Hoffer v. Sec'y, Fla. Dep't of Corr.

NEWSOM, Circuit Judge

[PUBLISH] D.C. Docket No. 4:17-cv-00214-MW-CAS Appeal from the United States District Court for the Northern District of Florida Before MARTIN, NEWSOM, and BALDOCK, Circuit Judges. NEWSOM, Circuit Judge:

Honorable Bobby R. Baldock, United States Circuit Judge for the Tenth Circuit, sitting by designation.

This case principally presents the question whether the Eighth Amendment requires Florida prison officials to treat all inmates with chronic Hepatitis C—including those who have only mild (or no) liver fibrosis—with expensive, state-of-the-art "direct acting antiviral" (DAA) drugs. The district court held that it does and entered a permanent injunction mandating across-the-board DAA treatment. We hold, to the contrary, that the officials' current treatment plan—pursuant to which they monitor all HCV-positive inmates, including those who have yet to exhibit serious symptoms, and provide DAAs to anyone who has an exacerbating condition, shows signs of rapid progression, or develops even moderate fibrosis—satisfies constitutional requirements. Accordingly, we reverse the district court's decision, vacate its injunction, and remand for further proceedings.

I A

Hepatitis C—here, HCV—is a bloodborne virus that is commonly spread, among other means, by sharing contaminated needles, utilizing unsterilized tattoo equipment, and engaging in risky sexual behavior. Only about 1% of the general population suffers from the disease, but its prevalence among prison inmates is much higher. Although estimates vary, it's safe to say that thousands—and quite possibly tens of thousands—of inmates confined in Florida state prisons have HCV.

HCV primarily attacks the liver and, in particular, can cause liver scarring, or "fibrosis." Liver fibrosis can be measured, or staged, on a five-step scale, in ascending order of severity, from F0 (no fibrosis) to F1 (mild fibrosis) to F2 (moderate fibrosis) to F3 (severe fibrosis) to F4 (cirrhosis). Fortunately, many people afflicted with HCV will "spontaneously clear" the virus without treatment. At least 50% of HCV cases, though, are "chronic," meaning that they can be cured only with medication. Among chronic HCV patients, the disease's rate of progression varies: 30% are stable, meaning that they aren't currently moving up the fibrosis scale; 40% progress slowly, taking several years to advance from one level to another and more than 20 years to reach full-blown cirrhosis; and 30% progress rapidly, reaching cirrhosis in fewer than 20 years and possibly as few as one.

In years past, HCV patients were prescribed Interferon, a weekly injectable medication that had a number of drawbacks—among them that it required patients to remain sober, caused several unpleasant side effects, and succeeded in eradicating the virus only about 30% of the time. In 2013, a new HCV treatment option arrived on the scene—direct acting antivirals. These DAAs brought great promise—the once- or twice-daily pills were easily administrable, had few side effects, and boasted a 95% cure rate. Unfortunately, and not surprisingly, they also came at great cost. Although the parties dispute the exact price tag of a single 12-week course of DAA treatment—in 2017, the Secretary put it between $25,000 and $37,000 per inmate, while the plaintiffs insisted that discounts and rebates reduced that cost—all agree that DAAs are very expensive.

B

In May 2017, Carl Hoffer, Ronald McPherson, and Roland Molina—chronic-HCV inmates incarcerated in Florida prisons—sued the Secretary of the Florida Department of Corrections on behalf of a putative class in the United States District Court for the Northern District of Florida. Among other claims, the plaintiffs alleged, pursuant to 42 U.S.C. § 1983, that the Secretary's HCV treatment plan—or, at the time, the lack thereof—was deliberately indifferent to inmates' serious medical needs in violation of the Eighth Amendment. As particularly relevant here, the plaintiffs sought a class-wide injunction requiring the FDC to "develop and adhere to a plan to provide direct-acting antiviral medication to all FDC prisoners with chronic HCV, consistent with the standard of care."

Not long after the plaintiffs filed, the Secretary hired Dr. Daniel Dewsnup to formulate a treatment plan for HCV-infected inmates. Dr. Dewsnup, who had earlier developed and implemented a similar HCV-treatment plan for the Oregon prison system, recommended treating some inmates with DAAs, but not all of them. In particular, he proposed (1) providing DAAs for all inmates at level F2 and above and (2) monitoring F0- and F1-level inmates and treating them with DAAs if they (a) have or develop an exacerbating condition like HIV, (b) exhibit signs of rapid fibrosis progression, or (c) advance to F2. The Secretary adopted Dr. Dewsnup's recommendations.

In October 2017, the district court held a five-day evidentiary hearing, which featured Dr. Dewsnup and the plaintiffs' expert, Dr. Margaret Koziel. The experts testified concerning a range of topics, but for present purposes our focus is on their dueling conclusions regarding the necessity of treating HCV-positive inmates who show little or no liver scarring—i.e., those at levels F0 and F1 on the fibrosis scale.

