to hire more individuals in a certain position than what the contract specified. (Doc. 4275 at 115). Defendant Gann responded with a general overview of his approach to staffing:

I think that [Centurion] really got into this contract in their best and final offer and they came back six months later and requested more staff. But you can't come back six months later and tell me – I'm just speaking --I wasn't here. But you can't tell me six months later you need more staff, and much of it's clerical related, not clinical ... and expect the

taxpayers to just go back and alter the contract by over 100 positions.

(Doc. 4275 at 115). In other words, Defendant Gann views the number of contracted positions as an absolute upper limit. If Centurion determines it needs more staff to provide adequate care, Defendant Gann is not interested. Rather than looking to the adequacy of care being provided, Defendant Gann is interested in ensuring the contracted-for staffing is not exceeded.

In addition to the testimony by Defendants Shinn and Gann, the undisputed evidence from the time the Stipulation was in effect establishes officials had no interest in taking sufficient actions to remedy known problems. During the Stipulation's life and beyond (discovery for trial went through September 2021), prison officials repeatedly claimed they could improve performance on their own. As explained by Dr. Wilcox, ADCRR identifies Corrective Action Plans allegedly as a mechanism to remedy deficiencies in health care. But Dr. Wilcox identified CAPs that repeated for months or years on end without improvement at Eyman (PM 13, PM 40, PM 44, PM 47, PM 52), Florence (PM 44, PM 50), Lewis (PM 44), Perryville (PM 13, PM 44, PM 51), Tucson (PM 37, PM 44, PM 47, PM 50, PM 51), Winslow (PM 44), and Yuma (PM 13, PM 37, PM 40, PM 51). In other words, ADCRR identified problems and identified solutions, but those solutions failed. Instead of trying different solutions, ADCRR simply repeated the already-failed "solutions" for months or years. This is evidence Defendants knew of failures but refused to take meaningful steps to those failures, often for years.

The foregoing evidence that Defendants Shinn and Gann knew of significant failures in the health care system builds on a unique aspect of this case involving the Stipulation. Defendants’ failures during the Stipulation are undisputed. After all, the failures were based on Defendants’ own records. Given that, Defendants cannot credibly claim they were unaware of the problems. Thus, Defendants only credible defense is to rely on legal, rather than factual, arguments. Those arguments fail.

IV. Defendants’ Legal Arguments are Unavailing

One of Defendants’ main arguments against a finding of liability is their belief that the "individual claims" brought by each of the named Plaintiffs fails. (Doc. 4309 at 98). Defendants do not provide a coherent explanation for this approach. In fact, Defendants even argue the named Plaintiffs’ "allegations were not raised in the Complaint." (Doc. 4309 at 98). Defendants seem to believe this litigation consists of determining whether the medical care provided to the named Plaintiffs was constitutional. But this case raises the question of whether Defendants’ policies and practices create a substantial risk of harm to all prisoners. As the Ninth Circuit stated when affirming class certification:

The cases cited in the defendants’ briefs, many of which involve individuals challenging particular instances of medical treatment or conditions of confinement, confirm that they (erroneously) view the plaintiffs’ claims as ultimately little more than a conglomeration of many such individual claims, rather than as a claim that central policies expose all inmates to a risk of harm.

Parsons v. Ryan , 754 F. 3d 657, 675 n.17 (9th Cir. 2014). The Supreme Court has also stated cases raising "systemwide deficiencies" do not turn on the care provided to a single prisoner on any particular occasion.

In Brown v. Plata , the Supreme Court explained:

Because plaintiffs do not base their case on deficiencies in care provided on any

one occasion, this Court has no occasion to consider whether these instances of delay-or any other particular deficiency in medical care complained of by the plaintiffs-would violate the Constitution under Estelle v. Gamble , if considered in isolation. Plaintiffs rely on systemwide deficiencies in the provision of medical and mental health care that, taken as a whole, subject sick and mentally ill prisoners in California to "substantial risk of serious harm" and cause the delivery of care in the prisons to fall below the evolving standards of decency that mark the progress of a maturing society.

Brown v. Plata , 563 U.S. 493, 505 n.3, 131 S.Ct. 1910, 179 L.Ed.2d 969 (2011). Defendants’ insistence that discrete care provided to particular prisoners forecloses their liability is misplaced. This case is not about the care provided to, for example, Kendall Johnson. While Defendants maintain, absurdly, that the care provided to her did not violate the Eighth Amendment, this case does not turn on whether Ms. Johnson was mistreated. Rather, the question is whether the policies and practices create a risk of harm to Ms. Johnson and all other prisoners. There is no question they do.

Defendants also attempt to rely on the Ninth Circuit's recent decision in Balla v. Idaho , 29 F.4th 1019 (9th Cir. 2022). There, the district court had terminated an injunction that was ongoing since 1984 requiring the Idaho State Correctional Institution to make changes in the areas of food preparation, medical facilities, safety, and many other aspects of prison operations. In the years following the injunction's issuance, there was lengthy litigation over Idaho officials’ compliance with the district court's orders and an increased focus on medical care and overcrowding. In 2020, Idaho officials moved to terminate the injunction, maintaining there was insufficient evidence they remained deliberately indifferent to the prisoners’ serious medical needs. The district court "found no current and ongoing constitutional violations" and terminated the injunction. The plaintiff appealed, arguing violations were continuing and the injunction should not have been terminated. The Ninth Circuit affirmed termination of the injunction.

Balla is distinguishable from the present case. Of significance, the plaintiffs’ expert in Balla was asked by the district judge whether the existing medical care resulted in injury or the unnecessary and wanton infliction of pain. The expert answered, "I don't know if I can answer that." Balla v. Idaho State Board of Correction , 2020 WL 2812564, at * 16 (D. Idaho May 30, 2020). Here, the evidence of constitutional violations is substantial. Both Dr. Wilcox and Dr. Stewart made clear the current system for the provision of medical and mental health care exposes all prisoners to a substantial risk of serious harm. Their testimony was corroborated by chart reviews by Dr. Murray and Dr. Penn demonstrating systemic deficiencies. Balla does not help Defendants.

Defendants offer a case from the Seventh Circuit, Rasho v. Jeffreys , 22 F.4th 703 (7th Cir. 2022), where an injunction regarding mental health care in the Illinois Department of Corrections was vacated. The Seventh Circuit found the prison officials did not have the required mental state to show they had acted with "deliberate indifference." Id. at 710. The Seventh Circuit concluded the prison officials "made reasonable efforts to cure the deficiencies identified" by the plaintiffs." Id. While those "efforts fell short" of actually correcting the constitutional violation, the Seventh Circuit concluded "reasonable measures," even if "ultimately unsuccessful," meant the officials had not acted with deliberate indifference. Id. at 711. It is important to note the extent of the "reasonable measures" undertaken in Rasho . After the suit was filed, prison officials "spent $45 million to build new residential treatment units at several facilities and $75 million to develop a new data system for intake assessments; [they] procured another $150 million to construct a new inpatient facility; [they] delivered mental-health training to [their] entire staff; and [they] hired administrative personnel to coordinate inmate care." Id. at 707. There is nothing remotely similar in the present suit. Defendants did not present evidence of new units being built or of requests to the legislature for additional funding to ensure better medical and mental health care. Defendants cannot be viewed as pursuing "reasonable measures" to correct the known problems.

An additional reason to doubt the applicability of Rasho is that it adopts a view of prison officials’ obligations that cannot be squared with existing Ninth Circuit law. In Peralta v. Dillard , 744 F.3d 1076, 1083 (9th Cir. 2014), the en banc majority held a prison dentist could not be held responsible for money damages in his individual capacity for staffing conditions over which he had no control. But in prohibiting the award of money damages, Peralta made clear "[l]ack of resources is not a defense to a claim for prospective relief because prison officials may be compelled to expand the pool of existing resources in order to remedy continuing Eighth Amendment violations." Id. Accordingly, contrary to Rasho ’s analysis that ineffective remedies may suffice, Peralta holds that in a suit for injunctive relief prison officials must remedy ongoing constitutional violations of which they are aware. Peralta is binding and it ensures adherence to constitutional standards. In the Ninth Circuit, state officials are not free to engage in ongoing violations of prisoner's constitutional rights merely because of policy decisions within the state officials’ control, such as funding.

To be clear, the Court does not take issue with the uncontroversial principle that prison officials can avoid liability by responding reasonably to a risk, even if the harm is not ultimately averted. But that principle is primarily aimed at situations where the claim is being viewed retrospectively and there is no ongoing constitutional violation to remedy. In injunctive relief suits, it often will not be enough for prison officials to attempt to remedy constitutional violations. They must actually do so.

Defendants also rely on Fraihat v. ICE , 16 F.4th 613 (9th Cir. 2021). In Fraihat , five immigration detainees with serious underlying medical conditions filed suit claiming ICE was acting with deliberate indifference to their medical needs and reckless disregard of known health risks. The plaintiffs claimed they were at increased risk if they contracted Covid-19 and they sought a nationwide preliminary injunction requiring ICE take actions to protect them from infection. Id. at 618. The district court certified two nationwide classes and issued a preliminary injunction that applied to all immigration detention facilities nationwide. Id. The injunction required ICE identify and track detainees with certain risk factors, required ICE issue a comprehensive performance standard regarding Covid-19 issues, and promulgate directives for release of detainees. The court later amended the preliminary injunction to impose more detailed requirements and procedures designed to result in the release of "substantial numbers of detainees from ICE custody." Id. The government appealed.

The Ninth Circuit set aside the injunction and held the plaintiffs had failed to demonstrate a likelihood of success on the merits of their claims. Id. It noted the plaintiffs had not sought individualized injunctive relief or relief specific to the conditions where they were housed but had instead challenged ICE's nationwide Covid-19 directives and essentially asked the district court "to assume control of top-level policies governing ICE's efforts" to combat the pandemic. Id. The plaintiffs had failed to present evidence of constitutional and statutory violations on a "programmatic, nationwide level." Id. The Ninth Circuit concluded "the slew of national guidance, directives, and mandatory requirements that [ICE] issued and then frequently updated in the spring of 2020" belied the notion that ICE had acted with "reckless disregard" necessary to support a finding of unconstitutional, system-wide deliberate indifference.

Defendants’ reliance on Fraihat is misplaced. The ICE facilities in that case consisted of "a nationwide network of over 250 detention facilities." Id. at 620. Those facilities differed dramatically from each other:

These facilities differ in various ways. ICE owns some of the detention facilities; others are operated under contract with state or local agencies or government contractors. Some of the centers are "dedicated" facilities, which hold only ICE detainees, whereas others are "non-dedicated" facilities, which also hold non-ICE detainees. ... Facilities also vary based on who provides medical care. Government employees, as part of the ICE Health Services Corps (IHSC), provide direct medical care at twenty facilities, which together hold about 13,500 detainees. The remaining facilities employ medical staff that the federal government does not directly employ.

Id. at 620. This type of analysis does not apply to the present case involving facilities owned and operated by ADCRR.

Prisoners incarcerated in ADCRR's ten complexes are subject to the same failed practices, which expose all of them to a substantial risk of serious harm. Understaffing is pervasive at every complex and the manner in which prisoners are assessed is common systemwide. Moreover, prisoners are routinely transferred between ADCRR's complexes. A healthy prisoner at one location might be transferred to another location when he is diagnosed with an illness or when he needs specialty care. Unlike Fraihat , ADCRR's complexes cannot be viewed as meaningfully distinct from each other.

Defendants’ legal arguments are no more persuasive than their factual arguments. Defendants know prisoners face a substantial risk of serious harm due to the policies and practices yet Defendants have refused for ten years to take the actions necessary to alleviate that risk.

CONCLUSION ON HEALTH CARE

The current policies and practices, of which Defendants are aware, require prisoners to repeatedly submit requests for treatment, with the response they will be seen on the nurse line. And while nurses are a vital component of any health care system, they are not substitutes for physicians, particularly for prisoners who develop serious conditions or sustain serious injuries (which could be any of them, at any time, in any location). Yet, despite these risks, Defendants provide no substantial explanation detailing why they have persisted in pursuing the same policies and practices regarding staffing. In particular, Defendants have not explained why they refuse to renegotiate the staffing matrix and, inexplicably, Defendants did not even explain how the staffing numbers were derived in the first instance.

Dr. Murray emphasized there is no evidence of indifference from individual health care staff, and Defendant Shinn noted their "tremendous will to succeed." (Doc. 4274 at 21:22-22:5). But this case is not about each individual health care provider's will to succeed. They are part of a system that is doomed to fail because, from the beginning, it was not designed to succeed. This case is about whether Defendants have a willingness to succeed. And Defendants remain deliberately indifferent not because they know about individual deficiencies in treatment but because they know of the widespread failures to provide care and still fail to take significant action. This has been evident since the inception of this action. Defendants’ failure to comply with the Stipulation, failure to undertake meaningful remedial action when their own mortality reviews reflect preventable deaths, and failure to take any meaningful follow up actions to ensure Centurion's compliance is what transforms their knowledge into conscious disregard of the risk of harm prisoners face.³⁹

39 Notably, the evidence demonstrates that the individual cases before the Court were the product of deliberate indifference to serious medical needs and the root cause of the deliberate indifference were the policies that prevented any substantive treatment of the serious medical needs of the individuals.

The fundamental conclusion is that ADCRR prisoners who develop life-threatening medical and mental health conditions are at a significant risk of serious harm. The ones that do develop such conditions may die prematurely, suffer prolonged pain or symptoms, or survive with lifelong disabilities. This risk is applicable to all prisoners because anyone is susceptible to serious injury or illness at any time. Moreover, the medical and mental health care systems are treated as interconnected such that prisoners housed at one location often are transferred to other locations for treatment. No prisoner, at any location, is safe.

Defendants’ refusal to acknowledge any shortcomings is foundational to these findings. If Defendants had admitted their obvious errors but attempted to demonstrate that, despite those errors, they were still capable of providing constitutionally adequate health care, that would have required a different analysis. If Defendants had acknowledged deficient care and attempted to demonstrate those were violations of existing policies, that would have required a different analysis. Instead, Defendants stubbornly maintain that there is nothing amiss (other than the need for a new EHR). In a system that produces this many catastrophic outcomes, where its own health care expert's chart reviews reflect care that meets the definition of deliberate indifference, when its mental health care expert's chart reviews reflected at least 23.8% of charts reviewed showed seriously deficient or an outright denial of care, when not one witness could validate the existing staffing allocation, and when nearly everyone agrees that ADCRR is critically understaffed, it is simply not possible to draw any other conclusion than the health care delivery system poses a substantial risk of serious harm to all prisoners. These issues are pervasive at all 10 ADCRR complexes. The overarching policies of critical understaffing, nurse-led health care, lack of adequate interpretation services, and a disastrous EHR are system-wide.

MAXIMUM CUSTODY

I. Subclass Definition and Admissibility of Expert Testimony

The Court certified a subclass defined as:

All prisoners who are now, or will in the future be, subjected by the ADC to isolation, defined as confinement in a cell

for 22 hours or more each day or confinement in the following housing units: Eyman–SMU 1; Eyman–Browning Unit; Florence–Central Unit; Florence–Kasson Unit; or Perryville–Lumley Special Management Area.

Two aspects of this class definition were subsequently addressed by the Ninth Circuit. First, the definition's reference to particular locations did not mean all prisoners housed at those locations were automatically members of the subclass. As explained by the Ninth Circuit, "[t]he touchstone for inclusion in the subclass is ... the amount of isolation experienced by inmates." Parsons v. Ryan , 912 F.3d 486, 503 (9th Cir. 2018). Thus, the crucial aspect is "the amount of time an inmate is ‘confine[d] in a cell’ each day," not where the prisoner is housed. Id. Those prisoners "who are confined in a cell for 22 hours or more each day (i.e., inmates who receive less than 14 hours of out-of-cell time each week)" are members of the subclass. Id. This clarification by the Ninth Circuit was important because some of the listed locations, such as Florence-Central and Eyman-SMU I, may house prisoners who receive offers to spend substantially more than 14 hours per week out of their cells at varying levels of restrictive housing. Id. at 502.

The second aspect of the subclass definition addressed by the Ninth Circuit was a prisoner's choice whether to leave his cell did not dictate membership in the subclass. In the Ninth Circuit's view, "[c]onfinement ... connotes a lack of control whether to leave a particular place." Id. Thus, if an inmate is "given an opportunity to participate in out-of-cell activities," he "cannot be considered ‘confined’ in a cell during that time even if the inmate may theoretically decide not to take advantage of the opportunity." Id. For example, if a prisoner were "offered 15 hours of out-of-cell time per week for education, but turn[ed] it down," he would be excluded from the subclass. Id. This clarification was useful to focus the subclass issues on whether prisoners were being "offered" out-of-cell time. It is "offers," not whether prisoners accept those "offers," that matter for determining out-of-cell time.

When the Ninth Circuit addressed "offers" for out-of-cell time, it was not asked to address, and did not address, what this Court has previously described as "unreasonable and unattainable offers" for out-of-cell time. (Doc. 3359 at 10). Offers for out-of-cell time conveyed in unreasonable ways or accompanied by unreasonable conditions or consequences could render such offers unavailable such that they cannot be used when determining membership in the subclass; nor can they be used when calculating how much out-of-cell time was available to prisoners. (Doc. 3359 at 9). For example, if a prisoner is offered fifteen hours of out-of-cell time in a given week, but all of the offers are accompanied by unreasonable consequences, the prisoner must be deemed as not being offered any out-of-cell time. Therefore, while recognizing it is "offered" out-of-cell time that matters, "offers" must be reasonable or else the offers are deemed not to have occurred.

Understanding the current conditions of confinement experienced by the subclass requires examining Defendants’ numerous policies regarding prisoner classification and placement. Unfortunately, those policies are complex and difficult to understand. Moreover, those policies routinely bear little resemblance to actual practices. Before considering those policies, however, the Court must address the admissibility and reliability of the testimony offered by two of Plaintiffs’ experts.⁴⁰ Those experts, Martin Horn and Dr. Craig Haney, testified extensively regarding the policies, practices, and other conditions of confinement relevant to the subclass. Thus, it is essential to ensure the admissibility of the experts’ testimony before making factual findings regarding the subclass.

40 Defendants did not offer any controverting expert testimony.

A. Martin Horn

Martin Horn earned a bachelor's degree in government from Franklin & Marshall College in 1969 and a master's degree in Criminal Justice from John Jay College in 1974. In 1969, Horn began his career as a New York State Parole Officer. He worked in that position for six years. From 1975 to 1977 he worked as an assistant professor of criminal justice at the State University College in Utica, New York. From 1978 to 1985 he held various positions in the New York State Department of Correctional Services, including serving as Superintendent of the Hudson Correctional Facility. He then became the Director of Parole Operations for the New York State Division of Parole where he served from 1985 to 1991. He was that entity's Executive Director from 1991 to 1995. He then became Pennsylvania's Secretary of Corrections until 2000.