Dr. Koziel "advocate[d]" in favor of treating all HCV-positive inmates with DAAs—even, she said, F0-level inmates, who have "no fibrosis." Her view, she said, comports with guidelines published by the American Association for the Study of Liver Diseases and the Infectious Diseases Society of America, which recommend DAA treatment for all chronic HCV patients, regardless of staging. Dr. Koziel further testified that "there are clearcut economic benefits to treating all-comers with Hepatitis C"—in particular, she reported that one paper she had recently reviewed estimated that "the U.S. health system would save about $13 billion by treating everybody."

Regarding medical outcomes, though, Dr. Koziel was more circumspect. As particularly relevant here, the district court judge asked her the following question: "[I]n terms of scoring, F0 to F4, . . . what's the tipping point where suddenly [a patient's] chance of getting cancer is dramatically much greater [or his] chance of permanent damage in terms of liver function . . . is greater?" Dr. Koziel began her response by stating that "there is some increased risk of mortality even in the earlier stages" and emphasizing that she "will always advocate for people." "But," she went on, "when I really, really start to get agitated . . . is about at F3, and that's when people really get into the irreversible complications because [of] the amount of scar tissue, and at that point I can no longer say to you that we are going to make you completely whole again." Even so, Dr. Koziel reiterated later in her testimony that her job as a physician "is to advocate for the best medical care" and agreed with the district court that "treating everyone who has [HCV] with DAA medication is the best possible management."

For his part, Dr. Dewsnup opined that F0- and F1-level inmates "don't need to be treated immediately" but, rather, "can be deferred." Usually, he said, the FDC only "see[s] them for a few years." At the same time, he recognized that it's different for "lifetime prisoners," whom the FDC is "going to have to treat eventually even if they are Stage 0 or Stage 1." Dr. Dewsnup also explained that the evidence concerning the relationship between the eradication of HCV and F0- and F1-level inmates' "mortality rates" may suffer from confounding; he acknowledged Interferon-era studies that suggest a correlation between the two, but clarified that "[w]e don't have the scientific data that say[s] that it's the viral eradication that does it." The reason, he said, is because Interferon treatment required patients "to make huge lifestyle changes"—in particular, "[t]hey had to be sober." Because "substance abuse . . . [is] the major reason" that HCV-positive inmates die, Dr. Dewsnup didn't believe that the correlation between Interferon treatment for F0s and F1s and decreased mortality rates established the necessary causal link.

Shortly after the hearing, the district court certified a plaintiff class of "all current and future prisoners in FDC custody who have been, or will be, diagnosed with chronic HCV" and entered a preliminary injunction. "[W]ith limited exceptions," the court ordered the FDC to "compl[y] with its own expert's recommendations" and to "formulate a plan" to ensure the treatment of HCV-positive inmates in accordance with a specified schedule. Later, having received the Secretary's plan, the court ordered DAA treatment for prisoners with decompensated cirrhosis (i.e., liver failure) within about a month, prisoners with F4-level cirrhosis within six months, and prisoners with F2 (moderate) and F3 (severe) fibrosis within about a year. The court's preliminary-injunction order did not require DAA treatment for F0- and F1-level inmates unless they had an exacerbating condition.

Several months later, the Secretary moved for summary judgment—in essence, against himself—seeking to make the district court's preliminary injunction permanent. As particularly relevant here, the Secretary reiterated his position that the Eighth Amendment doesn't require immediate DAA treatment of all F0- and F1-level inmates. As a general matter, the Secretary proposed to monitor F0s and F1s, to schedule infirmary visits and lab tests every six months, and, if necessary, to re-stage their fibrosis annually. F0s and F1s would be put on DAAs, however, if they (1) had or developed an exacerbating condition, (2) showed signs of rapid progression, or (3) advanced to F2. The plaintiffs cross-moved for summary judgment, seeking 15 additional forms of relief, all of which the Secretary opposed. Most significantly for present purposes, the plaintiffs asked the court to require the Secretary to provide DAA treatment for all F0- and F1-level patients, regardless of underlying comorbidities or the pace of disease progression.

The district court granted the plaintiffs' summary-judgment motion in substantial part, denying or altering only a few of their additional requests, and entered a permanent injunction. See Hoffer v. Inch, 382 F. Supp. 3d 1288 (N.D. Fla. 2019). As relevant here, the injunction requires the Secretary to treat all F0- and F1-level inmates with DAAs. In its order, the district court found that the Secretary "ha[d] not put forth any medical reason (nor does the record otherwise reveal a medical reason) why F0 and F1 inmates should not be treated." Id. at 1302. Instead, the court agreed with the plaintiffs that "[t]he only reason why [the Secretary] is electing not to provide treatment" to F0 and F1 patients "is due to the cost of treatment, which is per se deliberate indifference." Id. (first alteration in original) (quotation omitted). Accordingly, the court ordered the Secretary to begin DAA treatment of all F0s and F1s within two years of their initial staging. Id. at 1302-03.