As of January 1, 2002, Horn became Commissioner of the New York City Department of Probation and, one year later, he became the Commissioner of the New York City Department of Corrections. In that role he was responsible for "literally everything that happened in the agency." (Doc. 4266 at 135). He was responsible for "the health, safety and welfare of the inmates [and] for the security of the city jails." (Doc. 4266 at 135). He held the two New York City positions until July 31, 2009. Over the course of his career, Horn has published articles, delivered lectures, and testified before state and federal legislative bodies as well as state and federal courts. (Doc. 4166 at 4-6).

In preparation for his testimony, Horn visited ADCRR facilities at Eyman and Lewis, and interviewed more than 60 prisoners. (Doc. 4166 at 4). He took photographs to document the conditions. Plaintiffs offered Horn as an expert regarding the conditions of confinement and the operation of prisons. In particular, Horn was offered as an expert in the appropriate classification and housing of prisoners, including the impact of isolation on prisoners.

In their post-trial submissions, Defendants summarily argue "Horn does not have the requisite background, knowledge, or experience to provide opinions regarding the effects of isolation or other conditions of confinement on ADCRR inmates." (Doc. 4309). Without meaningful elaboration Defendants state only that Horn's "knowledge, skill, experience, training, or education" are insufficient to provide the opinions he offers.⁴¹ Pursuant to Federal Rule of Evidence 702, Horn is qualified because he has sufficient knowledge, training, and experience such that his opinions are admissible. The Court may or may not credit those opinions, but that is irrelevant to the admissibility issue. See Primiano , 598 F.3d 558, 564-65 (9th Cir. 2010) ("Under Daubert , the district judge is a gatekeeper, not a fact finder.").

41 Defendants also request the Court make strange findings regarding Horn's testimony. For example, Defendants request the Court find "Horn admits that per ADCRR's DO 812 governing maximum custody management, ADCRR maximum custody inmates are offered 7.5 hours a week of outdoor recreation." (Doc. 4309 at 340). There is no dispute what DO 812 clearly states and it is simply not true that Horn "admitted" prisoners in maximum custody were always being offered 7.5 hours each week of outdoor recreation.

B. Dr. Craig Haney

Craig Haney, Ph.D., is a Professor of Psychology at the University of California, Santa Cruz. (Doc. 4122 at 3). He has a bachelor's degree in psychology from the University of Pennsylvania, an M.A. and Ph.D. in psychology and a J.D. degree from Stanford University. Dr. Haney describes his area of specialization as "psychology and law" which he defines as "the application of psychological data and principles to legal issues." (Doc. 4122 at 3). He has published numerous articles and book chapters in law and psychology, with some emphasis on "the nature and consequences of solitary or ‘supermax’-type confinement." (Doc. 4122 at 3). Dr. Haney has served as a consultant to police departments, the California legislature, and the United States Department of Justice.

A special focus of his career has been "the psychological effects of living and working in real (as opposed to simulated) institutional environments, including juvenile facilities, mainline adult prison and jail settings, and specialized correctional housing units (such as solitary and ‘supermax’-type confinement)." (Doc. 4122 at 5). In performing his professional work, he has toured state prison facilities in twenty-nine states, multiple federal facilities, and international facilities in eight other countries. He has been qualified as an expert and testified in many state and federal courts. His work has been cited approvingly by the United States Supreme Court. Brown v. Plata , 563 U.S. 493, 518, 131 S.Ct. 1910, 179 L.Ed.2d 969 (2011).

Dr. Haney has been involved in this case since 2012. Over the years, he has toured Arizona facilities multiple times and spoken with many prisoners. He often selected prisoners at random but, over the years, he has made efforts to interview the same prisoners previously interviewed if they were still housed in facilities he visits. (Doc. 4122 at 9). He also reviewed medical and mental health records of the prisoners he interviewed. Finally, during his tours he took photographs. During this case, Dr. Haney filed numerous declarations containing his opinions regarding Arizona's policies and practices. (Doc. 4122 at 7).

In September 2021, Dr. Haney toured the following housing units: Eyman-SMU I, Eyman-Browning, Lewis-Rast Max, Lewis-Barchey and Morey (medium security units), Lewis-Stiner detention, and Lewis-Sunrise. During these tours, Dr. Haney interviewed approximately 75 prisoners. Dr. Haney attempted to re-interview prisoners he had previously interviewed but his selection of other prisoners to interview was "quasi random." (Doc. 4306 at 58). Those interviews did not follow a rigid structure. Rather, Dr. Haney asked each prisoner "what your life is like and how you feel you're being affected by the conditions and the procedures and your treatment." (Doc. 4306 at 121). Plaintiffs offered Dr. Haney as an expert regarding the subclass's conditions of confinement and, in particular, whether those conditions present a substantial risk of serious harm to the subclass members’ mental health.

As for Dr. Haney, Defendants again offered the summary argument that he "does not have the requisite background, knowledge, or experience to provide opinions regarding the psychological effects of isolation," without any meaningful detail. (Doc. 4309 at 327). Given Dr. Haney's knowledge, training, and experience, this argument is preposterous. Pursuant to Federal Rule of Civil Procedure 702, Dr. Haney has precisely the qualifications to testify to the exact subject matter in which he offers his opinions.

Further, Defendants did not bring timely motions to exclude Horn or Dr. Haney and, therefore, they have waived such challenges. See Skydive Arizona, Inc. v. Quattrocchi , 673 F.3d 1105, 1113 (9th Cir. 2012). The Court finds both Horn and Haney were credible. As appropriate, the Court will rely on Horn and Haney's testimony to describe the policies and conditions of confinement imposed on the subclass.

II. Policies Applied to Subclass

Defendants choose where to house prisoners, and the conditions at those locations, using five policies known as "Department Orders" ("DOs"). The five DOs relevant to the subclass issues are DO 801, DO 804, DO 807, DO 812, and DO 813.

A. DO 801

DO 801 sets forth Defendants’ "Classification Policy." (Ex. 1309). As explained by Horn, "classification is a term that is used very widely in prisons and jails all over the country" and it is "a fundamental building block of operating prisons and jails that are safe and secure." (Doc. 4266 at 159). The underlying premise of classification is to identify prisoners who can be placed in "the least restrictive situations" versus those prisoners who must be housed in more restrictive environments. (Doc. 4266 at 159-160). DO 801 requires each prisoner be assigned two scores, known as a "Custody Level" and an "Internal Risk Level." In general, the "Custody Level" "refers to that sort of gross discrimination between minimum custody and maximum custody and gradations in between." (Doc. 4266 at 163). The "Internal Risk Level" refers to "when an inmate is at a prison, how much independence of movement the inmate can have within that particular [prison]." (Doc. 4266 at 163).

1. Custody Classification

DO 801 outlines "four custody levels": Maximum Custody; Close Custody; Medium Custody; and Minimum Custody. (Ex. 1309 at 5). Determining a prisoner's custody level requires an individualized analysis using a document titled "Objective Classification: Custody & Internal Risk Technical Manual." (Ex. 1310). The criteria used in this process are:

(i) Most serious current offense;

(ii) Most serious prior/other offense;

(iii) Escape History;

(iv) History Institutional Violence;

(v) Gang Affiliation Status; and

(vi) Current Age.

(Ex. 1310 at 12). The technical manual assigns numeric "points" for gradations within each of the six criteria. The total number of points is then used to determine the appropriate custody level. For example, an inmate who scores a total of thirty-eight points under the six criteria likely will be assigned to maximum custody while an inmate who scores zero to nine points likely will be assigned to minimum custody. (Ex. 1310 at 60).

2. Internal Risk Level

A prisoner's "Internal Risk Level" is determined based on a process very similar to that regarding custody classification. The same six criteria used to determine the custody level are also used to determine the internal risk level. (Ex. 1309 at 15). And using those criteria, each inmate is assigned one of five risk levels. A score of 5 represents the highest risk and a score of 1 the lowest. (Ex. 1309 at 14-15). There are certain guidelines, however, that impose a particular internal risk level regardless of how the inmate otherwise scores. For example, a minimum custody inmate that has "a current or prior conviction for Murder or Kidnapping" must receive an internal risk level of no lower than 3. (Ex. 1309 at 15).

The custody classification and internal risk level scores are meant to address slightly different aspects of each prisoner's appropriate housing location but, given that they are using some of the same characteristics of each prisoner, the scores often overlap. Moreover, the internal risk level is secondary to the custody level in that a prisoner placed in maximum custody will have his movement drastically limited regardless of his internal risk level. (Doc. 4266 at 164).

3. Overrides

The custody classification and internal risk level scores are both subject to "overrides."⁴² "[O]verrides are basically exceptions," meant to address situations where the classification scheme may not be reliable. (Doc. 4266 at 165). For example, there may be "confidential information that is not reflected in the criminal history or in the crime itself" that should be considered when deciding how to classify a prisoner. (Doc. 4266 at 165). Overrides are either "discretionary" or "non-discretionary."

42 Pursuant to policy, internal risk level overrides may decrease, but not increase, a prisoner's internal risk level. (Ex. 1310 at 45).

Discretionary overrides that increase custody level are appropriate "whenever the inmate's behavior or new information indicates increased security measures are appropriate to ensure the safety of the public, staff, and/or other inmates." (Ex. 1309 at 11). Discretionary overrides may be appropriate if an inmate is a high-profile individual, his crimes were "depicted as heinous," or if there is some "other major reason." (Ex. 1309 at 12). There was no testimony or evidence offered at trial how these determinations are made. For example, there was no explanation which crimes will be deemed "heinous" for purposes of a discretionary override.

In addition to discretionary overrides, DO 801 provides for "non-discretionary overrides." (Ex. 1309 at 6). The non-discretionary overrides impose simple bright-line rules based on particular characteristics. Thus, regardless of how the technical manual would score a death-sentenced prisoner, he must always be classified as either maximum or close custody. (Ex. 1309 at 6). Another non-discretionary override is that an adult male, regardless of crime, who is serving a life sentence must always serve the first two years in maximum custody. This is the default starting point, but reduction to close custody is, allegedly, possible "on a case-by-case basis." (Ex. 1309 at 6). However, it was undisputed at trial that as a matter of practice, prisoners beginning a life sentence are never eligible to be reclassified until they have served two years in maximum custody. (Doc. 4268 at 49).

Defendants’ approach to overrides creates the possibility prisoners will be "overclassifie[d] ... based on who he or she is or who he or she harmed." (Doc. 4266 at 169). The classification and override processes also create the situation where a single characteristic is used repeatedly. For example, escape history is used when calculating a prisoner's custody classification as well as his internal risk level. But escape history can also be used to determine if an override is appropriate. (Doc. 4267 at 8). Defendants offered no evidence at trial explaining the reason their classification policies use the same characteristics over and over again. Overall, DO 801 is drafted and implemented with a strong predisposition towards housing prisoners in very restrictive housing.

B. DO 804

DO 804 sets forth Defendants’ "Inmate Behavior Control" policy. (Ex. 1312). DO 804 "describes the operation of prison Detention Units as a means to temporarily or permanently separate inmates from the general population to preserve the safe, secure and orderly operation of an institution." (Ex. 1312 at 3). The policy allows for an inmate to be placed "in detention status" if certain criteria are met.⁴³ There was no evidence offered at trial to explain the exact meaning behind each of the seven criteria for placement in detention status. A Deputy Warden testified DO 804's reference to "observation status" is implicated when a prisoner is acting strange "but mental health does not place them on a mental health watch." (Doc. 4273 at 53). That Deputy Warden did not know, however, what "to fulfill disciplinary sanctions" meant because ADCRR allegedly does not have "disciplinary detention." (Doc. 4273 at 54).

43 Detention may be merited:

1. To ensure the safe, secure and orderly operation of the institution/facility.

2. To ensure the integrity, and pending completion, of an ongoing investigation.

3. While determining eligibility for Protective Custody.

4. For observation status, to identify, minimize and intervene in the possibility of self-destructive behaviors.

5. Pending institutional review and classification placement, such as pending transfer to a higher custody level.

6. Pending revocation of parole, work furlough, home arrest, or temporary, mandatory or provisional release.

7. To fulfill disciplinary sanctions. (Ex. 1312 at 4).

Detention status pursuant to DO 804 is technically distinct from being classified as maximum custody. However, they are very similar in practice. As explained by a Deputy Warden, "[p]eople who are in detention are treated like they're in max custody." (Doc. 4264 at 184-185). In fact, prisoners in detention are treated as having the highest possible custody classification as well as the highest internal risk level. (Doc. 4264 at 185).

Significantly, pursuant to DO 804, meals for prisoners in detention units should be provided at the same time, and of the same quality, as meals provided to prisoners in general population. In addition, DO 804 states prisoners in detention will be given the opportunity to shower three times per week and the opportunity to exercise outside of their cell for two hours on three different days each week. (Ex. 1312 at 5). Thus, if DO 804 were followed, prisoners in detention would be offered three showers and no less than six hours of out-of-cell recreation per week.

C. DO 807

DO 807 is the policy regarding "Inmate Suicide Prevention, Mental Health Watches, and Progressive Mental Health Restraints." (Ex. 1315). DO 807 sets forth the training requirements for correctional officials regarding suicide prevention and outlines the various types of "Mental Health Watches" an at-risk prisoner may be subjected to when appropriate. For example, depending on a prisoner's needs, he or she may be placed on "Continuous Mental Health Watch," "10-Minute Mental Health Watch," or "30-Minute Mental Health Watch." (Ex. 1315 at 12-15). Those watches require observation by a correctional officer either continuously or at the specified intervals (i.e. , ten or thirty minutes). DO 807 states all prisoners on a mental health watch should be served meals "of the same quantity and nutritional quality as meals served to the general population." (Ex. 1315 at 11). And, "[u]nless determined contraindicated by a licensed mental health professional; [sic] showers, telephone privileges, recreation, and visitation shall be made available to the inmate." (Ex. 1315 at 12). DO 807 does not state the amount of recreation that should be provided to prisoners on a mental health watch.⁴⁴

44 Prisoners often "go back and forth between [mental health] watch and their maximum custody location." Those prisoners are called "frequent fliers." (Doc. 4273 at 110). One Deputy Warden knew of "between five and ten" prisoners who have gone back and forth between maximum custody and mental health watch. (Doc. 4261 at 114).

D. DO 812

DO 812 is the "Inmate Maximum Custody Management and Incentive System." (Ex. 1318). It establishes "a system that requires inmates in Maximum Custody to work through a program, utilizing a step incentive system, providing the opportunity to participate in jobs, programs, and other out of cell activities." (Ex. 1318 at 3). Prisoners in maximum custody "begin the program in Step I, with Step I being the most restrictive, and Step III being the least." (Ex. 1318 at 5). There are many different varieties of "maximum custody" and some varieties are subject to different limitations.⁴⁵ For the most common type of maximum custody, a prisoner must spend a minimum of 30 calendar days at Steps I or II before being eligible for advancement to the next step. And prisoners who maintain "Step III for a minimum of 30 consecutive days, without incident, are eligible for consideration for placement in a Close Custody housing location." (Ex. 1318 at 6). DO 812 is effectively a program whereby prisoners in maximum custody are told they can earn their way into more privileges and, eventually, be placed in less-restrictive housing settings.

45 DO 812 identifies one set of procedures and benefits for the "maximum custody population" and different procedures and benefits for prisoners in "restrictive/enhanced status housing program." (Ex. 1318 at 5). Because the vast majority of prisoners in maximum custody are not in the "restrictive/enhanced status housing program," the Court will focus on the policy's application to "maximum custody."

DO 812 specifies the exact privileges for prisoners at each step and those privileges vary slightly based on location. For example, prisoners in the Eyman-Browning unit who are at Step I are allowed one phone call per week as well as one non-contact visit. Prisoners at Step III at the same location are allowed three phone calls per week as well as three non-contacts visits. (Ex. 1318 at 11). The same limits apply to prisoners in the Lewis-Rast unit with the exception that Rast prisoners at Step III are allowed a weekly contact visit. (Ex. 1318 at 13). For present purposes, the most relevant aspect of DO 812 involves the amount and type of recreation opportunities each prisoner is supposed to be provided. Pursuant to DO 812, the table at Appendix 6 identifies the recreation that should be offered to a maximum custody prisoner at a particular location and at a particular step. The most crucial aspect of DO 812 is that no prisoner in maximum custody should be offered less than 7.5 hours of recreation per week while prisoners may, some weeks, receive up to 12.5 hours of recreation in a single week.

When asked at trial about DO 812's requirements, a Deputy Warden stated, with the exception of Step II prisoners at Browning, recreation opportunities must be in the enclosures of the size specified.⁴⁶ (Doc. 4259 at 104-05). Thus, if DO 812 were being faithfully applied, a general population Step III prisoner at Browning would receive one 2.5-hour block in a 20x40 enclosure every month as well as, each week, one 2.5-hour block in a 10x10 enclosure and two other blocks in the standard enclosure or larger. (Doc. 4259 at 105). That would mean at least 7.5 hours of out-of-cell recreation every week, some of which would be with other inmates to allow for socialization.

46 The Court reached the same conclusion regarding the mandatory nature of the enclosure sizes while the Stipulation was in effect. (Doc. 3861 at 10).

DO 812 contemplates "periodic classification reviews of inmates ... every 180 calendar days or less." (Ex. 1318 at 10). There was testimony explaining how these periodic reviews are conducted at Eyman-Browning as well as Eyman-SMU. At both locations, prisoners have their progression reviewed each month. To accomplish this, officials at both locations hold weekly meetings. (Doc. 4259 at 114, 117; 4273 at 22). Those meetings involve the deputy warden or the associate deputy warden, as well as other correctional officers. (Doc. 4264 at 150; 4273 at 22). Those officials review each prisoner, including that prisoner's "behavior, their participation in programming, their disciplinary and their overall interactions with staff." (Doc. 4264 at 150; Doc. 4273 at 22-23). At Eyman-Browning, each meeting lasts approximately 90 minutes, meaning approximately 360 minutes are spent on step reviews each month. (Doc. 4259 at 117). That means, because there are 700 maximum custody prisoners at Browning, on average each prisoner is reviewed for approximately thirty seconds each month. At Eyman-SMU, each meeting lasts between 30 and 60 minutes, meaning each month there is approximately 240 minutes spent on reviews. (Doc. 4273 at 23). Again that means, on average, each prisoner is reviewed for approximately 30 seconds each month.