Toward the end of its order, the district court included the following one-sentence paragraph, addressing its obligation to ensure that the injunction complied with the Prison Litigation Reform Act:

Finally, the permanent injunction satisfies the Prison Litigation Reform Act ("PLRA") because it is narrowly drawn, extends no further than necessary to effect the changes this Court concludes are constitutionally required, and is the least intrusive means of effecting such changes.

Id. at 1315 (citing 18 U.S.C. § 3626(a)(1)).

* * *

The Secretary now appeals the district court's summary-judgment order and accompanying injunction, arguing (1) that his plan for monitoring F0- and F1-level inmates, and treating them with DAAs if they develop an exacerbating condition, exhibit rapid disease progression, or advance to F2, does not violate the Eighth Amendment, and (2) that the district court's PLRA-related findings were not sufficiently particularized.

We review the district court's decision to grant summary judgment de novo, Ellis v. England, 432 F.3d 1321, 1325 (11th Cir. 2005), and its decision to enter an injunction for abuse of discretion, Prison Legal News v. Sec'y, Fla. Dep't of Corr., 890 F.3d 954, 964 (11th Cir. 2018). "A district court by definition abuses its discretion when it makes an error of law." Koon v. United States, 518 U.S. 81, 100 (1996).

II A

The Eighth Amendment forbids the "inflict[ion]" of "cruel and unusual punishments." U.S. Const. amend. VIII. And the Supreme Court has held that because the Cruel and Unusual Punishments Clause prohibits "the unnecessary and wanton infliction of pain," it also prohibits "deliberate indifference to serious medical needs of prisoners." Estelle v. Gamble, 429 U.S. 97, 103-04 (1976) (quotation omitted). "Federal and state governments therefore have a constitutional obligation to provide minimally adequate medical care to those whom they are punishing by incarceration." Harris v. Thigpen, 941 F.2d 1495, 1504 (11th Cir. 1991).

"To show that a prison official acted with deliberate indifference to serious medical needs, a plaintiff must satisfy both an objective and a subjective inquiry." Farrow v. West, 320 F.3d 1235, 1243 (11th Cir. 2003). To meet the first prong, the plaintiff must demonstrate an "objectively serious medical need"—i.e., "one that has been diagnosed by a physician as mandating treatment or one that is so obvious that even a lay person would easily recognize the necessity for a doctor's attention," and, in either instance, "one that, if left unattended, poses a substantial risk of serious harm." Id. (alteration adopted) (quotations omitted). To satisfy the second, subjective prong, the plaintiff must prove that the prison officials "acted with deliberate indifference to [his serious medical] need." Harper v. Lawrence Cty., 592 F.3d 1227, 1234 (11th Cir. 2010) (quotation omitted). "To establish deliberate indifference," a plaintiff must demonstrate that the prison officials "(1) had subjective knowledge of a risk of serious harm; (2) disregarded that risk; and (3) acted with more than gross negligence." Id. (quotation omitted). An inmate-plaintiff bears the burden to establish both prongs. Goebert v. Lee Cty., 510 F.3d 1312, 1326 (11th Cir. 2007).

We pause briefly to flag a tension within our precedent regarding the minimum standard for culpability under the deliberate-indifference standard. Although we have repeatedly noted that "a claim of deliberate indifference requires proof of more than gross negligence," e.g., Townsend v. Jefferson Cty., 601 F.3d 1152, 1158 (11th Cir. 2010) (emphasis added), another line of our cases favors the phrase "more than mere negligence," McElligott v. Foley, 182 F.3d 1248, 1255 (11th Cir. 1999) (emphasis added). A panel of this Court recently suggested that "the 'more than mere negligence' standard in McElligott is more consistent with [Farmer v. Brennan, 511 U.S. 825 (1994)] than the 'more than gross negligence' standard in Townsend." Melton v. Abston, 841 F.3d 1207, 1223 n.2 (11th Cir. 2016). These competing articulations—"gross" vs. "mere" negligence—may well represent a distinction without a difference because, as explained below, the Supreme Court itself has likened the deliberate-indifference standard to "subjective recklessness as used in the criminal law." Farmer, 511 U.S. at 839-40 (emphasis added). Accordingly, no matter how serious the negligence, conduct that can't fairly be characterized as reckless won't meet the Supreme Court's standard.

B

The Secretary seems to concede (we think wisely) that chronic HCV is, at least as a general matter, a "serious medical need" within the meaning of our precedents. See Reply Br. at 5 n.5. And although the parties spar over the subsidiary question whether chronic HCV constitutes a condition "that, if left unattended, poses a substantial risk of serious harm" to F0- and F1-level inmates, Farrow, 320 F.3d at 1243, we find that it is unnecessary to address that issue. For the reasons that follow, we conclude that even if the plaintiffs could satisfy the objective prong of the two-part Eighth Amendment standard, they haven't shown that the Secretary's response to F0- and F1-level inmates' medical needs was deliberately indifferent in the subjective sense—in particular, that the Secretary "acted with more than gross negligence." Harper, 592 F.3d at 1234.

First, a bit of housekeeping. Based on two stray statements in its summary-judgment motion, the plaintiffs assert that the Secretary has "conceded" that the Eighth Amendment requires DAA treatment for F0- and F1-level inmates. Br. of Appellees at 29. We don't think so—seems to us like a game of gotcha. As for the first statement—that the "present-day standard of care is to treat" chronic HCV patients "with DAAs as long as there are no contraindications or exceptional circumstances"—the Secretary clarified on the very same page of the very same filing its position, which squares with Dr. Dewsnup's testimony, that "[t]here has been no showing that delaying treatment with DAAs for inmates with fibrosis scores of F0 or F1, until they reach F2, will cause harm to those inmates." As for the second—that "[i]t is inappropriate to only treat those with advanced levels of fibrosis"—the plaintiffs' argument simply misunderstands the import of the word "advanced." As Dr. Koziel herself explained, "advanced fibrosis" can be distinguished from F2-level fibrosis, which is "in the middle of the spectrum." Notably, the district court didn't rely on this so-called "concession" in its summary-judgment order.

1

We think it worth reiterating at the outset the stringency of the deliberate-indifference standard—because, it seems to us, the district court here lost track of it and impermissibly evaluated the Secretary's treatment plan against a negligence (or perhaps even more lenient) benchmark. It's an easy enough mistake to make: Confronted with an inmate who has a serious medical condition, a reviewing court hears from experts about measures that (in their view) would provide the most effective treatment. When the court then sees evidence that prison authorities aren't taking those measures—that perhaps they could be doing more, doing better—it concludes that liability must presumably follow.

Intuitive as that line of thinking may be—especially for lawyers and judges educated and trained in the common-law tradition—it does not reflect the constitutional standard. As we recently reiterated, "deliberate indifference is not a constitutionalized version of common-law negligence." Swain v. Junior, 961 F.3d 1276, 2020 WL 3167628, at *7 (11th Cir. June 15, 2020). "To the contrary, we (echoing the Supreme Court) have been at pains to emphasize that 'the deliberate indifference standard . . . is far more onerous than normal tort-based standards of conduct sounding in negligence,' and is in fact akin to 'subjective recklessness as used in the criminal law.'" Id. (quoting, respectively, Goodman v. Kimbrough, 718 F.3d 1325, 1332 (11th Cir. 2013), and Farmer v. Brennan, 511 U.S. 825, 839-40 (1994)). With respect to prisoners' medical care, in particular, we have held that the Eighth Amendment doesn't require it to be "perfect, the best obtainable, or even very good." Harris, 941 F.2d at 1510 (quotation omitted). Rather, we have emphasized, "[m]edical treatment violates the [E]ighth [A]mendment only when it is so grossly incompetent, inadequate, or excessive as to shock the conscience or to be intolerable to fundamental fairness." Id. at 1505 (quotation omitted).

The question here, therefore, isn't whether, in the best of all possible worlds, F0- and F1-level HCV-positive inmates should receive treatment with DAAs. Nor is it whether, if we were doctors, we would prescribe DAAs to all F0 and F1 patients. Nor, for that matter, is it even whether, if we were sitting as a common-law court, we might conclude that ordinary prudence requires across-the-board DAA treatment. Rather, because the plaintiffs here have invoked the Eighth Amendment, the sole question before us is whether the Secretary's approach to the treatment of F0- and F1-level inmates is so reckless—so conscience-shocking—that it violates the Constitution. As explained below, it is not.

2

We should first recap briefly what the Secretary's treatment plan is. For starters, the Secretary treats all HCV-positive inmates at the F2 level and above with DAAs. Among those at the F0 and F1 levels, the Secretary monitors their conditions and provides DAA treatment to those who either (1) have an exacerbating illness, such as HIV, (2) exhibit signs of rapid fibrosis progression, or (3) advance to F2.

So, to be clear—and contrary to misimpressions that the district court's opinion might leave, see Hoffer, 382 F. Supp. 3d at 1302-03—the Secretary isn't refusing or denying medical care to any HCV-positive inmate. He may not be providing F0- and F1-level inmates the particular course of treatment that they and their experts want—or as quickly as they want it—but he isn't turning a blind eye, either.