Given the number of prisoners, the length of meetings, and that the relevant records can be incomplete and hard to read, it is not possible officials are conducting a careful and meaningful review of each prisoner each month. DO 812 contemplates step progression and eventual release from maximum custody but the evidence shows actual operation establishes DO 812 is not being applied in a systematic and consistent way. Prisoners sometimes change step levels or are reclassified to lower custody levels but those actions occur almost at random. Prisoners may remain at certain steps for lengthy periods, even though they have not been subject to any disciplinary actions. Prisoners also attain Step III and remain at that step for years. For an egregious example, one prisoner attained Step III in October 2015 and remained in maximum custody, at Step III at least through August 2020. (Ex. 1220 at 33).

DO 812 contemplates step progression but ADCRR does not calculate how long prisoners spend in maximum custody or how long prisoners spend at each step. (Doc. 4259 at 120). A Deputy Warden testified ADCRR does not track these figures, but ADCRR does maintain an electronic database containing information on when a prisoner entered maximum custody, when a prisoner had a change in his step level, and when officials conducted a review of his current step level. (Ex. 1220). The evidence indicates that database is used solely to record relevant dates and events, but ADCRR does not use that data to calculate the amount of time spent in maximum custody or at a particular step level for each prisoner. A conscientious effort to follow DO 812 would, at the very least, require ADCRR have a way of ensuring prisoners progressed through the steps on a timely basis.

E. DO 813

DO 813 outlines ADCRR's "Close Management" process.⁴⁷ (Ex. 1319). When a prisoner participates in one of five identified behaviors, such as being an active participant in a riot, he may be placed in "Close Management status." That status is designed for prisoners "unable to live in general population" but who do not merit "Maximum Custody placement." (Ex. 1319 at 3). DO 813 establishes three "phase levels" that prisoners may advance through but there was no evidence explaining how these phases work. All prisoners in close management status are eligible for six hours per week of outdoor exercise. (Ex. 1319 at 11).

47 "Close management" is different from "close custody," one of the four custody classifications.

F. Summary of Restrictive Housing Policies

Based on Defendants’ classification and housing practices proven at trial, the following prisoners qualify as subclass members:

(a) Formally classified as "maximum custody" pursuant to DO 801;

(b) Housed in a detention unit pursuant to DO 804;

(c) Placed on mental health watch pursuant to DO 807; and

(d) Placed in close management status pursuant to DO 813.

III. Subclass Population

In their post-trial submissions, the parties agree these four classifications cover the subclass population. Using these four classifications, and the population figures contained in Exhibit 1304, the subclass consists of the 2,667 individuals reflected in Appendix 7.⁴⁸ That number is slightly larger than either side offers but it is important to note the exact number is not crucial. Defendants concede the subclass consists of at least 2,472 prisoners and that number undoubtedly undercounts to some extent. Moreover, the subclass size fluctuates as prisoners move in and out of particular classifications or are released from custody. Thus, the Court will establish 2,667 as the subclass population while recognizing the exact number on any given date might be slightly higher or lower.

48 While both parties rely on Exhibit 1304 to calculate the size of the subclass, they have inexplicable disputes regarding the total numbers calculated from that exhibit. According to Defendants, Exhibit 1304 establishes there were 1,636 prisoners in maximum custody, 750 prisoners on detention status, 81 prisoners on mental health watch status, and 7 prisoners on close management status. (Doc. 4309 at 346). Thus, Defendants argue there were a total of 2,474 subclass members as of September 30, 2021. Plaintiffs counter that Exhibit 1304 shows 1,801 prisoners in maximum custody, 750 in detention, and 22 in close management for a total of 2,573 subclass members. (Doc. 4314 at 90). Plaintiffs do not reference prisoners on mental health watch and do not explain the omission.

Beyond disputing the precise number of subclass members, the parties disagree on the relevant total number of ADCRR prisoners to use when comparing the subclass population to the entire prison population. Defendants argue the Court should consider all prisoners "housed in both ADCRR-operated and contract prison complexes." (Doc. 4315 at 102). Doing so would mean a total prison population of 35,410. Plaintiffs argue the Court should consider only prisoners in the ten ADCRR-operated facilities, which would mean a total prison population of 27,794. (Doc. 4314 at 89). Given that this lawsuit challenges conditions in ADCRR-operated facilities, the population in those facilities will be used. In addition, using the population in ADCRR-operated facilities allows for a helpful comparison between Arizona and other states using one of Defendants’ trial exhibits.

Defense exhibit 3530 is a report titled Time-In-Cell 2019: A Snapshot of Restrictive Housing based on a Nationwide Survey of U.S. Prison Systems. The report was published by The Correctional Leaders Association & The Arthur Liman Center for Public Interest Law at Yale Law School. The drafters of that report aimed to provide a statistical overview of the number of prisoners in each state's prison facilities and the number of those prisoners subject to "restrictive housing," defined as "in a cell for an average of twenty-two hours or more per day for fifteen days or more." (Ex. 3530 at ADCRR00231467). The drafters provided a definition for survey respondents to use that excluded individuals in privately run prisons. (Ex. 3530 at ADCRR00231472). At the time of the study, the average percentage of the prison population in restrictive housing across all responding states was 3.8%. As for Arizona, it reported a total prison population of 42,312 with 1,934 (4.6%) in restrictive housing. Arkansas reported the highest percentage in restrictive housing of 11% while Colorado, Delaware, North Dakota, and Vermont the lowest percentage of 0%. Arizona's percentage of 4.6% was the 14th highest. (Id. at ADCRR00231475).

Arizona's prison population decreased significantly since the 2019 report. Now, using the definitions provided by the study's authors, Arizona has 27,794 total prisoners. However, as of 2021 there are more prisoners in restrictive housing (i.e. , 2,667) than what Arizona reported for the 2019 study. Using the current number of prisoners in restrictive housing and the number of prisoners in ADCRR-operated facilities means Arizona now has 9.5% of its prisoners in restrictive housing which is over double the 2019 national average. Because of the high percentage of its prisoners in restrictive housing the Court agrees with Horn that "the ADCRR overuses isolation. There are more inmates held in restrictive housing than are necessary." (Doc. 4166 at 101).

The number of prisoners in maximum custody (1849) and detention (746) dwarf the number of prisoners on mental health watch (50)⁴⁹ and close management status (22). Thus, the Court's findings and analysis will focus on maximum custody and detention. For those two statuses, the manner in which activities are tracked is different and the out-of-cell offers are quite different. The Court will analyze the treatment of these two statuses separately.

49 Based on the evidence presented at trial, there likely are even more prisoners who should be on mental health watch.

A. Maximum Custody ⁵⁰

50 Maximum custody and close management are, in effect, the same. The trial evidence focused on the manner in which the activities of prisoners in maximum custody were tracked. But it is unclear whether the activities of the prisoners in close management are tracked using the same form.

1. Classification Policies and Practices Lead to Overclassification

Before examining the conditions in maximum custody, it is important to understand how prisoners are classified to see how classification procedures impact operations. As set forth earlier, DO 801 and its related practices and policies are designed such that large numbers of prisoners are classified as maximum custody, even when the prisoner's characteristics would support a lower custody level. Horn provided examples of the "overclassification" produced by DO 801 and related practices. These examples show how Defendants’ policies and practices result in a large maximum custody population even when officials believe there is no penological justification that supports keeping certain prisoners at that custody level.

Prisoner S.C. entered Defendants’ custody on April 1, 2019, to begin serving a life sentence. (Doc. 4166 at 30). S.C. "scored" for placement in medium custody but, employing an override, S.C. was recommended to be placed in maximum custody because he had "currently served less than 2 years on a Life Sentence." (Doc. 4166 at 31; Exhibit 3611 at 3). S.C.’s placement was reviewed in October 2019. His classification score recommended minimum custody but he remained in maximum custody because he was "serving first two years in max." (Ex. 3611 at 19). S.C.’s placement was reviewed in March 2020 with identical results. (Ex. 3611 at 17).

S.C.’s placement was reviewed again in October 2020. At that time, an official erroneously stated S.C. had served two years. Based on that, and the lack of any disciplinary history, the official recommended S.C. be reclassified to the lower level of "close custody." (Ex. 3611 at 15). Two deputy wardens agreed S.C. could be managed at close custody. The "Classification Administrator," however, disagreed and, without explanation, decided S.C. was to remain in maximum custody. (Ex. 3611 at 14). For unknown reasons, officials completed forms on May 3, 2021, May 17, 2021, and June 16, 2021, recommending S.C. be reclassified to close custody based on his custody score and lack of disciplinary history. (Ex. 3611 at 8-12). All those recommendations were rejected. At some unknown time past the two-year period, S.C. was eventually reclassified to a lower custody level and left maximum custody. Overall, S.C. spent a significant amount of time in maximum custody despite multiple prison officials repeatedly recognizing a maximum custody placement was unnecessary.

Another prisoner, S.M., had a similar experience with the classification system. S.M. entered custody on June 14, 2019, to begin serving a life term. He was placed in maximum custody despite his custody scores showing he was eligible for medium custody. (Doc. 4166 at 34). His maximum custody placement was reviewed over the following years, with his custody score not changing, but S.M. remained in maximum custody because of his "sentence structure." (Doc. 4166 at 34). In June 2021, his placement was reviewed, and his scores justified minimum custody placement. But officials recommended he be transferred to close custody which occurred on July 15, 2021. (Doc. 4166 at 35).

Regarding S.C. and S.M., Horn stated "there was no penological justification" for keeping both in maximum custody for two or more years. (Doc. 4166 at 33, 35). Horn further opined, in his "experience and [ ] opinion," there was no penological justification for the policy that all life-sentenced prisoners spend the first two years in maximum custody. (Doc. 4267 at 26). Horn was unaware of any prison system that has a similar requirement. (Doc. 4267 at 27). The only penological justification offered by Defendants for this policy was from Director Shinn who testified, without elaboration, the policy was "[t]o prevent harm in our facilities." (Doc. 4274 at 69). And when Director Shinn had been asked for the penological justification behind the policy at his deposition, he responded "I don't have that information with me." (Doc. 4274 at 70). Director Shinn's attempted justification was not credible given the documented instances of other correctional officials concluding individuals subject to the policy could safely be housed at lower custody levels (even minimum custody). In other words, officials with actual knowledge of the prisoners’ behavior and risks believed the prisoners could be housed at lower custody levels without presenting a risk of harm.

Horn highlighted problems with other classification policies of Defendants. A prisoner, V.S., entered custody in August 2018, but the records regarding him were incomplete and reflected he was reclassified (maximum, close, medium) over a three-year period. (Doc. 4166 at 40). In July 2020, V.S.’s custody scores reflected medium custody and medium risk. But in September 2020 V.S. told officials he "did not feel safe" in the housing unit where he was living. (Doc. 4166 at 40). V.S. was then "approved for maximum custody because of ‘exhausting all other housing options.’ " (Doc. 4166 at 41). As correctly observed by Horn, "[n]ot being able to figure out how to keep a medium security inmate safe is not a justification for placing that person into isolation." (Doc. 4166 at 41).

Horn also referenced an inmate, J.J., who in 2012 committed a serious violation by kidnapping an individual working at the prison for a private contractor and holding her hostage for 30-40 minutes. (Doc. 4166 at 36). As a result, J.J. was placed in maximum custody. His file included a note that he was "not to be removed from max custody without approval from the [Offender Standards Bureau Administrator]." (Doc. 4166 at 36). Over the following years, J.J.’s placement was reviewed and, while his scores reflected "medium custody and medium internal risk," he remained in maximum custody. Based on how Defendants’ policies operate, despite that the 2012 kidnapping offense was included when calculating J.J.’s custody scores, he still scored the lower medium custody. (Doc. 4267 at 39). In April 2018, a classification officer stated J.J. had remained "disciplinary free since 2012" and he had "shown the ability to be managed at a lower custody unit." (Ex. 3616 at 64). J.J.’s custody scores at that time reflected medium custody and medium internal risk. The officer recommended J.J. be transferred to close custody. That recommendation was rejected, J.J. remained in maximum custody, and, at the time of trial, J.J. was still in maximum custody.⁵¹

51 The records indicate J.J.’s offense was sometimes noted as having occurred in 2015 instead of 2012. Horn noted the failure to keep accurate records could lead to overclassification because offenses may appear more recent than the reality. (Doc. 4267 at 33-34).

While acknowledging the seriousness of J.J.’s 2012 offense, Horn stated there was "no rational basis for keeping [J.J.] in maximum custody" given the nine subsequent years of good behavior. (Doc. 4166 at 38). Defendants offered no evidence of a penological justification for keeping J.J. in maximum custody despite his custody scores and the recommendation by officials to reclassify him. Instead, Defendants merely stated, without record citation, that "J.J. poses a substantial risk of harm to staff and thus his classification is justified" despite their own internal seemingly valid recommendation that J.J. could be reclassified. (Doc. 4315 at 144).

Horn cited a similar situation involving inmate T.A. While in a lower custody level, T.A. received six disciplinary tickets and, in June 2019, he was reclassified to maximum custody. (Doc. 4166 at 38-39). In June 2020, a classification officer stated T.A. had "shown great improve[ment,] he can be managed at close custody." (Ex. 3671 at 14). T.A. was approved to be reclassified as close custody on July 15, 2020. (Ex. 3671 at 14). Despite the reclassification, T.A. remained in maximum custody. T.A. subsequently committed another violation in April 2021 that again triggered maximum custody status. That violation was based on an incident where T.A. refused to return handcuffs after he was in his cell. (Ex. 3671 at 30). If T.A. had been treated according to his classification in "close custody," he would not have been restrained and the disciplinary incident would not have occurred.

Regardless of the April 2021 disciplinary charge, T.A. was meant to leave maximum custody in July 2020 but, inexplicably, he remained in maximum custody for the following nine months. Horn accurately described this situation as "evidence of the irrationality and unfairness of maximum custody in ADCRR." (Doc. 4166 at 39). During cross-examination, defense counsel asked Horn if he was aware that, as of the time of this trial, T.A. had been classified and was being housed in a close custody unit. (Doc. 4268 at 59). Why defense counsel thought it helpful that T.A.’s placement as of November 2021 had changed was not explained. Horn's point was that T.A. had remained in maximum custody for months after being reclassified. The reclassification of T.A. just in time for trial shows consciousness of guilt.

A final aspect of how DO 801 is being implemented involves prisoners who are classified as maximum custody, and subsequently classified to a lower level, but remain housed in maximum custody. The Deputy Warden at Eyman SMU, Lori Stickley, testified there are many prisoners currently housed in maximum custody at that location who have already been reclassified to a lower level and were merely waiting for "beds [to] become available." (Doc. 4264 at 192). At trial, Stickley testified there were approximately 150 prisoners in maximum custody at Eyman SMU who had not been classified as maximum custody. (Doc. 4264 at 193). Stickley did not know how long those prisoners had been waiting to be moved. The Deputy Warden at Eyman Browning, Travis Scott, testified there were 44 prisoners at that location "that were currently waiting to go to a close custody unit" but space was unavailable. (Doc. 4264 at 148-149). Scott did not know how long those prisoners had been waiting. (Doc. 4264 at 172). Added together, Eyman SMU and Eyman Browning established approximately 200 prisoners being held in maximum custody despite officials conceding they should not be there.

Overall, DO 801 and its related policies and practices provide a mechanism for the classification of prisoners as maximum custody and housed in these restrictive settings without penological justification and even when DO 801 is correctly implemented and prisoners are reclassified to lower custody levels, they may well not leave maximum custody. In short, Defendants keep hundreds of prisoners in maximum custody housing despite all prison officials admitting there is no penological justification for doing so.

The Court does not question the wisdom or lawfulness of Defendants’ policies regarding classification of prisoners and where prisoners are housed. Defendants may well have the authority to automatically place prisoners in maximum custody for the first two years of a life sentence, regardless of that prisoner's custody score. The basic wisdom of such a policy is not before the Court. However, the specific operation of the policies regarding who is placed in restricted housing are crucial to assessing the conditions imposed on the subclass. In particular, the policies and implementation of them regarding the subclass population establish the subclass does not consist solely of uniquely dangerous individuals. At trial, defense counsel suggested prisoners in maximum custody or other restrictive housing units are there because they were too dangerous to house anywhere else. (See Doc. 4263 at 41; Doc. 4268 at 54). That is simply not true for all subclass members. Defendants’ own policies show many subclass members do not present such dangers. In fact, many subclass members are not even formally classified as "maximum custody" prisoners. Yet these subclass members remain in restrictive housing.

2. Subclass Custody Conditions

The two maximum custody housing locations with the largest populations are Eyman-SMU I ("SMU") and Eyman-Browning ("Browning"). Altogether, SMU and Browning house approximately 1,300 members of the 2,667 subclass members.⁵² Plaintiffs provided significant detail regarding these units. The descriptions of other facilities was not as detailed. Thus, the most detailed factual findings regarding physical conditions and general practices relate to SMU and Browning. There is sufficient evidence to make limited factual findings regarding other locations.

52 At trial, Deputy Wardens testified Browning usually houses approximately 700 prisoners while SMU houses over 500. (Doc. 4259 at 92; 4264 at 179).

(a) SMU and Browning

Plaintiffs’ experts testified there may be "subtle differences" between SMU and Browning but, in general, "[t]hey feel the same when you enter them." (Doc. 4306 at 43). SMU and Browning are "very severe, dehumanizing units." (Doc. 4306 at 42). The units were "designed to be maximally controlling" and are "maximally depriving" in practice. (Doc. 4306 at 42).

The cells in these units are "less than 8 feet by 12 feet" and do not have windows.⁵³ (Doc. 4306 at 42; 4122 at 61). Inside the cells are "a bare concrete box with a metal stool, shelf, toilet/sink and either a single or double slab for a bed." "The doors to the cells have no windows but are made of perforated steel." (Doc. 4122 at 61). Thus, the only natural light in the units comes from "gritty skylights in the ceiling outside the cells." (Doc. 4122 at 60). The cells "have 24-hour illumination" and prisoners often cover the light. (Doc. 4122 at 52). The constant illumination makes it "difficult to sleep." (Doc. 4263 at 44). The lack of natural light, combined with the constant illumination, "adds to [the] disorienting nature of the conditions." (Doc. 4122 at 52).

53 The cells for subclass members were estimated at trial to be approximately 80 square feet. (Doc. 4268 at 90). Defense counsel hinted the cells are between 88 and 89 square feet. (Doc. 4268 at 90). It is undisputed the cells are somewhere between 80 and 90 square feet.

Evidence was not offered regarding the limitations on personal property for prisoners in these units other than that, when they are brought to the Browning "intake" unit, prisoners initially do not have any property. Officials "try" to get prisoners their property "in the first week" of arriving. (Doc. 4264 at 139).

The cells in these units are "infested with insects, roaches and crickets, and infested also with rodents." (Doc. 4306 at 44). During his visit, a prisoner showed Dr. Haney a rodent the prisoner had captured. (Doc. 4306 at 44). There is also mold growing on the walls throughout the SMU unit.⁵⁴ Dr. Haney encountered "a large puddle of musty-smelling and moldy water outside one person's cell that came from a leak inside." (Doc. 4122 at 60). Overall, the "cells give a sense of being entombed in a small, concrete box" and are "barren, stark, and ... dehumanizing." (Doc. 4122 at 61; Doc. 4306 at 42).

54 The showers at both units also have mold growing in them and are in a state of "disrepair." (Doc. 4306 at 46).

The cells have slots through which food trays are passed. Those food trays are "worn and dirty." (Doc. 4122 at 59). Prisoners are supposed to be fed twice per day. (Doc. 4273 at 55-56). This is not consistent with general industry practice, which suggests prisoners should receive three meals a day, at "regular mealtimes," including two hot meals. (Doc. 4166 at 77). Around 4 or 5 in the morning, prisoners receive a "mega sack" containing breakfast and lunch. (Doc. 4273 at 64-65). A mega sack is "mostly bread and served cold." (Doc. 4166 at 77). A mega sack may also contain milk, but if it is not consumed immediately, the milk will likely curdle. (Doc. 4166 at 77). Prisoners are supposed to receive a "hot dinner." (Doc. 4273 at 64); (Doc. 4278 at 29). The meals are "repetitive, unappetizing, and insufficient." (Doc. 4166 at 77). Prisoners "remain hungry most of the time." (Doc. 4122 at 52); (Doc. 4263 at 65) (prisoner testified he did not receive enough food while at SMU I). Prisoners often must supplement their meals by purchasing food, if they have funds to do so.⁵⁵ (Doc. 4166 at 77). The Warden of one complex, who had years of experience in maximum custody units, testified the "top complaint" made by prisoners was about "food, commissary." (Doc. 4278 at 36). In particular, prisoners complain "they're not getting the portions of food they're supposed to be getting." (Doc. 4278 at 36).

55 One prisoner testified he was unable to buy additional food because he could not afford to do so. (Doc. 4263 at 65-66).

Prisoners in these units must be strip searched each time they leave their cells. (Doc. 4259 at 100). When out of their cells, most of the prisoners in Browning (and presumably SMU as well) must be handcuffed behind their backs and escorted by one officer. (Doc. 4259 at 99). However, if the prisoner is housed in a cell with another prisoner, two officers are required to remove the prisoner from the cell and to return him to the cell. (Doc. 4259 at 99). Some prisoners must be escorted by three officers and wear leg irons and handcuffs in back. (Doc. 4259 at 100).

Prisoners in these units have effectively no regular social contact. The only "regular contact" prisoners have are "the brief, routinized ‘interactions’ that occur twice a day, when they receive their meals," assuming prisoners are fed that day. (Doc. 4165 at 53). Prisoners may have some contact with mental health staff, but that contact is rare. (Doc. 4165 at 53). For prisoners who are visited by mental health staff, those visits may occur through "cell-front encounters." Such encounters are not confidential and prisoners are often uncomfortable talking about their issues. (Doc. 4306 at 74). If a prisoner requests a "confidential, private setting," he will be taken to a small room and placed in a "small telephone booth-like cage, still in restraints." (Doc. 4122 at 47). Maximum custody prisoners are reluctant to request these individual sessions because they must be strip searched before leaving their cells, placed in restraints, and placed in the cage, still in restraints. The treatment cage is "obviously inhospitable" and it makes it "difficult" for a prisoner to "relax and develop rapport or express their ... deep psychological concerns or fears or issues." (Doc. 4306 at 73).

In these units, a housing area, also known as a "pod" or "run," refers to an area comprised of "10 cells, five stacked on top of five."⁵⁶ (Doc. 4259 at 90-91). The term "cluster" refers to a group of pods. (Doc. 4259 at 91). Browning has six pods in each cluster and there are twelve clusters. (Doc. 4259 at 91). Each cluster has a "control room" that is intended to be staffed by an officer. Officers in control rooms are responsible for opening the doors in the cluster and documenting when certain activities occur. (Doc. 4261 at 120).

56 Some pods may have eight cells "four on the lower tier, four on the top tier." (Doc. 4273 at 95).

At the end of each pod is a recreation enclosure known as "the chute." (Doc. 4259 at 96). Each chute is "approximately 11 feet by 24 feet, with solid concrete enclosure walls approximately 15 feet high with a covering of metal mesh over the top."⁵⁷ (Doc. 4122 at 62; Doc. 4259 at 96). A chute has "no exercise equipment in it, it's just sort of a barren area that [the prisoners] are allowed to go out in."⁵⁸ (Doc. 4306 at 42). There is very little breeze and, during warmer months, entering the chute "feels like stepping into an extremely hot sauna" and the chutes have insects in them. (Doc. 4122 at 62; Doc. 4306 at 45).

57 Chutes are "probably the size of at least two cells, maybe larger." (Doc. 4267 at 46).

58 Dr. Haney did observe "a small, blue handball" in one enclosure. (Doc. 4122 at 62).

In addition to the chutes, there are recreation areas known as the "10 by 10s," a "20 by 40" enclosure, and a "50 by 90" enclosure. (Doc. 4264 at 152). The 10 by 10s are approximately ten "individual recreation cages." (Doc. 4122 at 63). Those cages are roughly the same size as a cell. (Doc. 4306 at 76). The cages have shaded portions and there is a misting system. (Doc. 4264 at 153). When viewed shortly before trial, the cages appeared to have been unused for a lengthy period, with some doors open and "rusted in place." (Doc. 4122 at 63). During four days of touring units in 2021, Dr. Haney did not observe the outdoor exercise cages being used. (Doc. 4264 at 57). Prisoners prefer the 10 by 10 enclosures to the chute because the 10 by 10s allow them to socialize.

The 20 by 40 enclosure has a basketball hoop and the 50 by 90 enclosure has a basketball hoop, a table, and an area for exercises such as pull-ups and push-ups. (Doc. 4264 at 152). There are water jugs available at some of the enclosures. (Doc. 4264 at 153).

Per ADCRR policy, officers are required to conduct periodic health and welfare checks, also known as "security checks" or "living, breath flesh checks" when prisoners are in their cells.⁵⁹ Those checks require an officer to walk through a pod to determine if there are problems with the facilities (e.g., doors working) and to ensure the safety of the prisoners. (Doc. 4273 at 33). As explained by one Deputy Warden, these checks require an officer "look inside of a cell" and determine if the prisoner is "alive" and "breathing." (Doc. 4265 at 29). Current policy requires these checks be performed every 30 to 59 minutes, at irregular intervals. (Doc. 4273 at 34). Previously, these checks were required to be performed every 30 minutes, but because of low staffing it has been reduced. (Doc. 4273 at 34-35). The industry standard is that prisoners in housing of this sort be observed twice per hour. (Doc. 4166 at 81).

59 The phrase "living, breathing flesh" is used in the formal policy. (Ex. 1742 at 1).

In addition to ADCRR adopting a policy permitting checks far less often than industry standards, records establish that when the checks do occur, they are often performed in remarkably little time. For example, a security check in a pod consisting of 8 or 10 cells may take only one minute. (Doc. 4273 at 38). According to one Deputy Warden, an officer spending only one minute on a security check was not cause for concern. (Doc. 4273 at 95). However, given the need to stop at eight or ten cells to determine the status of each prisoner and the need to climb the stairs to the upper tier, a one-minute status check would be exceptionally cursory.

The staffing levels at SMU and Browning are far below what prison officials acknowledge as necessary to operate the units safely. The Deputy Warden for Browning stated Browning has about 60% of the staff it needs. (Doc. 4261 at 115). The Deputy Warden for SMU agreed that some necessary positions and tasks are performed only because existing staff are willing to work overtime. (Doc. 4273 at 33). Plaintiffs cited the "Daily Post Sheet" for Browning on July 30, 2021, to illustrate staffing levels. Based on the Deputy Warden for Browning's reaction to this record, the day did not reflect an unusual level of staffing.

The "Daily Post Sheet" identifies each "required post" where a single officer should have been assigned. (Ex. 2131). The Deputy Warden for Browning described each of these posts as "critical." (Doc. 4261 at 117). On July 30, 2021, there was a total of "33 posts collapsed." "Collapsed means that the post wasn't staffed by a single person ... or there wasn't one person assigned to that post." (Doc. 4261 at 123). The sheet shows no officer was assigned to a variety of positions. Four officers should have been assigned as "Recreation Escort." There were none. Seven officers should have been assigned as "Mental Health/Medical Programs." There was one. Four officers should have been assigned to "Recreation Officer." There was one. Six officers should have been assigned to "Rover." There was one. Finally, there was supposed to be an officer assigned to each of the twelve control rooms. (Ex. 2131). There were seven officers assigned to control rooms, leaving five control rooms unstaffed. But according to Defendants the control rooms were not actually unstaffed because officers are assigned to multiple control rooms at once. (Doc. 4264 at 162). A single officer may be assigned to two control rooms during the day, and, at night, a single officer may be assigned to six control rooms at the same time. (Doc. 4264 at 176). With a control room in each cluster, and each cluster including six pods of 10 cells each, an officer in a control room is intended to be responsible for the housing area of 60 prisoners. When a single officer is assigned to six control rooms, that officer is responsible for the housing area of 360 prisoners. Inadequate staffing of control rooms happens "every day." (Doc. 4264 at 162).

(b) Lewis

Dr. Haney visited the Lewis complex which houses approximately five hundred subclass members. While there, he toured the Rast unit (maximum custody), the Barchey and Morey units (not maximum custody), the Stiner detention unit (maximum custody), and the Sunrise unit (juveniles). (Doc. 4306 at 41). The cells in the Rast unit are similar to those found at SMU and Browning. That is, the cells are "stark and largely barren," with no window. (Doc. 4306 at 47). Similar to SMU and Browning, being housed in the Rast unit subjects prisoners "to profound levels of deprivation." (Doc. 4122 at 65).

The Stiner detention unit is "a very stark, oppressive environment." (Doc. 4306 at 47). Unlike many other prisons, the Stiner detention unit does not have a day room or a common area where prisoners are allowed out of their cells to socialize. (Doc. 4306 at 47-48). However, at least some of the cells in this unit have a window. (Doc. 4122 at 67). The cells have a single or double bunk. (Doc. 4306 at 49). The Sunrise unit houses juveniles. That unit consists of "windowless cells built of concrete blocks, with solid doors with one small window in the door." (Doc. 4122 at 67). One cell photographed by Dr. Haney shows a bunkbed, a sink, a toilet, and a small seat which is bolted to the ground. The hallway has no natural light. Instead, it is lit only by fluorescent lights. (Doc. 4122 at 67).

(c) Other Locations

Plaintiffs’ experts toured the locations housing approximately 1800 subclass members. Those experts did not tour the other locations housing the remaining subclass members,⁶⁰ most of which consisted of the mental health watch or detention units at smaller facilities. (Doc. 4315 at 104). Defendants argue the failure to tour, and present evidence describing, every location means Plaintiffs have "provided no evidence regarding conditions of confinement at" the facilities not toured. (Doc. 4315 at 104). Defendants even seem to fault Plaintiffs’ experts for not touring and interviewing subclass members at locations such as the Florence Globe Unit and the Winslow Apache Unit. (Doc. 4315 at 104). But as of September 30, 2021, those units were empty. Complaining that the experts did not tour empty units is not the only problem with Defendants’ argument regarding the conditions of confinement at every location.

60 The eight facilities that house subclass members but were not toured immediately prior to trial are Douglas, Florence, Perryville, Phoenix, Safford, Tucson, Winslow, and Yuma.

Class actions often involve, to some degree, an amount of "representative evidence." Tyson Foods, Inc. v. Bouaphakeo , 577 U.S. 442, 455, 136 S.Ct. 1036, 194 L.Ed.2d 124 (2016). Here, there is no evidence regarding the precise physical layout of cells at some of the facilities nor is there evidence regarding the exact recreation enclosures at those facilities. However, the precise physical layout or exact recreation enclosures are not the core issues in this case. Rather, it is the amount of out-of-cell time offered to subclass members and whether the amount of isolation experienced by subclass members creates a substantial risk of harm. Resolving those issues will require examining the policies and practices applied to all subclass members. The evidence at trial established that, as a matter of systemwide practice, subclass members are treated the same regardless of exact location. That is, the policies and practices applied to subclass members are the same at the toured facilities as well as the facilities not toured. To satisfy their burden of proof, Plaintiffs were not required to present evidence and proof of every cell and every prisoner. Instead, Plaintiffs established through testimony and corroborating evidence the same practices apply throughout the system. In fact, Defendants did not even dispute this point during trial nor offer countervailing evidence. For example, Defendants did not present evidence the policies and practices differed across facilities. Accordingly, less evidence regarding some locations does not overcome the preponderance of evidence establishing subclass members are treated the same.

3. Maximum Custody Tracking

The activities of prisoners in all maximum custody units are tracked using a form titled "Maximum Custody Daily Out-of-Cell Time Tracking." (Ex. 1290). That form is meant to reflect a prisoner's activities on a weekly basis. It requires officers fill in when a prisoner is offered recreation, the size of the enclosure where the recreation would take place, if the prisoner refused the offer, and, if the prisoner accepted the offer, the time spent at recreation. The form also requires officers document if the prisoner was offered and attended other out-of-cell programs or classes. Finally, if the prisoner was categorized as Seriously Mentally Ill, the form requires officers track any additional hours of out-of-cell time. (Ex. 1290). During the life of the Stipulation, the forms for twenty randomly selected prisoners in maximum custody units were compiled monthly into what the parties refer to as "maximum custody notebooks."⁶¹ (Doc. 4264 at 126).

61 Those notebooks contained a variety of other documents as well, such as reports explaining which activities were canceled. (Ex. 1681 at 2). Each notebook contains hundreds of pages. For example, the June 2021 notebook for Eyman-Browning contains 302 pages. (Ex. 1681).

The forms purport to indicate the total amount of out-of-cell time offered to each inmate. But they do not. The form requires officers track all out-of-cell time that was offered, accepted, refused, or canceled. The out-of-cell time that was canceled, however, was still counted when determining the total amount of time offered to inmates. For example, the form found at page 112 in Exhibit 1681 identifies the prisoner as having been "offered" four, three-hour recreation periods during that week. Two of those offers allegedly were made but refused by the prisoner. The other two offers, however, were accompanied by a "C" notation, indicating those recreation periods were canceled. A separate form indicates those recreation periods were canceled "due to staff shortage." (Ex. 1681 at 41). Despite those cancelations, the bottom of the form states the prisoner was offered a total of 12 hours of out-of-cell time. This consciously-adopted practice of counting out-of-cell time that was not offered was the result of Defendants’ inexplicable interpretation of the Stipulation.

The Stipulation required Defendants provide specific amounts of out-of-cell time. The Stipulation also provided:

If out of cell time ... is limited or cancelled for legitimate operational or safety and security reasons such as an unexpected staffing shortage, inclement weather or facility emergency lockdown, Defendants shall make every reasonable effort to ensure that amount of out of cell time shall be made up for those prisoners who missed out of cell time.

(Doc. 1185 at 10). Defendants unilaterally, without Court authorization, interpreted this provision favorably as allowing them to include out-of-cell time canceled because of chronic staffing shortages when determining the total amount of out-of-cell time offered to inmates. In other words, a prisoner such as that found at page 112 of Exhibit 1681 was documented as being offered 12 hours of out-of-cell time, even though it is undisputed the prisoner was offered, at most, 6 hours. At trial a Warden explained ADCRR interpreted the Stipulation such that if a particular "cancelation was because of poor staffing that day and they did not have the staff to complete that class," it was to be included in the total recorded out-of-cell time offered. (Doc. 4277 at 132). If, however, the "staff just didn't want to do [the class or activity], that's not acceptable," and the time could not be included as out-of-cell time.⁶² (Doc. 4277 at 132). In February 2021, the Court by Order made clear Defendants could not rely on chronic "staff shortages" to justify cancellations of out-of-cell time. (Doc. 3861 at 9-10). However, as of June 2021 just prior to the repudiation of the Stipulation, Defendants were still invoking longstanding staff shortages as the reason for cancellation of some recreation and still counting non-existent recreation offers when determining the total amount of out-of-cell time. The fact that Defendants unilaterally decided indisputably unavailable recreation time should be counted as "offered," is a blatant illustration of how, throughout the ten years this case has been pending Defendants have not in good faith undertaken their obligations under the Stipulation, or compliance with Court Orders.

62 Deputy Warden Stickley also testified about the practice of counting out-of-cell time that was canceled. (Doc. 4273 at 48). She explained such time "counted" because it was not ADCRR that decided to cancel the class. Instead, "Centurion did not have the staff available to teach the class." (Doc. 4273 at 94). But Centurion is ADCRR's chosen agent to uphold ADCRR's commitments.

Given the manner in which Defendants calculated out-of-cell time, the numbers reflected on the maximum custody records for total out-of-cell time are inaccurate. To correct for Defendants’ obvious miscalculation regarding the amount of out-of-cell time offered, Plaintiffs presented a summary document that compiled the information contained in maximum custody notebooks from four locations from 2019 through 2021.⁶³ (Ex. 1980). That document calculated the amount of recreation allegedly offered to prisoners, i.e. , canceled recreation was excluded. This summary document establishes a few critical points.

63 One of the locations included in the summary, Florence-Kasson, was closed in September 2021. (Doc. 4277 at 113). While the conditions at Florence-Kasson are no longer at issue, the information regarding conditions in that unit are included to provide additional evidence that practices are similar across multiple facilities.

First, it establishes the total amount of weekly recreation allegedly offered to prisoners at the summarized locations from 2019 through 2021 varied from zero to over twelve hours.⁶⁴ For the most part, however, the total recreation allegedly offered to prisoners was between five and ten hours.

64 Eyman-Browning shows multiple inmates were offered no recreation in the monitored week for April 2020. (Ex. 1980 at 16). But the records for Eyman-Browning during the monitored week in October 2019 show one prisoner was offered over twelve hours of recreation. (Ex. 1980 at 24).

Second, the summary establishes recreation was routinely offered only in the "chute" or standard enclosure. For example, in Browning and SMU for the monitored week in January 2021, the only recreation offered was in the chute. (Ex. 1980 at 4-5). Thus, even before any Covid-based limitations, the chute was offered most of the time.⁶⁵

65 Browning and SMU in January 2019 indicate the chute was the offered location well over 75% of the time. (Ex. 1980 at 36-37).

Third, the rate at which prisoners were recorded as refusing offers for recreation was high, often so high as to render it impossible the recreation offers were legitimate. The table below shows the total recreation refusal rates at four units across ten points in time from 2019 through 2021.⁶⁶ The table shows the refusal rates were routinely above 80% and often above 90%.