Two fundamental considerations—both of which reflect the steep hill that deliberate-indifference plaintiffs must climb—combine to demonstrate that the Secretary's treatment of F0- and F1-level patients is neither criminally reckless nor conscience-shocking. First, we have recognized that "[w]here a prisoner has received some medical attention and the dispute is over the adequacy of the treatment, federal courts are generally reluctant to second guess medical judgments and to constitutionalize claims which sound in state tort law." Harris, 941 F.2d at 1507 (quoting Westlake v. Lucas, 537 F.2d 857, 860 n.5 (6th Cir. 1976)); accord, e.g., Hamm v. DeKalb Cty., 774 F.2d 1567, 1575 (11th Cir. 1985) ("Although [the inmate] may have desired different modes of treatment, the care the jail provided did not amount to deliberate indifference."). Here, as already explained, the Secretary is providing "some medical attention" to F0- and F1-level inmates—diagnosing their illnesses, assessing their risk of future harm, and regularly monitoring and managing their disease progression. The plaintiffs' complaint isn't that the Secretary is providing no care, just that he isn't providing the more aggressive—and they say better—care that they desire.

To be clear, "some medical attention" doesn't necessarily demand curative care. Rather, medical intervention exists along a spectrum. At one end is ignoring medical needs entirely, which our decisions have rightly and repeatedly condemned: "'Choosing to deliberately disregard' an inmate's complaints of pain 'without any investigation or inquiry,'" we have held, constitutes deliberate indifference. Taylor v. Hughes, 920 F.3d 729, 734 (11th Cir. 2019) (quoting Goebert, 510 F.3d at 1328). The plaintiffs here, understandably, demand care at the other end of the spectrum—a prompt and an effective, albeit expensive, cure. There is, though, a range of responsible treatment options between the two poles that will satisfy the Eighth Amendment.

Consider an inmate with kidney disease. Surely the Constitution doesn't require prison authorities to schedule an immediate transplant, even though that might be the most effective, and permanent, solution. Rather, even for an inmate with end-stage renal disease, a regular course of dialysis treatments would doubtlessly pass constitutional muster. And for those whose condition hasn't progressed to near-complete kidney failure, even less aggressive measures—say, monitoring and managing diet and exercise—would presumably suffice. Or consider an inmate in the early stages of progressive hearing loss. Prison authorities needn't jump straight to cochlear implants, or even hearing aids. At least for a spell—until his condition worsens, anyway—the inmate may have to content himself with asking people to speak up. The long and short of it is that diagnosing, monitoring, and managing conditions—even where a complete cure may be available—will often meet the "minimally adequate medical care" standard that the Eighth Amendment imposes. Harris, 941 F.2d at 1504.

Second, and relatedly, we have emphasized—as have our sister circuits—that "a simple difference in medical opinion between the prison's medical staff and the inmate as to the latter's diagnosis or course of treatment [fails to] support a claim of cruel and unusual punishment." Id. at 1505; accord, e.g., Lamb v. Norwood, 899 F.3d 1159, 1162 (10th Cir. 2018) ("We have consistently held that prison officials do not act with deliberate indifference when they provide medical treatment even if it is subpar or different from what the inmate wants."); Kosilek v. Spencer, 774 F.3d 63, 82 (1st Cir. 2014) (en banc) ("[The Eighth Amendment] does not impose upon prison administrators a duty to provide care that is ideal, or of the prisoner's choosing."). That, at bottom, is exactly what we have here. The plaintiffs' expert, Dr. Koziel, opined that all F0- and F1-level inmates should be treated with DAAs, regardless of underlying condition or disease progression. The Secretary's expert, Dr. Dewsnup, testified, to the contrary, that it is sufficient to prescribe DAAs to inmates at level F2 and above, to monitor F0- and F1-level inmates, and to provide DAAs to F0s and F1s who either have underlying co-conditions or whose disease seems to be progressing rapidly.

The Secretary disputes the upshot of Dr. Koziel's testimony. In particular, the Secretary emphasizes (1) that all Dr. Koziel really said was that treating all HCV patients with DAAs is the "best possible management for Hepatitis C"—not that doing so is the only permissible course of action—and (2) that she acknowledged that she only "really, really start[s] to get agitated" when a patient reaches F3. See Br. of Appellant at 20. In response, the plaintiffs (1) challenge the Secretary's "implication" that Dr. Koziel's testimony suggests "that there is some other adequate form of treatment that is not the 'best'"—which, they insist, "is simply untrue"—and (2) that the Secretary has taken Dr. Koziel's "agitat[ion]" remark out of context. Br. of Appellees at 35. For present purposes, the parties' debate is beside the point. We can simply assume for present purposes that Dr. Koziel believes that all HCV-positive inmates should be treated with DAAs.