66 Allegedly due to Covid, at Browning "all programming, class group education, outside recreation, SMI classes, and education were cancelled ... from March, 2020, through June, 2021." (Doc. 4261 at 115). During that time, the only recreation offered was in the chute and even that recreation was canceled at times. (Doc. 4261 at 116). As of June 2021, ADCRR started its "phase reopening" to restart some officer-led "programming and mental health programming" after Covid. (Doc. 4264 at 161). Despite reopening, the "SMI classes were canceled for July, August," and part of September 2021 because Centurion did not have sufficient staff available. (Doc. 4264 at 161).

Recreation Refusal Rate Table
Month and Year	Lewis Rast	Florence Kasson	Eyman SMU	Eyman Browning
Jan-19	46	80	73	83
Apr-19	38	71	80	84
Jul-19	67	58	83	80
Oct-19	78	63	86	93
Jan-20	68	90	84	62
Apr-20	80	87	95	68
Jul-20	64	92	92	61
Oct-20	70	73	90	56
Jan-21	80	77	93	46
Apr-21	n/a	63	90	29

As stated by Horn, recreation refusal rates over 80% are "striking." (Doc. 4166 ¶ 149). Based on Horn's expertise and pure common sense,

inmates like the opportunity to get out in the fresh air. They like the opportunity and there are some circumstances which they have the opportunity to at least socialize with other inmates if they use what we refer to as the 10-by-10 enclosures. And people generally need fresh air and sunshine and a change of scenery and the opportunity to get out of their cell.

(Doc. 4267 at 44-45). Thus, refusal rates as high as reflected in the records indicate officers either did not actually make the offers or officers used "any excuse to record a refusal to take advantage of outdoor exercise." (Doc. 4166 ¶ 146). The tactic of interpreting certain behavior as a refusal of a recreation offer was explained by a sign posted in a maximum custody unit. It reads:

Beginning 06-01-2020, All inmates and their Cells shall be in 704 Compliance before inmates will be allowed to go to Recreation. Also, the Inmates will be awake and ready to exit the cell when the officers arrive. Failure to be in compliance, or failure to be ready to submit to the strip search or any unreasonable delay once the officer arrives will be considered a refusal of Recreation and will be documented on the DO812 forms as such.

(Doc. 4166 ¶ 147). The term "704 Compliance" refers to Department Order 704 which is "essentially the rules of inmate behavior," such as personal grooming and cell cleanliness. (Doc. 4267 at 43-44). This policy is enforced such that a prisoner is documented as having "refused" recreation if, for example, the prisoner is not fully dressed or otherwise not "ready on the spot" when offered recreation. (Doc. 4267 at 43). Similarly, if a prisoner is asleep, that is documented as having refused recreation. (Doc. 4267 at 41). In short, "[o]fficers record things as refusals when inmates do not intend to refuse." (Doc. 4267 at 44).

While many offers for recreation may not actually be made or a prisoner is interpreted, inaccurately, as rejecting the offer, many offers are also accompanied by unreasonable consequences. Prisoners who accept an offer for recreation may be left in the recreation area for hours. (Doc. 4166 ¶ 148). That may happen even during hot temperatures in the recreation areas where prisoners lack access to water. (Doc. 4166 ¶¶ 148, 160). In addition, if a prisoner needs to use the restroom during the recreation time, officers may not take him to the restroom. Alternatively, the prisoner may be taken back to his cell, forfeiting the rest of the recreation time. (Doc. 4166 ¶¶ 148, 160).

Horn's testimony that the records do not indicate legitimate offers for recreation is credible and comports with common sense. Understandably at times, prisoners may have idiosyncratic reasons to decline recreation offers. But refusal rates of 80 to 95%—calculated based on Defendants’ own documents—make it more likely than not the recreation offers were never given or offered on unacceptable terms. For example, it is not plausible that in April 2020, the monitored prisoners in SMU received 42 legitimate offers for recreation but only two prisoners accepted, and those prisoners accepted only one recreation period each. Put in terms of a refusal rate, it is not plausible these prisoners received legitimate offers for recreation but refused 95% of those offers. Equally problematic is Defendants’ failure to question refusal rates of 95%, strongly suggesting they are aware legitimate offers are not conveyed. In short, recreation offers recorded on the Maximum Custody Daily Out-of-Cell Time Tracking forms are not made, are not legitimate, or are accompanied by unreasonable consequences (i.e. , being left outside in the heat for 6 hours without water and/or restroom access).

4. Detention Records

Defendants do not track how long inmates remain in detention units. Some inmates spend less than a month in a detention unit while others may spend six months or more in a detention unit. The detention records in Exhibit 1694 establish many inmates remained in detention from February 2021 through August 2021. (Doc. 4308 at 130 n.60). Detention is not necessarily a short-term placement.

ADCRR uses a form titled "Individual Inmate Detention Record" to "document out-of-cell time and activities for people in detention." (Doc. 4273 at 56). Officers are expected to document when inmates in detention receive meals, are allowed to shower, or are offered recreation. (Doc. 4273 at 57). The form is the only way ADCRR tracks recreation offers to inmates in detention. (Doc. 4273 at 58). During trial, the Deputy Warden at SMU was asked about the contents of particular detention records. She conceded that "according to the sheet," one prisoner did not receive dinner on four nights from February 15, 2021, through February 22, 2021. (Doc. 4273 at 66; Exhibit 1694 at 539). During that same week, the record shows the prisoner was not offered any showers or any recreation time. (Doc. 4273 at 66-67). The records for two other prisoners, in the same cluster of cells for the same week, also showed no dinner for four nights, no showers, and no recreation offers. (Ex. 1694 at 541, 545). The record for yet another prisoner in the same cluster for the same week indicates no dinner on four nights but he was offered one shower and one recreation period. (Ex. 1694 at 547). Records for even more prisoners indicate a similar pattern of missed meals and, at most, one offer for shower and one offer for recreation. (Ex. 1694 at 551-570).

Another record for a prisoner in that cluster of cells contained a notation that the prisoner "moved out," presumably meaning he left the detention unit in the middle of that week. However, the record also indicates this prisoner received meals in the detention unit, and also exchanged his laundry and linens, for two days after he left the unit. (Ex. 1694 at 583). Another form does not have a prisoner name or any other identifying information filled in but that form indicates two meals were delivered on all seven days, dinner was delivered on two days, and the inmate had laundry and linen exchanges multiple times. When asked if this particular record "could have just been filled out for no one," Stickley responded "I don't know." (Doc. 4273 at 79). Overall, the evidence establishes the detention forms often are pre-filled for the entire week and do not indicate what truly happened. (Ex. 1694 at 593).

In an attempt to summarize the conditions recorded by the forms at various locations, Horn analyzed 125 detention records at SMU detention unit for a week in February 2021. (Doc. 4166 ¶¶ 190-191). Based on his expert review, "fewer than half reflect 2 offers of recreation, about half reflect a single offer of recreation, and about a tenth reflect no offers of recreation." (Doc. 4166 ¶ 191). Those records do not "reflect[ ] any person actually went to recreation" at this location during this week. (Doc. 4166 ¶ 191). The records regarding showers were similar, with the majority of prisoners offered showers one or two times but a few prisoners were not offered any showers. While there was a high rate of purported refusal of showers, the records indicate that if prisoners accepted, they would be left in the shower for approximately an hour. (Doc. 4166 ¶ 192).

After the Stipulation was repudiated and during preparation for trial, the September 2021 records for SMU showed some improvement but still "not a single person went to recreation." (Doc. 4166 ¶ 196). In addition, the records again had entries for a prisoner who left the unit midweek, indicating the records are pre-filled. (Doc. 4166 ¶ 197).

The records for the Yuma Cheyenne detention unit for a week in February 2021 show prisoners were given three offers to shower and either two or three offers for recreation. However, not a single prisoner went to recreation that week. (Doc. 4166 ¶ 199). The records for this location in July 2021 indicate about half of the prisoners were offered two recreation periods, but not a single prisoner went to recreation. (Ex. 1700-2 at 2234-2283).

The records for the Lewis Morey detention unit in July 2021 show approximately half of the prisoners received either no or one offer for recreation, and about half received two offers, with a few prisoners receiving three offers. (Doc. 4166 ¶ 204). Three prisoners went to recreation during this week and the recreation time is listed as less than one hour. (Doc. 4166 ¶¶ 207-208). The prisoners were offered two or three showers. (Doc. 4166 ¶ 204). And the records show some prisoners did not receive three meals every day, including one prisoner who was documented as receiving no meals on one day, one meal on one day, and two meals on two days. (Doc. 4166 ¶ 205). The September 2021 records for this location show most prisoners received two shower offers and most prisoners were offered recreation twice but almost all offers were recorded as refused. The two prisoners who accepted recreation offers received 45 minutes of recreation. (Ex. 1697 at 1773, 1775).

The records for the Lewis Bachman detention unit for July 2021 are, as characterized by Horn, "alarming." (Doc. 4166 ¶ 215). The records document profoundly cruel treatment of prisoners this week. The weekly record for one prisoner indicates he was offered a shower once, which he accepted, as well as recreation once, which he refused. But for the seven-day period, this prisoner allegedly received only nine meals. (Ex. 1697 at 1275). But that prisoner was one of the lucky ones. A different prisoner that week was not offered a shower, not offered recreation, and received only seven meals. (Ex. 1697 at 1277). Other prisoners received no offers for out-of-cell time and only received one or two meals for the entire week. (Ex. 1697 at 1315-1319). Thus, officials at this unit affirmatively documented they confined many prisoners in their cells for 168 straight hours and provided them with only a few meals for the entire week. That affirmative acknowledgment is shocking enough. But the fact that no supervisory official took any action in response to this level of inhumane disregard for even feeding the prisoners indicates a profound disinterest in caring for the prisoners.

The records for the Lewis Bachman detention unit for August 2021, which was during trial preparation, are mildly better than for July 2021. Again, many prisoners were not offered any out-of-cell time. (Doc. 4166 ¶ 219). The records also indicate the prisoners routinely did not receive three meals per day. Many records show days prisoners had no meals. (Doc. 4166 ¶ 220). It is incomprehensible that prison officials were not concerned when their own records admitted at trial established blatant violations of policy regarding recreation and nutrition.⁶⁷

67 The records for other locations and other time periods establish it was commonplace for prisoners to not receive three meals per day. (Exhibits 1696, 1697, 1699, 1700). Plaintiffs accurately summarize this information. (Doc. 4308, pages 85-87).

Based on the evidence presented at trial, the treatment of prisoners in detention units is shocking. According to Defendants’ own records, prisoners in detention units are routinely offered no out-of-cell time and are routinely denied meals. In their post-trial briefing, Defendants argue the detention records simply reflect "deficient documentation," and the prisoners must have at least been fed.⁶⁸ (Doc. 4315 at 130). Defendants even steadfastly claim if officials had "determined that inmates actually were not fed or that there was a pattern of deficient completion of paperwork by a particular officer, corrective action would be taken." (Doc. 4315 at 129-130). Defendants offered no explanation, however, how the records recounted above do not reflect either a pattern of not feeding prisoners or a "pattern of deficient completion of paperwork." Defendants did not present any evidence they took corrective action in response to their own unrefuted documentation, which reflected weeks of denying prisoners in detention units any out-of-cell time and adequate nutrition.

68 Defendants blithely also argue "Plaintiffs’ reliance on cherry picked detention records for only seven out of seventeen detention units across the state proves nothing." (Doc. 4315 at 123). Records showing Defendants, across multiple locations, deprive prisoners of recreation and food certainly prove more than "nothing." It establishes their constitutional rights are being violated.

5. Mental Health Watch

When a prisoner is placed on mental health watch, all his or her activities are tracked on a form known as an "Observation Record." When Plaintiffs’ expert reviewed the Observation Records for the mental health watch unit at SMU, he noted none of the forms reflected the prisoners were receiving recreation. (Doc. 4166 at 51). The Deputy Warden at Browning testified that individuals on mental health watch at that location are supposed to be provided recreation opportunities but there are no facilities at Browning that would allow for recreation. (Doc. 4261 at 114).

6. Expert Opinion Regarding Subclass Conditions

Plaintiffs offered the expert opinions of Horn and Dr. Haney that the conditions imposed on the subclass members lack a penological purpose and present a substantial risk of serious harm. Defendants did not offer their own experts. Nor did they offer any remotely credible lay opinions contradicting Plaintiffs’ expert opinions. Thus, the Court did not have the opportunity to critically examine competing views from qualified experts and then determine which view is more credible. Rather, the Court was left with either crediting Plaintiffs’ experts who have spent their entire careers studying these issues, or rejecting those experts based solely on the unsubstantiated arguments presented by defense counsel. Based on careful review of written testimony, as well as in-court observation of direct and cross-examination, the opinions offered by Horn and Dr. Haney are credible.

(a) Horn

Horn credibly testified ADCRR "overuses isolation." (Doc. 4166 at 101). ADCRR keeps more prisoners in isolation than almost any other state and often there is no rational basis supporting placement of a particular prisoner in isolation. For example, ADCRR places numerous prisoners in isolation who, according to ADCRR's own policies, merit a lower level of custody placement. ADCRR has promulgated policies that purport to provide a path for prisoners to earn their way out of isolation. However, those policies are "unusually and unnecessarily complicated," leaving prisoners and officials confused regarding their implementation. (Doc. 4166 at 102). Indeed, the evidence reflects that ADCRR houses prisoners in isolation because it lacks lower custody cells and staff as evidenced by the hundreds of prisoners awaiting transfer out of maximum custody.

ADCRR keeps prisoners in isolation for years, including prisoners who have "been free of any discipline for years." (Doc. 4166 at 103). ADCRR claims to review the isolation placement for many prisoners on a monthly basis but those "reviews are brief to the point of perfunctory." (Doc. 4166 at 104). Prisoners are not involved in the review process and do not receive information regarding what he should do in the future to merit a lower classification. (Doc. 4166 at 104). The step-progression system contained in DO 812 "offers inmates the promise of advancement if they maintain a good disciplinary record and participate in prescribed programs, and [officials] do not meet that promise. [Officials] do not fulfill that promise" because prisoners have no way of knowing when, or if, they will be able to advance. (Doc. 4268 at 68).

The units housing subclass members do not have sufficient staff to operate in a safe manner. Staff are not available to conduct sufficient security checks "to properly determine whether an inmate is in distress or not." (Doc. 4166 at 104). The staffing shortage is exacerbated by the physical layout of the many units which preclude staff from having adequate "sight lines" or the "ability to hear inmates." (Doc. 4166 at 104). On cross-examination, defense counsel asked Horn whether New York and Pennsylvania had "the staffing luxury" such that sufficient staff could respond to an emergency without leaving critical posts vacant. (Doc. 4268 at 104). No doubt to defense counsel's surprise, Horn responded he "wouldn't call it a luxury" for the prison to have adequate staff to handle emergencies without leaving critical posts abandoned. (Doc. 4268 at 104). When pressed on this point by defense counsel, Horn explained other systems ensure sufficient staffing so any emergency response will not require staff involved in "direct supervision of inmates" abandon their posts. (Doc. 4268 at 105).

The lack of staff is also the obvious explanation for why prisoners do not receive the out-of-cell time promised by ADCRR policy. Based on the "actual practices," prisoners "do not get to leave their cells for the one hour per day five days a week recommended by the profession" nor do they get the hours promised by ADCRR policy. (Doc. 4166 at 105). Staff use "a variety of means to obscure the reality of what is happening." (Doc. 4166 at 105). Despite what may be indicated on the forms, prisoners "are often recorded as refusing recreation when they have not" and, in any event, out-of-cell time is "often cancelled." (Doc. 4166 at 105). On cross-examination, Horn explained prisoners should be allowed outside recreation even if they "are not following the rules and don't have their cells in compliance." (Doc. 4268 at 109). When asked whether staff should accommodate any prisoner who is not immediately ready to accept an offer for recreation, Horn stated "outside exercise, the ability to get fresh air when you're confined in these conditions, is so important that it is worth some level of effort" by the staff. (Doc. 4268 at 111).

In contrast, Defendants did not present any meaningful evidence or testimony casting doubt on Horn's conclusions. Defendants did not justify why they keep prisoners in isolation who, by their own policy and their own admission, should be housed elsewhere. Defendants did not explain why the policy regarding step-progression is implemented in a random and chaotic way nor did Defendants explain why individuals may remain at particular steps for years on end. In addition, Defendants concede they lack sufficient staff to operate the facilities as contemplated by ADCRR's own policies. Defendants’ own deputy wardens acknowledged they lack sufficient staff to have officers posted at all "critical" locations and rely on existing staff who agree to volunteer to work overtime shifts even to meet the present staffing levels. (Doc. 4261 at 117). And Defendants did not dispute that out-of-cell time is routinely canceled because of staffing shortages nor did Defendants offer evidence that the out-of-cell offers are, in fact, legitimate ones that prisoners in isolation can accept. Horn's testimony and opinions were credible and incontestable.

(b) Dr. Haney

Based on his training and experience, Dr. Haney explained "we now know that isolated confinement is painful. People who are exposed to it suffer when they experience it." (Doc. 4306 at 18). The amount of suffering, however, "[d]epends upon the conditions and the length of time that people are [in isolated confinement]." (Doc. 4306 at 18). And while some people experience harm that appears "to be reversible," others experience "very severe" harm that "can be life threatening." (Doc. 4306 at 18).

In Dr. Haney's view, the most "widespread" reaction to isolated confinement is "depression. People become despondent." (Doc. 4306 at 21). This is especially pronounced when prisoners are placed in isolated confinement for an indeterminate amount of time. (Doc. 4306 at 21). In addition to depression, prisoners placed in isolated confinement may become "unpredictably and inexplicably anxious, nervous, on edge."⁶⁹ (Doc. 4306 at 21). Alternatively, prisoners may get "angry and irritable," sometimes leading to "explosive behavior." (Doc. 4306 at 22). Isolated confinement may lead "to self-harm and it can lead to suicide." (Doc. 4306 at 23).

69 Travis Scott, a Deputy Warden with twenty years’ experience, was asked whether out-of-cell time was important. He responded "Yes. Very important." (Doc. 4264 at 151). The Deputy Warden then went on to explain "When inmates are occupied or have out-of-cell time or classroom, it helps decrease tension amongst the inmates. It makes the unit a safer place to work. The inmates are less agitated. It makes it safer for the staff for when they're interacting with those inmates. We want them to be able to get out to do their exercises or to go to class. Keeps them occupied." (Doc. 4264 at 151).

"Social contact and social interaction are essential components in the creation and maintenance of normal social identity and social reality." (Doc. 4122 at 26). Deprivation of social contact "is psychologically destabilizing" and "[i]t undermines a person's sense of self or social identity and erodes his connection to a shared social reality." (Doc. 4122 at 26). Individuals deprived of social contact "are apt to confuse reality with their idiosyncratic beliefs and fantasies and likely to act upon such fantasies, including violent ones." (Doc. 4122 at 26).

While Dr. Haney believes isolated confinement presents a substantial risk of harm to all prisoners, those prisoners with mental illness "are more vulnerable to the pains and stresses of solitary confinement." (Doc. 4306 at 25). Dr. Haney further opined prisoners under the age of 18 are especially susceptible to harm from being placed in isolated confinement. (Doc. 4306 at 25). Dr. Haney believes that even during the time the present litigation has been pending, there has developed a "broader and deeper consensus, scientific consensus," that isolated confinement is harmful. (Doc. 4306 at 26).

The American Psychiatric Association believes "[p]rolonged segregation of adult inmates with serious mental illness, with rare exceptions, should be avoided due to the potential for harm to such inmates." (Doc. 4306 at 28). The American Psychiatric Association takes a similar position regarding solitary confinement of juveniles. (Doc. 4306 at 31). The National Commission on Correctional Health Care takes the position:

(i) Prolonged, greater than 15 consecutive days, solitary confinement is cruel, inhuman, and degrading treatment and harmful to an individual's health.

(ii) Juveniles, mentally ill individuals, and pregnant women should be excluded from solitary confinement of any duration.

(Doc. 4306 at 36).

Based on his experience studying approximately 29 state prison systems, Dr. Haney credibly testified Arizona's system is "among the most severe and the most depriving solitary confinement units that [he has] encountered." (Doc. 4306 at 56; Doc. 4264 at 65). The conditions in Arizona "place[ ] the mental health of even psychologically strong incarcerated persons in jeopardy and creates especially high risks of harm for those whose mental illness makes them especially vulnerable." (Doc. 4122 at 67). He stressed this opinion was taking "into account not what the regulations provide for, but what [he] learned is actually being delivered" in Arizona. (Doc. 4306 at 56). Dr. Haney found the requirements of DO 812, in particular the amount of out-of-cell time set out in that policy, were not being followed. (Doc. 4263 at 31).

The physical attributes of the units combine with the "deprivation of social contact" to impose "severe" limitations. And "[t]he amount of out-of-cell time the prisoners are getting is severe by any measure." (Doc. 4306 at 56). While there were a few improvements shortly after the parties entered into the Stipulation, those improvements began to disappear "within a year or two." (Doc. 4122 at 91). The conditions are now the same, or worse, than the conditions present at the outset of this litigation in 2013. (Doc. 4122 at 91). The only exception is that prisoners now have tablets, which they still have to pay to send emails from. (Doc. 4306 at 78). Even then, not every prisoner is entitled to a tablet and use of the tablets is not free. (Doc. 4264 at 51-52).

On cross, Dr. Haney noted the better state systems provide "at least two hours a day" out-of-cell time. (Doc. 4306 at 99). When pressed on how long would be "too long" to spend in maximum custody, Dr. Haney noted the United Nations believes 15 days is the limit. He did not agree with that specific limit, but he noted a "month or years" is "too long." (Doc. 4263 at 16). Defense counsel asked if keeping someone in isolation longer than 15 days would mean "there is going to be psychological harm to that person." (Doc. 4263 at 17). Dr. Haney explained that causation is not that simple. Rather, when keeping individuals in isolation, "[t]here's risk of harm. There is risk of harm. So all of this is about risk of harm. And the longer somebody's exposed to these kind of conditions, the greater the risk of harm." (Doc. 4263 at 17).

Of particular concern to Dr. Haney is the Sunrise unit which houses "juveniles who have been committed to the adult prison system." (Doc. 4306 at 50). On the day of his visit, the unit contained one girl and approximately twelve boys. (Doc. 4306 at 50). Dr. Haney opined, in extremely strong terms, that the conditions in this unit were intolerable for juveniles. Dr. Haney was "shocked" that Defendants would house juveniles "in such gratuitously harsh" and "abominable" conditions. (Doc. 4122 at 67). Doing so "is singularly inappropriate and highly dangerous." (Doc. 4122 at 67). Dr. Haney noted Defendants had "learned nothing" from a previous tragic suicide with which he was personally involved.⁷⁰ (Doc. 4122 at 68). Dr. Haney opined that continuing to subject juveniles to harsh isolation practices is "playing with fire." (Doc. 4122 at 69). And the conditions in the Sunrise unit "create[ ] a very substantial risk of harm." (Doc. 4306 at 81).

70 During a visit in 2016, Dr. Haney noted a 17-year-old girl had been held in solitary confinement for more than two months. (Doc. 4122 at 68). Dr. Haney requested defense counsel be informed that this girl was at "serious risk of self-harm due to the prolonged period of time she was held in solitary." (Doc. 4122 at 68). Shortly after his visit, the girl turned 18 and was transferred to a different maximum custody unit. The girl committed suicide a few weeks later. (Doc. 4122 at 68). Dr. Haney believes Defendants’ "learned nothing from that young woman's tragic and preventable death." (Doc. 4122 at 68).

Defendants offered no contrary opinions or testimony. Defendants did not offer an expert stating isolated confinement creates no concerns. Nor did Defendants even proffer a penological purpose for keeping many prisoners in isolated conditions. Thus, for purposes of this case, it is undisputed isolated confinement is harmful for all subclass members. The conditions specific to this case, both the physical layout of the cells and the amount of out-of-cell time actually available to subclass members, present a substantial risk of serious harm to the mental health of every subclass member.

If ADCRR were following its own policies, including the amount of out-of-cell time promised and required by DO 812 and the chance to leave isolated confinement, the situation might be different. Dr. Haney's testimony cannot be read as prohibiting isolated confinement regardless of the circumstances. And the Court recognizes the need and value of isolated confinement in some circumstances. However, the actual practices imposed on the subclass and analyzed by Dr. Haney present a substantial risk of harm. That risk of harm is elevated for prisoners with a mental illness. And the risk of harm is extremely high for juveniles.

IV. Legal Standard for Subclass

The legal standard and inquiry applicable to subclass members is somewhat different from that applicable to the class. "The Eighth Amendment's prohibition against cruel and unusual punishment protects prisoners not only from inhumane methods of punishment but also from inhumane conditions of confinement." Morgan v. Morgensen , 465 F.3d 1041, 1045 (9th Cir. 2006). See also Farmer v. Brennan , 511 U.S. 825, 832, 114 S.Ct. 1970, 128 L.Ed.2d 811 (1994) (noting "[t]he Constitution does not mandate comfortable prisons, but neither does it permit inhumane ones"). "[W]hile conditions of confinement may be, and often are, restrictive and harsh, they ‘must not involve the wanton and unnecessary infliction of pain.’ " Morgan , 465 F.3d at 1045 (quoting Rhodes v. Chapman , 452 U.S. 337, 347, 101 S.Ct. 2392, 69 L.Ed.2d 59 (1981) ). Under this standard, the conditions of confinement "must not be devoid of legitimate penological purpose, or contrary to evolving standards of decency that mark the progress of a maturing society." Id. (citation and quotation marks omitted).

"Where the conditions of confinement are challenged rather than the confinement itself, a plaintiff must make two showings. First, the plaintiff must make an objective showing that the deprivation was sufficiently serious to form the basis for an Eighth Amendment violation. Second, the plaintiff must make a subjective showing that the prison official acted with a sufficiently culpable state of mind." Johnson v. Lewis , 217 F.3d 726, 731 (9th Cir. 2000).

A. Objective Prong

Restrictions or deprivations "inherent in the prison setting" cannot "satisfy the objective prong of an Eighth Amendment inquiry." Johnson , 217 F.3d at 731. Only restrictions or deprivations of "the minimal civilized measure of life's necessities are sufficiently grave to form the basis of an Eighth Amendment violation." Id. (quotation marks and citation omitted). The "life's necessities" include, among others, "adequate shelter, food, clothing, sanitation, medical care, and personal safety." Id. When a prisoner is denied a necessity, "[t]he circumstances, nature, and duration of [the] deprivation ... must be considered in determining whether a constitutional violation has occurred." Id. In general, "[t]he more basic the particular need, the shorter the time it can be withheld." Id. (quoting Hoptowit v. Ray , 682 F.2d 1237, 1259 (9th Cir. 1982), overruled on other grounds by Sandin v. Conner , 515 U.S. 472, 115 S.Ct. 2293, 132 L.Ed.2d 418 (1995) ). For example, forcing prisoners to lie "motionless in the dirt in subfreezing temperatures for five to nine hours" could be sufficient to establish a deprivation of the "basic human need[ ]" for shelter. Johnson , 217 F.3d at 732. But housing prisoners for "short periods of time" in a padded cell that was "dark, scary, and smelled bad" containing a "pit toilet encrusted with excrement and urine" will not qualify as deprivation of the basic human need for shelter or sanitation. Anderson v. Cty. of Kern , 45 F.3d 1310, 1313 (9th Cir. 1995). See also Hutto v. Finney , 437 U.S. 678, 686–87, 98 S.Ct. 2565, 57 L.Ed.2d 522 (1978) ("[T]he length of confinement cannot be ignored in deciding whether the confinement meets constitutional standards. A filthy, overcrowded cell and a diet of ‘grue’⁷¹ might be tolerable for a few days and intolerably cruel for weeks or months.").

71 "Grue" was described as "a substance created by mashing meat, potatoes, oleo, syrup, vegetables, eggs, and seasoning into a paste and baking the mixture in a pan." Hutto , 437 U.S. at 683, 98 S.Ct. 2565.

B. Subjective Prong

Conditions of confinement such as those challenged here require Plaintiffs establish the deprivations "occurred with deliberate indifference to [their] health or safety." Foster v. Runnels , 554 F.3d 807, 812 (9th Cir. 2009). This "involves a two part inquiry." The first part requires Plaintiffs show Defendants "were aware of a substantial risk of serious harm to [their] health or safety." Thomas v. Ponder , 611 F.3d 1144, 1150 (9th Cir. 2010). Whether Defendants "had the requisite knowledge of a substantial risk is a question of fact subject to demonstration in the usual ways, including inference from circumstantial evidence." Farmer , 511 U.S. at 842, 114 S.Ct. 1970. And Defendants’ knowledge of a substantial risk may be established "from the very fact that the risk was obvious." Id.

V. Conditions Imposed on the Subclass

In the class certification order, the Court identified the seven practices subclass members were allegedly being subjected to contrary to constitutional requirements. Those seven practices were:

(i) Inadequate psychiatric monitoring because of chronic understaffing;

(ii) Use of chemical agents against inmates on psychotropic medications;

(iii) Lack of recreation;

(iv) Extreme social isolation;

(v) Constant cell illumination;

(vi) Limited property; and

(vii) Insufficient nutrition.

When assessing whether Plaintiffs have carried their burden of proof regarding these practices, the Court cannot combine various practices to conclude the "overall conditions" violate the Eighth Amendment. Wilson v. Seiter , 501 U.S. 294, 305, 111 S.Ct. 2321, 115 L.Ed.2d 271 (1991). As explained by the Supreme Court, "[s]ome conditions of confinement may establish an Eighth Amendment violation in combination when each would not do so alone, but only when they have a mutually enforcing effect that produces the deprivation of a single, identifiable human need such as food, warmth, or exercise—for example, a low cell temperature at night combined with a failure to issue blankets." Id. at 304, 111 S.Ct. 2321. Thus, the Court can only combine various conditions when, together, they deprive prisoners of a "single, identifiable human need." Id. The challenged practices involve some well-established "human needs" such as the need for exercise, ability to sleep, and sufficient food. But the subclass also challenges practices involving matters that have not yet been widely recognized as "human needs," such as extreme limitations on social contact or limited property.

It is important to note Defendants admit keeping all subclass members in the restrictive environments at issue is not justified. There are approximately 200 subclass members who Defendants’ employees admitted at trial should no longer be held in maximum custody units. In other words, there simply is no penological justification for holding those subclass members under harsh and restrictive conditions. Defendants’ concession they house prisoners in these restrictive conditions without penological justification means Defendants should have remedied it long ago. See Wood v. Beauclair , 692 F.3d 1041, 1050 (9th Cir. 2012) (finding conduct "totally without penological justification" satisfies the objective prong of the Eighth Amendment analysis).

Further, policies allow prisoners to be placed in isolated confinement without penological justification. As recounted earlier, the policy requiring prisoners start serving life sentences in maximum custody is not supported by a valid penological interest. In addition, Defendants’ policies that some prisoners must remain in maximum custody because of misconduct years ago is not supported by a legitimate penological interest.

Despite the subclass including many prisoners who obviously should not be housed in isolated confinement, the Court will conduct the practice-by-practice analysis on the subclass as a whole. Thus, the Court will not analyze separately the practices and policies as applied to prisoners who everyone agree do not merit isolated confinement. The Court will assess whether Plaintiffs have carried their burden of proving certain policies or practices, as applied to members of the subclass Defendants believe should be in isolated confinement, violate the Eighth Amendment.

1. Defendants do not have a policy or practice of using chemical agents against inmates on psychotropic medications.

Neither at trial nor in their post-trial submissions did Plaintiffs identify the exact contours of the practice they were attempting to prove. While the class certification order described the practice as use of chemical agents against prisoners on particular medication, Plaintiffs’ response to Defendants’ proposed findings of fact describes the challenged practice as "widespread and indiscriminate use of force on people who engage in self-harm." (Doc. 4314 at 109). That was not the practice certified. Plaintiffs have not proven Defendants have a policy or practice of using chemical agents against inmates on psychotropic medications that violates the Eighth Amendment.

In reaching this conclusion, the Court notes Plaintiffs presented disturbing evidence of the use, and overuse, of chemical agents against prisoners with obvious mental health needs. For example, the February 12, 2020, email between medical providers quoted earlier involved I.C., a "mental health patient" at the Phoenix facility, who had been "self harming by banging his head for the past several days resulting in ... the use of OC [pepper] spray." (Doc. 4109 at 79). The "Regional Director of Mental Health" recommended officials "continue using OC spray as needed while the on site mental health team comes up with a treatment plan." (Doc. 4109 at 79). Using pepper spray in lieu of immediate psychiatric treatment was obviously inappropriate. And this was not the only evidence of such practices.

During trial, as discussed above, Plaintiffs presented multiple videos and other evidence that Rahim Muhammed was repeatedly subjected to OC spray when he was experiencing mental health crises. The evidence established prison officers did not know how to handle Muhammed and they were unable to convince him to stop his self-harming behaviors. Given Muhammed's behavior, it was obvious prison officials should have arranged for much more attention from mental health providers.⁷² Once Muhammed received additional mental health care, and was transferred to a more suitable location, he was able to function at a much higher level as shown by his trial testimony being coherent and convincing. The treatment of Muhammed might have violated the Eighth Amendment, but that treatment does not prove the existence of a general policy regarding using OC spray against prisoners on medication.

72 During trial, the Court asked the Deputy Warden about the failure to obtain additional mental health care for Muhammad. (Doc. 4279 33-36). The Deputy Warden claimed "mental health" personnel were involved with Muhammad "every single time" Muhammad was subjected to OC spray. (Doc. 4279 at 33). But there was no evidence establishing such interventions. The Deputy Warden stated Muhammad first had an "issue" approximately "three or four weeks" after he arrived. (Doc. 4279 at 36). After repeated instances of self-harm and OC spray, the Deputy Warden reported to the "mental health lead" that Muhammad needed "more care than what" was available at his location. (Doc. 4279 at 34). Muhammad was transferred "shortly thereafter." (Doc. 4279 at 34). Muhammad was at that location for approximately seven months. (Doc. 4279 at 37). Thus, the Deputy Warden knew of Muhammad's difficulties after approximately one month, but waited five or six months to speak to the "mental health lead" about arranging additional services.

The evidence regarding I.C. and Muhammed make it more likely than not that prison officers overuse chemical agents. Those chemical agents are used because adequate mental health care is not available. With appropriate mental health care, the need for chemical agents would be significantly reduced. Accordingly, remedying the patently insufficient mental health resources will reduce the need for interventions by officers involving chemical agents and likely provide the relief Plaintiffs sought regarding the use of chemical agents.

2. Subclass members do not receive adequate recreation and exercise opportunities.

The Ninth Circuit has recognized, repeatedly, "exercise is one of the basic human necessities protected by the Eighth Amendment." Norbert v. City & Cty. of San Francisco , 10 F.4th 918, 929 (9th Cir. 2021). The amount of exercise that must be offered to prisoners, as well as whether that exercise may be indoors or outside, "must be evaluated on the full extent of the available recreational opportunities." Id. at 930. In other words, "[d]etermining what constitutes adequate exercise requires consideration of the physical characteristics of the cell and [prison] and the average length of stay of the inmates." Pierce v. Cty. of Orange , 526 F.3d 1190, 1212 (9th Cir. 2008) (quotation marks and citation omitted). For example, in a case where prisoners were held in "continuous segregation, spending virtually 24 hours every day in their cells with only meager out-of-cell movements and corridor exercise," they were entitled to "one hour per day, five days a week" of outdoor exercise. Spain v. Procunier , 600 F.2d 189, 199 (9th Cir. 1979). By contrast, prisoners were not entitled to outdoor exercise when they were held in a far less restrictive setting that allowed out-of-cell "recreation time" of at least one hour every day. Norbert v. City and County of San Francisco , 10 F.4th 918, 934 (9th Cir. 2021).

In general, prisoners must be provided sufficient time and access to locations such that they enjoy "meaningful" recreation opportunities. Shorter v. Baca , 895 F.3d 1176, 1185 (9th Cir. 2018). The Ninth Circuit has noted offers for recreation in a room with "space constraints" that lacks "appropriate equipment" may not "constitute an exercise opportunity." Pierce v. Cty. of Orange , 526 F.3d 1190, 1212 n.22 (9th Cir. 2008).

The Court is not assessing the constitutionality of the amounts of recreation set forth in DO 812. If subclass members were receiving legitimate offers for recreation in the amount and locations set forth in DO 812, and prisoners were able to progress through the step program set forth in DO 812, it is unlikely prisoners would have a viable claim based on inadequate recreation. A minimum of 7.5 hours of recreation per week, with the promise that additional and better recreation offers could be earned, would present a very different situation from what is presented here. Instead, the actual practice is that DO 812 is not followed. Subclass members are routinely denied recreation of even 7.5 hours per week and, even when prisoners should be provided recreation in different enclosures, recreation is routinely offered in the chute which is only marginally better than their cells.

The tracking forms for individuals in maximum custody establish it is more probable than not the subclass members are routinely receiving legitimate offers for five hours or less of recreation per week. Even five hours is an overstatement based on the implausibly high refusal rates and the practice to interpret, whenever possible, responses to offers as denials. The detention forms establish it is more probable than not that prisoners in detention units routinely receive no recreation offers. While the forms sometimes indicate recreation offers, every single prisoner is often recorded as refusing recreation. That indicates it is more probable than not the offers for recreation, even if documented on the form, are overstating the offers actually made.