Because the plaintiffs here are receiving medical care—and because the adequacy of that care is the subject of genuine, good-faith disagreement between healthcare professionals—we are hard-pressed to find that the Secretary has acted in so reckless and conscience-shocking a manner as to have violated the Constitution. "Measured against constitutional minima," the Secretary's plan "evidence[s] at least tolerable and responsive medical treatment." Harris, 941 F.2d at 1507

C

In just two short pages, the district court reached the opposite conclusion. It did so principally on two bases. First, the court accepted the plaintiffs' contention that "[t]he only reason why FDC is electing not to provide [DAA] treatment is due to the cost of treatment, which is per se deliberate indifference." Hoffer, 382 F. Supp. 3d at 1302 (alteration in original) (quotation omitted). And second, it emphasized that the "FDC has not put forth any medical reason (nor does the record otherwise reveal a medical reason) why F0 and F1 inmates should not be treated." Id.

We conclude that the district court erred on both counts. We take up its errors in reverse order, because the second—concerning the Secretary's supposed failure to demonstrate a medical basis for declining DAA treatment for F0s and F1s—is easy to explain, while the first—regarding the role of "cost" in a deliberate-indifference analysis—will take bit of unpacking.

1

So first, the easy(ish) part. In stating that the Secretary "ha[d] not put forth any medical reason . . . why F0 and F1 inmates should not be treated," the district court triply erred. As an initial matter, the court impermissibly flipped the burden of proof. Our precedent is clear that an inmate bears the burden of proving all aspects of his Eighth Amendment claim—as relevant here, he "must satisfy the subjective component by showing that the prison official acted with deliberate indifference to h[is] serious medical need." Goebert, 510 F.3d at 1326; accord, e.g., Hill v. Dekalb Reg'l Youth Ctr., 40 F.3d 1176, 1190 (11th Cir. 1994) (noting that the plaintiff "submitted no medical evidence explaining how the four-hour delay in taking [him] to the hospital detrimented or worsened his medical condition"), abrogated on other grounds by Hope v. Pelzer, 536 U.S. 730 (2002). So it wasn't the Secretary's burden to demonstrate that treatment of F0s and F1s isn't medically necessary; it was the plaintiffs' burden to prove that such treatment is necessary. See H.C. by Hewett v. Jarrard, 786 F.2d 1080, 1086 (11th Cir. 1986) ("[T]he failure to provide diagnostic care and medical treatment known to be necessary [i]s deliberate indifference.").

To be sure, our decisions have said—incanted, really—that "a defendant who delays necessary treatment for non-medical reasons may exhibit deliberate indifference." Bingham v. Thomas, 654 F.3d 1171, 1176 (11th Cir. 2011) (citing Hill, 40 F.3d at 1187. That rule, of course, isn't absolute. It seems to us inconceivable, for instance, that the Eighth Amendment could be read to disable prison administrators from refusing an inmate's treatment request (no matter how medically sound) on the ground that it would threaten institutional security. To take just one example, would the Constitution require prison authorities to allow an inmate with severe allergies to keep an EpiPen® in his cell? If only "medical" considerations counted, then the answer would have to be yes—but of course the answer is no. In any event, none of our decisions has ever formally imposed a burden on prison officials to prove a medical reason for delaying or withholding an inmate's preferred treatment, let alone suggested that the failure to discharge that burden is fatal.

To compound matters, in connection with its determination that the Secretary had failed to discharge its phantom burden, the district court misconstrued Dr. Dewsnup's testimony in two key respects. First, immediately after faulting the Secretary for "not put[ting] forth any medical reason" for delaying DAA treatment for F0- and F1-level inmates, the court stated that Dr. Dewsnup had "explicitly noted . . . [that] F0 and F1 inmates must be treated eventually"—thereby implying that even the Secretary's own expert believed that medical considerations counsel in favor of treating all inmates. Hoffer, 382 F. Supp. 3d at 1302. Context, though, reveals the district court's error. As already explained, what Dr. Dewsnup actually said was (1) that in the main, F0s and F1s "don't need to be treated immediately" and instead "can be deferred," but (2) that "there are others who are lifetime prisoners that . . . [the FDC is] going to have to treat eventually even if they are Stage 0 or Stage 1." In context, it's clear that Dr. Dewsnup simply meant to acknowledge that many inmates who are currently at F0 or F1 will, if they remain incarcerated long enough, "eventually" progress to F2 and need DAA treatment.

Second, and even more conspicuously, as part of the same discussion, the district court emphasized that "both Dr. Koziel and Dr. Dewsnup agreed that—even for F0 and F1 inmates—successful treatment of HCV tends to decrease mortality rates." Hoffer, 382 F. Supp. 3d at 1302. That is demonstrably incorrect. In fact, Dr. Dewsnup said precisely the opposite. As already explained, he testified that (1) because the only available studies were from the Interferon era, when treatment required HCV patients "to be sober," and (2) because "substance abuse . . . [is] the major reason" that HCV-positive inmates die, there is simply no way to know whether it was the HCV treatment, the "huge lifestyle changes" that the Interferon regimen required, or some combination of the two that led to a decrease in mortality rates.

For all these reasons—legal and factual—we conclude that the district court erred in concluding that the Secretary had somehow failed to prove a medical justification for delaying DAA treatment to F0- and F1-level inmates.

2

Now for the tougher part—the role of "cost." Again, although the parties squabble over the exact price of a single DAA course of treatment—the Secretary pegged it at between $25,000 and $37,000 per inmate in 2017, while the plaintiffs insisted that it was lower as a result of discounts and rebates—there is no dispute that DAAs are expensive. The parties' real debate is over the legal questions whether, at what point, and to what extent the cost of a particular medical treatment should factor into an Eighth Amendment deliberate-indifference analysis.

As already noted, the district court agreed with the plaintiffs that "[t]he only reason why FDC is electing not to provide [DAA] treatment is due to the cost of treatment, which is per se deliberate indifference." Hoffer, 382 F. Supp. 3d at 1302 (alteration in original) (quotation omitted). On appeal, the parties can't even agree on how to interpret the court's order. For his part, the Secretary complains that the district court here improperly held that "consideration of the 'cost of treatment' in making medical decisions is 'per se deliberate indifference.'" Br. of Appellant at 16. In response, the plaintiffs insist that the district court merely—and properly—held that "[t]he Eighth Amendment prohibits the refusal of medical treatment solely because of costs." Br. of Appellees at 22. Because the issue of cost has become such a bone of contention in this case—and because the district court's decision and the parties' arguments reflect confusion about the meaning and application of our precedent—we will examine it in some detail.

The contest over cost's proper place in a deliberate-indifference analysis centers on our decision in Harris v. Thigpen. There, in the course of rejecting Eighth Amendment claims brought by HIV-positive inmates, the panel offered the following observations in dicta:

[W]e are troubled by and reject any suggestion . . . that a state's comparative wealth might affect an HIV-infected prisoner's right to constitutionally adequate medical care. We do not agree that "financial considerations must be considered in determining the reasonableness" of inmates' medical care to the extent that such a rationale could ever be used by so-called "poor states" to deny a prisoner the minimally adequate care to which he or she is entitled.

941 F.2d at 1509.

We agree with the Secretary that nothing in Harris—or in any of our other decisions—precludes prison authorities from "consider[ing] . . . the cost of treatment in making medical decisions." Br. of Appellant at 16 (quotation omitted). Although we have never had occasion to say so expressly, other courts have recognized the commonsense notion that "the civilized minimum" level of care required by the Eighth Amendment "is a function both of objective need and cost." Ralston v. McGovern, 167 F.3d 1160, 1162 (7th Cir. 1999). Need and cost are correlated, such that the more serious and exigent an inmate's need, the more likely it is that "the civilized minimum" might be deemed to require expensive treatment—and vice versa. See id.

Cf. also Woodall v. Foti, 648 F.2d 268, 272 (5th Cir. Unit A 1981) (holding that in deciding whether denial or deferral of medical treatment constitutes deliberate indifference, a court should "take into account a number of competing considerations," among them "the seriousness of the prisoner's illness [and] the need for immediate treatment" and the "availability and expense" of the requested care, keeping in "mind[] that the essential test is one of medical necessity and not simply of desirability").

The First Circuit's recently addressed this very issue in Zingg v. Groblewski, 907 F.3d 630 (1st Cir. 2018). There, an inmate suffering from severe psoriasis asked prison officials to give her Humira, an injectable "systemic treatment that targets the immune system," which she had been taking before her detention and to which she had responded well. Id. at 633. The prison officials denied her requests, treating her instead—and unsuccessfully—with topical steroids. See id. Eventually, after the inmate's psoriasis continued to deteriorate, the officials approved treatment with Humira, which led to "significant improvement in her condition." Id. The inmate sued under 42 U.S.C. § 1983, alleging that by delaying the authorization of Humira, the officials had acted with deliberate indifference in violation of the Eighth Amendment. Id. at 634. The district court granted summary judgment for the officials, and the inmate appealed. Id.

As particularly relevant here, the First Circuit addressed the inmate's contention that a jury could find that the prison officials "denied Humira because of its cost and that such a finding would suffice to permit a jury to find that [the prison officials were] deliberately indifferent to her serious medical needs." Id. at 638. For support, the inmate cited evidence, for instance, that Humira "is an expensive medication," that "cost containment was important to" prison administrators, and, indeed, that she had been told—point blank—that "Humira would not be approved because of its high cost." Id. Notwithstanding that smoking-gun evidence, the First Circuit rejected the inmate's legal argument: "We are not aware of any authority . . . to support the proposition that there is a per se Eighth Amendment prohibition against corrections officials considering cost, even when considered only in the course of selecting treatment that is aimed at attending to an incarcerated person's serious medical needs." Id.