Beyond the insufficient amount of recreation time actually offered to subclass members, the offered location is insufficient under the particular facts of this case. Here, maximum custody units are "barren, stark, and ... dehumanizing." (Doc. 4122 at 61; 4306 at 42). When allegedly given the opportunity to leave their cells, the offer is merely to be placed in one of the chutes for a potentially extended period. The chutes are only marginally larger than the cells, contain no exercise equipment, and have no toilet. The chutes are open to the sky, meaning they are stifling in the summer and attract insects. And prisoners are taken to the chute one at a time, precluding any socialization. Accordingly, the chutes are a better location for prisoners to obtain some amount of exercise, but exercise in the chute is not sufficiently "meaningful" on a long-term basis.

Prisoners often spend years or decades in maximum custody. Prisoners may spend months in detention units. Given the length of time prisoners remain in isolated confinement, the physical conditions of the cells (i.e. , cell size and pest infestations), and the stark conditions of the chute, the amount of recreation offered to subclass members deprives them of the essential human need to exercise. See Pierce v. Cty. of Orange , 526 F.3d 1190, 1212 (9th Cir. 2008) (noting courts have held prisoners "are ordinarily entitled to daily exercise, or five to seven hours of exercise per week, outside their cells"). The objective prong of the Eighth Amendment test is met.

The subjective prong is easily met. It is undisputed the records for maximum custody as well as detention establish subclass members routinely receive few, if any, recreation offers. Even if accurate, the records show prisoners refusing recreation at implausibly high rates. Officials must have known prisoners were not being offered legitimate opportunities to exercise. And because it has been established for decades that prisoners in long-term housing must be provided some meaningful amount of recreation, the failure to provide that exercise must have been done with deliberate indifference to the subclass members’ rights. Shorter v. Baca , 895 F.3d 1176, 1185 (9th Cir. 2018). 3. Subclass members are subject to unconstitutional amounts of social isolation.

Plaintiffs claim that subclass members are subject to profound levels of social isolation. The evidence established prisoners are subject to "extreme social isolation" creating a substantial risk of serious harm. (Doc. 4165 at 93). No other conclusion is possible.

Subclass members may be celled individually or with one other person.⁷³ Beyond a cellmate, subclass members often have effectively zero social contact. For many, the only interactions that might be described as social contacts are when meals are delivered. (Doc. 4165 at 53). Beyond meal delivery, subclass members may remain in their cells with absolutely no interaction with another human for days at a time. Some additional interaction may occur on the days when subclass members are offered showers or recreation. But even if those offers are accepted, there likely will be no opportunity for meaningful social interaction.

73 As noted by Dr. Haney, having a cellmate "does not mitigate, and indeed may exacerbate, the psychological impact of" living in a maximum custody unit. (Doc. 4165 at 54). Living in such close proximity to a single other individual is not "meaningful social contact" and "may become an additional stressor" because those relationships may create "tension and even conflict" between prisoners. (Doc. 4165 at 54). "[C]onstant, forced, inescapable, and unremitting contact with another person in such a small and enclosed space" can become "intolerable." (Doc. 4165 at 55).

Individuals who accept an offer to shower or to go to recreation will have brief interactions with one officer while the prisoner is searched, handcuffed, and escorted to and from the shower or recreation enclosure.⁷⁴ Both showers and recreation, if offered in the chute, are done alone. The amount of social isolation that may be enforced on subclass members is illustrated by a Detention Record covering the week beginning July 11, 2021. (Doc. 1697 at 1277). During that seven-day period, prisoner M.H. received only seven meals. Thus, M.H. interacted with prison staff for a few seconds, seven times, when receiving his meals. That was the entirety of M.H.’s social contact that week. M.H. did not receive a single offer to shower nor did he receive a single offer for recreation. M.H. was kept in his cell for at least the seven straight days reflected in the record, receiving minimal food, and experiencing no contact with another person.

74 Some prisoners might require more than one officer be present during this process. But, at most, there will be three officers.

The record for M.H. is on the extreme end of the conditions experienced by subclass members. But all subclass members are subjected to profound deprivations of social interaction. Subclass members have interactions when they receive meals and when they receive offers for showers or recreation. But those interactions are with one or two staff members. And because recreation is often offered in the chute, no social interactions will occur during the recreation beyond the interaction with the staff member while being transported to and from the chute.

The deprivation of social contact imposed on subclass members places all subclass members "at significant risk of serious psychological harm." (Doc. 4165 at 50). Defendants did not present any evidence, expert or otherwise, questioning this conclusion. Presumably because it has been established for decades that profound levels of social isolation are harmful. As one court noted almost thirty-five years ago, it "seems pretty obvious ... that isolating a human being from other human beings year after year or even month after month can cause substantial psychological damage, even if the isolation is not total." Davenport v. DeRobertis , 844 F.2d 1310, 1313 (7th Cir. 1988). Another court noted in 1995 that "conditions of extreme social isolation and reduced environmental stimulation ... will likely inflict some degree of psychological trauma upon most inmates confined there for more than brief periods."⁷⁵ Madrid v. Gomez , 889 F. Supp. 1146, 1265 (N.D. Cal. 1995). And yet another court noted, in 2020, "the substantial risks of prolonged solitary confinement are obvious, longstanding, pervasive, well-documented, [and] expressly noted by prison officials in the past." Porter v. Pennsylvania Dep't of Corr. , 974 F.3d 431, 445 (3d Cir. 2020) (quotation marks and citation omitted). In reaching that conclusion, the Third Circuit quoted approvingly from an earlier case that had concluded prolonged deprivation of social contact creates a situation where a prisoner's "very identity is at risk of disintegration." Id. (quoting Williams v. Sec'y Pennsylvania Dep't of Corr. , 848 F.3d 549, 566 (3d Cir. 2017) ).

75 The Madrid court found the lack of social isolation and environmental stimulation were not "per se violative of the Eighth Amendment." Id. Instead, those conditions violated the Eighth Amendment rights of only those prisoners with existing mental illness or similar conditions. The conditions in Madrid , however, differed from those here. For example, prisoners in Madrid were entitled to "an exercise period five times each week" and were generally out of their cells for 90 minutes every day. 889 F. Supp. at 1156, 1230.

Social contact is a need protected by the Eighth Amendment. That does not mean the same amount of social contact is required for all prisoners or that, in some circumstances, it cannot be denied for periods of time. For purposes of the subclass, the only issue is whether all prisoners in isolated confinement can be deprived of effectively all social contact for an indeterminate amount of time, up to years and decades. Requiring subclass members go decades without meaningful social contact presents a substantial risk of serious psychological harm. The objective prong is satisfied.

The subjective prong is also satisfied given the decades-long recognition by academics, prison administrators, and courts that social isolation of the sort imposed on the subclass creates a significant risk of serious harm. The risk is so obvious that Defendants must have known their practices were creating a risk of harm.

4. Subclass members are not subjected to unconstitutional cell illumination practices.

"Adequate lighting is one of the fundamental attributes of ‘adequate shelter’ required by the Eighth Amendment." Keenan v. Hall , 83 F.3d 1083, 1090 (9th Cir. 1996) (quotation marks and citation omitted). Requiring prisoners live "in constant illumination" can be unconstitutional. Id. Existing caselaw links excessive lighting to sleep deprivation. See, e.g. , id. at 1091 (holding excessive cell lighting allegedly caused "grave sleeping problem"). The Court will assume Plaintiffs challenge the cell illumination practices as a deprivation of their right to adequate shelter or sleep.⁷⁶

76 Plaintiffs repeatedly reference that subclass members do not experience "natural light." (Doc. 4308 at 38). If Plaintiffs’ theory was the deprivation of natural light, that theory fails because Plaintiffs did not prove deprivation of "natural light" violates the objective prong of the Eighth Amendment.

Plaintiffs presented very little evidence regarding the cell illumination experienced by subclass members. Dr. Haney stated there was constant illumination in certain cells. (Doc. 4122 at 52). And one prisoner testified that, while on mental health watch, his cell was illuminated "24/7" that made him feel "insane." (Doc. 4263 at 89). But Dr. Haney noted prisoners often cover the lights in their cells, presumably reducing the amount of light to at least some degree. (Doc. 4122 at 52).

Plaintiffs presented no evidence regarding how bright the lighting in the cells actually is nor did they present evidence establishing subclass members, in general, experience negative effects such as sleeping problems because of the lighting. See Grenning v. Miller-Stout , 739 F.3d 1235, 1240 (9th Cir. 2014) (reversing grant of summary judgment because record did not indicate how bright the lighting in cells was). In addition, it is undisputed Defendants have a legitimate penological interest in lighting the subclass members’ cells so officers "can actually look in the cell and make sure that the inmate is alive and okay." (Doc. 4263 at 46).

Without evidence regarding how bright each cell is, the impact of lighting on subclass members, and the existence of a penological purpose for some amount of lighting, Plaintiffs have not carried their burden of providing the current cell illumination practices violate the objective prong of the Eighth Amendment. See Chappell v. Mandeville , 706 F.3d 1052, 1058 (9th Cir. 2013) (holding cell illumination claim failed because there was undisputed penological purpose).

5. Subclass members are not subjected to unconstitutional limitations on their property.

Plaintiffs did not present evidence regarding the exact limitations on property they are challenging. At trial, there was evidence that entitlement to property varies significantly across the subclass. For example, subclass members in detention, close management, and in an intake facility are permitted very little, if any, property. (Doc. 4264 at 139; Doc. 4165 at 52). Similarly, subclass members in a mental health watch unit are not entitled to any "personal property." (Doc. 4122 at 72). But there is no evidence regarding deprivation of property in the general maximum custody units. Plaintiffs have not established, on a subclass-wide basis, any practice or policy regarding property that violates a basic human need.

6. Subclass members are subjected to practices resulting in inadequate nutrition.

"Adequate food is a basic human need protected by the Eighth Amendment," but that "food need not be tasty or aesthetically pleasing." Keenan v. Hall , 83 F.3d 1083, 1091 (9th Cir. 1996) (quotation marks and citations omitted). In one case, the Ninth Circuit addressed an Eighth Amendment claim by an inmate who alleged he had been "denied 16 meals in 23 days." Foster v. Runnels , 554 F.3d 807, 812 (9th Cir. 2009). The Ninth Circuit concluded that, if true, that was a sufficiently serious deprivation of a life necessity. Id. at 813. Moreover, the knowing denial of food easily meets the subjective prong because "[t]he risk that an inmate might suffer harm as a result of the repeated denial of meals is obvious." Id. at 814.

As with other aspects of the subclass's claims, the issue is not whether Defendants’ policies require prisoners receive nutritionally adequate food. It is undisputed the menus and policies contemplate the delivery of meals that would be nutritionally adequate. (Ex. 4406). In other words, if Defendants implemented their own policies, there would likely be no plausible claim for relief. But the evidence establishes Defendants’ actual practices differ dramatically from its policies. Subclass members routinely receive less food than policy requires and less food than appropriate. Subclass members routinely report they are receiving insufficient food. The "top complaint" by prisoners regarding their conditions of confinement is that they do not receive the portions of food outlined in Defendants’ policies and menus. (Doc. 4278 at 36). In addition, the detention records show prisoners are routinely denied meals. Some of those records establish prisoners were denied far more than the "16 meals in 23 days" the Ninth Circuit found to be an obvious constitutional violation. Foster v. Runnels , 554 F.3d 807, 812 (9th Cir. 2009). For example, one prisoner received nine meals over a seven-day period. (Ex. 1697 at 1275). Another prisoner received only seven meals over a seven-day period. (Ex. 1697 at 1277).

Defendants claim the records are inaccurate. Defendants argue they are not correct because prisoners would "report if they were not being fed." (Doc. 4315 at 129). But they do and the "top complaint" from prisoners regarding their conditions of confinement was insufficient food. Next, Defendants argue Plaintiffs merely "cherry picked examples of deficient documentation" of meals. (Doc. 4315 at 130). According to Defendants, there are "thousands of pages" showing prisoners received their meals and only some records establish prisoners were fed the appropriate number of meals. (Doc. 4315 at 130). However, the records across locations proved subclass members, as a matter of practice, routinely do not receive adequate meals. At the very least, the records show every subclass member is subject to food practices that create a substantial risk of harm. And because the failure to deliver adequate food is so obvious in the records, Defendants are deliberately indifferent in denying prisoners a basic human need.

INJUNCTIVE RELIEF UNDER THE PRISON LITIGATION REFORM ACT

Because the Court has determined by a preponderance of the evidence that Defendants are deliberately indifferent to a substantial risk of serious harm, it must fashion an appropriate remedy. As noted at the beginning of its decision, the Court is mindful of its responsibilities in light of the PLRA's requirements, namely that "[t]he court shall not grant or approve any prospective relief unless the court finds that such relief is narrowly drawn, extends no further than necessary to correct the violation of the Federal right, and is the least intrusive means necessary to correct the violation of the Federal right." 18 U.S.C. § 3626(a)(1)(A).

In this case, there is no question remedial measures are necessary to correct constitutional deficiencies and the Court will meet its constitutional obligations. Thus, the Court will employ an expert to assist with crafting an injunction that remedies the constitutional violations—no more and no less.

IT IS THEREFORE ORDERED that the Court finds in favor of Plaintiffs and against Defendants as to their claims in Counts One, Count Two, Count Four, and Count Five. Count Three is dismissed.

IT IS FURTHER ORDERED that the Court finds in favor of Plaintiffs and against Defendants as to the following certified class practices:

• Failure to provide timely access to health care;

• Failure to provide timely emergency treatment;

• Failure to (timely) provide necessary medication;

• Insufficient health care staffing;

• Failure to provide care for chronic diseases;

• Failure to provide timely access to medically necessary specialty care;

• Failure to provide mentally ill prisoners medically necessary mental health treatment (i.e., psychotropic medication, therapy, and inpatient treatment); and

• Failure to provide suicidal and self-harming prisoners basic mental health care.

IT IS FURTHER ORDERED that the Court finds in favor of Plaintiffs and against Defendants as to the following certified subclass practices:

• Inadequate psychiatric monitoring because of chronic understaffing;

• Lack of recreation;

• Extreme social isolation; and

• Insufficient nutrition.

IT IS FURTHER ORDERED within 14 days of this Order the parties are invited to nominate proposed experts to assist the Court with crafting an injunction that complies with 18 U.S.C. § 3626(a)(1)(A). The Court will consider the proposals and will appoint a qualified expert on or before August 15, 2022 .

IT IS FURTHER ORDERED Plaintiffs’ Motion to Exclude Evidence Due to Rule 30(b)(6) Designee's Lack of Knowledge (Doc. 4163) is denied as moot ; Plaintiffs’ Motion in Limine Re: Admission of Exhibits (Doc. 4215) is denied except as explained herein; Defendants’ Motion to Strike Portions of Plaintiffs’ Experts’ Declarations (Doc. 4219) is denied ; the parties’ Joint Motion to Admit Exhibits (Doc. 4224) is granted ; and Plaintiffs’ Motion in Limine to Admit Exhibits 556 and 941 (Doc. 4225) is denied except as explained herein.

Appendix 1 Medical Staffing

Douglas Complex ⁷⁷ Population: 1,524 POSITION CONTRACT HIRED FTE FTE FTE VARIANCE Assistant Director of Nursing 1.00 1.00 Director of Nursing 1.00 1.00 Licensed Practical Nurse/Medical Assistant 4.00 4.00 Medical Assistant - 0.00 Medical Director 1.00 1.00 Midlevel Practitioner 1.50 2.00 0.50 Nursing Assistant/Patient Care Technician 4.00 4.00 Registered Nurse 8.00 7.30 (0.70) Staff Physician N/A N/A TOTAL 20.50 20.30 [REDACTED/] Eyman Complex Population: 5,219 POSITION CONTRACT HIRED FTE FTE FTE VARIANCE Assistant Director of Nursing 6.00 3.00 (3.00) Director of Nursing 1.00 1.00 Licensed Practical Nurse/Medical Assistant 30.00 19.70 (10.30) Medical Assistant - 3.00 3.00 Medical Director 1.00 1.00 Midlevel Practitioner 5.50 5.00 (0.50) Nursing Assistant/Patient Care Technician 9.00 9.35 0.35 Registered Nurse 20.00 9.90 (10.10) Staff Physician 1.00 1.00 TOTAL 73.50 52.95 [REDACTED/] Florence Complex Population: 2,534 POSITION CONTRACT HIRED FTE FTE FTE VARIANCE Assistant Director of Nursing 6.00 4.00 (2.00) Director of Nursing 1.00 0.00 (1.00) Licensed Practical Nurse/Medical Assistant 30.00 15.90 (14.10) Medical Assistant - 0.00 Medical Director 1.00 1.00 Midlevel Practitioner 6.00 6.75 0.75 Nursing Assistant/Patient Care Technician 20.00 14.90 (5.10) Registered Nurse 36.00 20.25 (15.75) Staff Physician 2.00 1.25 (0.75) TOTAL 102.00 64.05 [REDACTED/] Lewis Complex Population: 4,220 POSITION CONTRACT HIRED FTE FTE FTE VARIANCE Assistant Director of Nursing 6.00 2.00 (4.00) Director of Nursing 1.00 1.00 Licensed Practical Nurse/Medical Assistant 34.00 17.30 (16.70) Medical Assistant - 0.90 0.90 Medical Director 1.00 1.00 Midlevel Practitioner 6.00 5.50 (0.50) Nursing Assistant/Patient Care Technician 14.00 12.70 (1.30) Registered Nurse 30.00 24.35 (5.65) Staff Physician 2.00 0.00 (2.00) TOTAL 94.00 64.75 [REDACTED/] Perryville Complex Population: 3,318 POSITION CONTRACT HIRED FTE FTE FTE VARIANCE Assistant Director of Nursing 6.00 2.00 (4.00) Director of Nursing 1.00 1.00 Licensed Practical Nurse/Medical Assistant 24.00 18.50 (5.50) Medical Assistant - 0.00 Medical Director 0.80 1.00 0.20 Midlevel Practitioner 5.00 6.00 1.00 Nursing Assistant/Patient Care Technician 14.00 12.60 (1.40) Registered Nurse 30.00 15.10 (14.90) Staff Physician 1.20 1.50 0.30 TOTAL 82.00 57.70 [REDACTED/] Phoenix Complex Population: 468 POSITION CONTRACT HIRED FTE FTE FTE VARIANCE Assistant Director of Nursing 3.00 0.00 (3.00) Director of Nursing 1.00 1.00 Licensed Practical Nurse/Medical Assistant 3.00 4.80 1.80 Medical Assistant - 0.00 Medical Director 1.00 1.00 Midlevel Practitioner 4.00 5.00 1.00 Nursing Assistant/Patient Care Technician 5.75 1.90 (3.85) Registered Nurse 12.00 9.65 (2.35) Staff Physician 1.00 0.00 (1.00) TOTAL 30.75 23.35 [REDACTED/] Safford Complex Population: 1,035 POSITION CONTRACT HIRED FTE FTE FTE VARIANCE Assistant Director of Nursing 2.00 2.00 Director of Nursing 1.00 1.00 Licensed Practical Nurse/Medical Assistant 6.00 2.80 (3.20) Medical Assistant - 0.00 Medical Director 1.00 1.00 Midlevel Practitioner 1.00 2.00 1.00 Nursing Assistant/Patient Care Technician 4.00 4.00 Registered Nurse 8.00 12.80 4.80 Staff Physician N/A N/A TOTAL 23.00 25.60 2.60 Tucson Complex Population: 4,420 POSITION CONTRACT HIRED FTE FTE FTE VARIANCE Assistant Director of Nursing 8.00 4.00 (4.00) Director of Nursing 1.00 1.00 Licensed Practical Nurse/Medical Assistant 40.00 24.30 (15.70) Medical Assistant - 0.00 Medical Director 1.00 1.00 Midlevel Practitioner 8.00 7.00 (1.00) Nursing Assistant/Patient Care Technician 19.00 16.80 (2.20) Registered Nurse 36.00 18.30 (17.70) Staff Physician 2.00 0.75 (1.25) TOTAL 115.00 73.15 [REDACTED/] Winslow Complex Population: 1,481 POSITION CONTRACT HIRED FTE FTE FTE VARIANCE Assistant Director of Nursing 2.00 1.00 (1.00) Director of Nursing 1.00 1.00 Licensed Practical Nurse/Medical Assistant 4.00 1.90 (2.10) Medical Assistant - 1.00 1.00 Medical Director 1.00 1.00 Midlevel Practitioner 2.00 2.00 Nursing Assistant/Patient Care Technician 3.00 5.60 2.60 Registered Nurse 6.00 11.10 5.10 Staff Physician N/A N/A TOTAL 19.00 24.60 5.40 Yuma Complex Population: 3,374 POSITION CONTRACT HIRED FTE FTE FTE VARIANCE Assistant Director of Nursing 5.00 4.80 (0.20) Director of Nursing 1.00 1.00 Licensed Practical Nurse/Medical Assistant 10.00 12.00 2.00 Medical Assistant - 2.00 2.00 Medical Director 1.00 1.00 Midlevel Practitioner 4.00 5.90 1.90 Nursing Assistant/Patient Care Technician 6.00 14.75 8.75 Registered Nurse 14.00 16.55 2.55 Staff Physician 1.00 1.00 TOTAL 42.00 59.00 17.00

[Editor's Note: The preceding image contains the reference for footnote⁷⁷ ]. Appendix 2

77 Does not include Administrative Assistant, Assistant Facility Health Administrator, Clinical Coordinator, Facility Health Administrator, Healthcare Delivery Facilitator, Inventory Coordinator, Lab Technician, Lead Inventory Coordinator, Medical Records Clerk and Supervisor, Physical Therapist, Release/Discharge Planner, Scheduler, X-Ray Technician, Mental Health care staff (addressed in Part Two), or Dental staff. "N/A" means the position does not exist at the facility. Medical Assistant, Education Coordinator, Physical Therapist, Clinical Coordinator, and Recreation Therapist positions exist in excess of the staffing allocation, even though they may not be filled.

Dr. Murray's Scale

A rating of 5—Excellent was assigned to care that was timely and reflected good decision-making. The care may not have been perfect in all instances, but the imperfections found did not affect the outcomes. This rating may allow for a lack of preventive health care. Regarding documentation, this rating was used if the notes were clear and comprehensive (usually free-typed instead of utilizing the template).

A rating of 4—Very Good may represent care that was mostly solid, but there may be some important care that was delayed slightly or a decision that should have been made a bit earlier eventually got made and there was no harm to the patient by the delay.

A rating of 3—Good may represent a mix of care that was sometimes poor and sometimes great. However, if there was a problem with care that markedly increased risk to a patient, a lower score was assigned, since it does not help to provide excellent care for some problems and to not address a problem that had a significant risk of causing an adverse outcome.

A rating of 2—Fair would be assigned if the decision-making represented a lack of basic practice rules or put the patient at some risk. It may represent good decision-making regarding referrals, with delay in the referral. Often this rating reflected that there was some good care, but it was overshadowed by the presence of care that put the patient at risk and could not be overlooked. Using this rating for documentation would represent incomplete notes where the reader cannot follow the care very easily at all.

A rating of 1—Poor represents care that cannot be excused or overlooked. It represents risk to patients, lack of basic knowledge or follow through, or a timeline that does not realistically apply to the patient's problem. In documentation, this rating might be used when the patient's care was fragmented, to a point where one cannot follow what should happen. This may be due to the shortcomings of eOMIS or may be due to a lack of comprehensive documentation by the provider.

(Doc. 4206 ¶¶ 209-213).

Appendix 3 Mental Health Staffing

Douglas Complex ⁷⁸ POSITION CONTRACT HIRED FTE FTE FTE VARIANCE Behavioral Health Technician N/A N/A Behavioral Specialist N/A N/A Mental Health Lead N/A N/A Mental Health Clerk N/A N/A Mental Health Midlevel (NP/PA) N/A N/A Mental Health Registered Nurse N/A N/A Psychiatrist N/A N/A Psychologist N/A N/A Psychology Associate 1.00 0.00 (1.00) TOTAL 1.00 0.00 [REDACTED/] Eyman Complex POSITION CONTRACT HIRED FTE FTE FTE VARIANCE Behavioral Health Technician 4.00 7.00 3.00 Behavioral Specialist N/A N/A Mental Health Lead 1.00 1.00 Mental Health Clerk 1.00 0.00 (1.00) Mental Health Midlevel (NP/PA) 3.50 4.00 0.50 Mental Health Registered Nurse 2.00 0.90 (1.10) Psychiatrist 1.00 1.00 Psychologist 3.00 2.00 (1.00) Psychology Associate 13.00 8.00 (5.00) TOTAL 28.5 23.9 [REDACTED/] Florence Complex POSITION CONTRACT HIRED FTE FTE FTE VARIANCE Behavioral Health Technician 4.00 0.00 (4.00) Behavioral Specialist 0.00 0.00 Mental Health Lead 1.00 0.00 (1.00) Mental Health Clerk 1.00 1.00 Mental Health Midlevel (NP/PA) 3.50 3.00 (0.50) Mental Health Registered Nurse 1.00 0.00 (1.00) Psychiatrist 1.00 1.00 Psychologist 3.00 0.90 (2.10) Psychology Associate 8.00 2.00 (6.00) TOTAL 22.50 7.90 [REDACTED/] Lewis Complex POSITION CONTRACT HIRED FTE FTE FTE VARIANCE Behavioral Health Technician 4.00 3.00 (1.00) Behavioral Specialist N/A N/A Mental Health Lead 1.00 0.00 (1.00) Mental Health Clerk 1.00 1.00 Mental Health Midlevel (NP/PA) 3.50 3.00 (0.50) Mental Health Registered Nurse 2.00 1.00 (1.00) Psychiatrist 1.00 1.00 Psychologist 3.00 2.00 (1.00) Psychology Associate 12.00 10.25 (1.75) TOTAL 27.50 21.25 [REDACTED/] Perryville Complex POSITION CONTRACT HIRED FTE FTE FTE VARIANCE Behavioral Health Technician 3.00 2.00 (1.00) Behavioral Specialist N/A N/A Mental Health Lead 1.00 1.00 Mental Health Clerk 1.00 0.00 (1.00) Mental Health Midlevel (NP/PA) 3.50 3.20 (0.30) Mental Health Registered Nurse 5.20 4.70 (0.50) Psychiatrist 1.00 1.00 Psychologist 2.00 1.75 (0.25) Psychology Associate 10.00 9.00 (1.00) TOTAL 26.70 22.65 [REDACTED/] Phoenix Complex POSITION CONTRACT HIRED FTE FTE FTE VARIANCE Behavioral Health Technician 5.00 2.00 (3.00) Behavioral Specialist N/A N/A Mental Health Lead N/A N/A Mental Health Clerk N/A N/A Mental Health Midlevel (NP/PA) 3.50 3.50 Mental Health Registered Nurse 15.80 8.80 (7.00) Mental Health Registered Nurse Charge 1.00 0.90 (0.10) Psychiatrist 1.00 1.00 Psychologist 4.00 2.50 (1.50) Psychology Associate 11.00 9.00 (2.00) TOTAL 41.30 27.70 [REDACTED/] Safford Complex POSITION CONTRACT HIRED FTE FTE FTE VARIANCE Behavioral Health Technician N/A N/A Behavioral Specialist N/A N/A Mental Health Lead N/A N/A Mental Health Clerk N/A N/A Mental Health Midlevel (NP/PA) N/A N/A Mental Health Registered Nurse N/A N/A Psychiatrist N/A N/A Psychologist N/A N/A Psychology Associate 1.00 1.00 TOTAL 1.00 1.00 0.00 Tucson Complex POSITION CONTRACT HIRED FTE FTE FTE VARIANCE Behavioral Health Technician 6.00 4.00 (2.00) Behavioral Specialist 0.00 0.00 Mental Health Lead 1.00 1.00 Mental Health Clerk 1.00 1.00 Mental Health Midlevel (NP/PA) 3.50 4.50 1.00 Mental Health Registered Nurse 2.00 2.00 Psychiatrist 1.00 0.00 (1.00) Psychologist 4.00 3.90 (0.10) Psychology Associate 14.00 11.00 (3.00) TOTAL 32.50 27.40 [REDACTED/] Winslow Complex POSITION CONTRACT HIRED FTE FTE FTE VARIANCE Behavioral Health Technician N/A N/A Behavioral Specialist N/A N/A Mental Health Lead N/A N/A Mental Health Clerk N/A N/A Mental Health Midlevel (NP/PA) N/A N/A Mental Health Registered Nurse N/A N/A Psychiatrist N/A N/A Psychologist N/A N/A Psychology Associate 1.00 1.00 TOTAL 1.00 1.00 0.00 Yuma Complex POSITION CONTRACT HIRED FTE FTE FTE VARIANCE Behavioral Health Technician 3.00 3.00 Behavioral Specialist N/A N/A Mental Health Lead 1.00 1.00 Mental Health Clerk 1.00 1.00 Mental Health Midlevel (NP/PA) 3.00 3.00 Mental Health Registered Nurse 1.00 1.00 Psychiatrist 1.00 1.00 Psychologist 1.00 0.00 (1.00) Psychology Associate 9.00 6.75 (2.25) TOTAL 17.00 16.75 [REDACTED/]

[Editor's Note: The preceding image contains the reference for footnote⁷⁸ ].

78 Does not include Administrative staff, Health Care staff, or Dental staff. "N/A" means the position does not exist at the facility. Behavioral Specialist positions exist in excess of the staffing allocation, even though they may not be filled.

Appendix 4

Pennington-Stallcup's Email Eyman Complex POSITION CONTRACT HIRED FTE FTE FTE VARIANCE Psychology Associate 13.00 7.5 [REDACTED/] Florence Complex POSITION CONTRACT HIRED FTE FTE FTE VARIANCE Psychologist 3.00 0.90 [REDACTED/] Psychology Associate 8.00 3.00 Lewis Complex POSITION CONTRACT HIRED FTE FTE FTE VARIANCE Psychologist 3.00 2.00 [REDACTED/] Psychology Associate 12.00 9.0 Perryville Complex POSITION CONTRACT HIRED FTE FTE FTE VARIANCE Psychologist 2.00 1.75 [REDACTED/] Psychology Associate 10.00 8.00 Phoenix Complex POSITION CONTRACT HIRED FTE FTE FTE VARIANCE Psychologist 4.00 1.50 [REDACTED/] Psychology Associate 11.00 10.00 Tucson Complex POSITION CONTRACT HIRED FTE FTE FTE VARIANCE Psychologist 4.00 2.0 [REDACTED/] Psychology Associate 14.00 11.00 Yuma Complex POSITION CONTRACT HIRED FTE FTE FTE VARIANCE Psychologist 1.00 0.00 [REDACTED/] Psychology Associate 9.00 7.50 (Ex. 2141).

Appendix 5

Dr. Penn's Chart Reviews

Patient Access to Notes Trial Tr. No. Care? Page 1 No 3102 6 No 3102 11 No 3103 "death by probable suicide, possible problems with medical access to care'";⁷⁹ 12 Yes patient submitted HSRs on 3103 9/2 and 9/12 but was not seen until 9/16 patient died by hanging while in segregation; "many late entries but seen 13 Yes regularly by mental health 3103 for ... health and welfare rounds"; "possible medical access to care problem" 14 No 3104 16 No 3104 17 No 3104 21/22⁸⁰ No 3104 27 No 3104 28 No 3105 29 No 3105 30 No 3105 34 No 3105 38 No 3106 49 No 3106 died by suicide by hanging; "ambulance team refused to go to the patient's location 52 Yes due to their policy and he 3106-07 was brought to medical on a gurney" 55 No 3106 56 No 3107-08

61 No 3108 62 No 3108 72 No 3108 80 No 3108 81 No 3109-10 82 No 3109-10 "two weeks on [suicide] 86 Yes watch but didn't see 3110 psychiatric clinician" 87 No 3109-10 90 No 3110 92 No 3110 patient waited almost three 95 No weeks to have psychotic 3111 complaints addressed 97 No 3110 "`on [suicide] watch 6-21-19 through 7-8-19 but saw no psychiatric provider 3111, 98 Yes under 7-18-19'"; patient 3114 was on olanzapine, an anti-psychotic medication 100 No 3111 "Depakote prescribed at too 114 Yes low a dose as revealed by 3116 disturbed behavior in lab." patient committed suicide on a date early in 2021; 115 Yes "[patient] might have 3116 benefited from a prison inpatient unit" 118 No 3116 patient had died by suicide before the relevant time for 120 Yes this review; "answered yes 3117 only so I could submit the review" "lithium continued at low 600 milligram daily dose 129 Yes despite low level of .3 and a 3122-23 suicide attempt on this dose"

"health care request written in Spanish, yet most mental 131 Yes health meetings say no 3123 interpreter was used and do not state whether interview was conducted in Spanish" "has inexplicably been 136 Yes prescribed clonidine for 3123 depression[/]anxiety" 140 No 3122 "low dose of Risperdal consta 12.5 mg [every two] 145 Yes weeks despite chronic 3126 auditory hallucinations" "patient was taken off of Risperdal but had to be returned due to resurgent 146 Yes psychosis"; "decision to 3126 stop Risperdal was probably a mistake, but the reasoning was documented" 147 No 3122 148 No 151 No 3128 "low dose haloperidol, 1 milligram [twice daily] by psych APN is inadequate as evidenced by symptoms. 152 Yes May have catatonia, which 3128 would benefit from benzo or [electro-convulsive therapy]." 153 No 3128 "lithium was increased prescribed once daily at night, which is not 155 Yes optimal"; diagnosis of 3129 bipolar I was not well substantiated 156 Yes "marginally adequate care" 3129 158 No 3130 159 Yes "marginal adequate [sic] 3130

access to care" "concerns about 160 Yes polypharmacy and 3130 combining Cogentin with Zyprexa" "access to care adequate but 167 Yes concerns about medication 3132 decisions" "access to care adequate but 168 Yes concerns about decision-making" 3132 169 No 3135 AIMS examination (a physical examination for possible neurological side 179 Yes effects of psychotropic 3135 medications) is performed only once a year 228 No patient died by suicide 3137 "It is concerning that patient is not on meds and carries a diagnosis of schizophrenia or bipolar DO"; patient has been on 235 Yes involuntary medication in 3137 the past; "unmedicated, [patient] is likely going to have difficulties functioning" 252 Yes "documentation is poor" 3141-42 258 No 3142 259 No 3142 "admission in 2006 but encounters only go back to 271 Yes January 2020, limited 3143 documentation available" "no scheduled visit with the 272 Yes psychiatric midlevel since 3143 August 2020"

[Editor's Note: The preceding image contains the reference for footnotes^{79 , 80} ].

79 Direct citations to Exhibit 3 to the Penn Report appear in quotation marks.

80 "Patient 21" and "Patient 22" referred to the same individual.

Appendix 6

Recreation per Step/Location Step I Step II Step III Browning Three, 2.5-hour Three, 2.5-hour For GP — One, 2.5-hour blocks per week blocks per week, block per month in 20×40 in the standard⁸¹ one of which can basketball enclosure (up to enclosure be in the 10×10 eight inmates) enclosure For GP and STG — Three, 2.5-hour blocks per week, to include one-time per week, in the 10×10 enclosures For Condemned Row — Four, 2.5-hour blocks per week of outdoor recreation — All can be in the 10×10 interactive enclosures For all inmates: In pod recreation;⁸² Recreation with another inmate in accordance with Department Order #704, Inmate Regulations SMU I/Rast Three, 2.5-hour Three, 2.5-hour Three, 2.5-hour blocks per blocks per week blocks per week week to include one-time per to include 10×10 10×10 enclosure month in 10×10 enclosure enclosure (two (two inmates), (two inmates), inmates) and; and; One, 2.5-hour One, 2.5-hour block per block per month in month in 20×40 basketball 20×40 basketball enclosure (up to eight enclosure (up to inmates), and; eight inmates) One, 2.5-hour block per month in 50×90 (up to 32 inmates)

[Editor's Note: The preceding image contains the reference for footnotes^{81 , 82} ].

81 The "standard enclosure" is also referred to as the "chute." (Doc. 4259 at 104).

82 "In pod recreation" refers to time when prisoners may be out of their cells in their pods without restraints. (Doc. 4259 at 105-106).

Appendix 7 Maximum Custody Population

M/H Close Location Max Detention Watch Management Douglas Complex Detention 31 Eyman Rynning Detention 25 Rynning Close Mgt. 15 SMU I 350 SMU I SO 130 SMU I P.C. 16 SMU I Detention 153 SMU I M/H Watch 13 Browning Intake 31 Browning Unit 435 Browning STG 144 Browning D/Row 15 Browning M/H Watch 3 Browning BMU 11 Browning Enhanced 31 Browning RSHP 6 Browning Close 0 Management Florence Housing Unit 8 36 Health Unit 12 Globe Detention 0 Perryville Perryville Watch Cells 7 Santa Maria Detention 9 Phoenix Reception 212 B-Ward 32 Flamenco Ida Watch M 0 Lewis Morey Detention 73 Rast PC 311

Rast Close 7 Management Stiner Detention 64 Bachman Detention 67 Safford Miles Detention 18 Tonto Detention 0 Tucson Cimarron Detention 77 Rincon MH Watch 27 Rincon MH Program II 77 Rincon MH Watch II 0 Manzanita Detention 3 Winchester Detention 22 Complex Detention 69 Winslow Complex Detention 23 Apache Detention 0 Yuma Cheyenne Detention 52 Dakota Detention 60 TOTALS 1849 746 50 22