We agree. Indeed, the law could hardly be otherwise. It is surely uncontroversial that "the deliberate indifference standard . . . does not guarantee prisoners the right to be entirely free from the cost considerations that figure in the medical-care decisions made by most non-prisoners in our society." Reynolds v. Wagner, 128 F.3d 166, 175 (3d Cir. 1997) (Alito, J.). Every minute of every day, ordinary Americans forgo or delay beneficial—and even life-altering—medical treatment because it's just too expensive. A couple decides to pass on in vitro fertilization in favor of less expensive (if also less effective) fertility treatment. A woman suffering from an autoimmune condition postpones an intravenous-immunoglobulin infusion because her insurance hasn't come through. Parents opt to delay reconstructive surgery for a physically disabled child. Healthcare can be expensive—sadly, sometimes prohibitively so. What a topsy-turvy world it would be if incarcerated inmates were somehow immune from that cold—and sometimes cruel—reality. See Maggert v. Hanks, 131 F.3d 670, 671 (7th Cir. 1997) ("A prison is not required by the Eighth Amendment to give a prisoner medical care that is as good as he would receive if he were a free person, let alone an affluent free person.").

So, to be clear, the Eighth Amendment does not prohibit prison officials from considering cost in determining what type (or level) of medical care inmates should receive. Nor, correlatively, are cost considerations off-limits to reviewing courts charged with determining whether prison officials have acted in so reckless and conscience-shocking a manner as to violate the Constitution.

Now, having said that, our cases do recognize an outer limit. While it is clear that cost can (and often will) be a relevant criterion in determining what the Eighth Amendment requires in a particular circumstance—what "minimally adequate care" entails in the first instance, Harris, 941 F.2d at 1509—it is also clear that cost can never be an absolute defense to what the Constitution otherwise requires. Put differently, if a particular course of treatment is indeed essential to "minimally adequate care," prison authorities can't plead poverty as an excuse for refusing to provide it. See, e.g., Anderson v. City of Atlanta, 778 F.2d 678, 688 n.14 (11th Cir. 1985) ("Lack of funds for facilities cannot justify an unconstitutional lack of competent medical care or treatment of inmates.").

But importantly, that outer-limit principle itself has limits. Cost isn't a sufficient excuse for failing to provide "minimally adequate care" only (and precisely) because "minimally adequate care" is—as Harris itself repeatedly emphasized and as we have reiterated here—quite minimal. See Harris, 941 F.2d at 1505 (observing that an inmate's "[m]edical treatment violates the [E]ighth [A]mendment only when it is so grossly incompetent, inadequate, or excessive as to shock the conscience or to be intolerable to fundamental fairness" (quotation omitted)). If the Constitution required prison authorities to provide all inmates with "perfect, the best obtainable, or even very good" care, id. (quotation omitted), then a rule prohibiting the state from raising cost as a defense would be inconceivable—and unsustainable. It would also impermissibly (and perversely) create a world—already discussed and dismissed—in which incarcerated prisoners would be constitutionally entitled to medical care, at taxpayer expense, that many private citizens can't get; there are, as already explained, countless instances in which ordinary Americans forgo particular medical treatments for the sole and exclusive reason that they can't afford them.

Accordingly, the two—the constitutionally required level of care and the propriety of a cost defense—necessarily go hand in hand. Here, because we have held that the Eighth Amendment's "minimally adequate care" does not require the Secretary to prescribe DAAs to all F0- and F1-level inmates, without respect to co-morbid conditions or disease progression, the issue of a cost "defense" never arises.

* * *

Because the Secretary has implemented a treatment plan that provides "minimally adequate care," we cannot say that his conduct is so reckless or conscience-shocking as to constitute deliberate indifference. Accordingly, the plaintiffs' Eighth Amendment claim fails as a matter of law. We reverse the district court's decision inasmuch as it requires DAA treatment for all F0- and F1-level inmates, vacate that portion of the injunction, and remand with instructions to grant summary judgment for the Secretary on that issue.

III

In addition to requiring DAA treatment of all F0- and F1-level HCV-positive inmates, the district court's permanent injunction also granted, in whole or in part, many of the plaintiffs' 14 other requests for relief. Although the Secretary challenged those additional items in the district court, he has narrowed his appeal (insofar as the Eighth Amendment is concerned) to the DAA issue. Even so, the Secretary maintains, with respect to all 15 items, that the district court's single "threadbare recitation" that its injunction satisfied the Prison Litigation Reform Act's narrowness, necessity, and non-intrusiveness requirements was insufficiently particularized. Accordingly, he insists, we must remand so that the district court can make (and explain) the PLRA-mandated findings with respect to all aspects of the permanent injunction other than the (now vacated) directive that all F0 and F1 inmates be treated with DAAs. We agree.

We respectfully disagree with our dissenting colleague, see Dissenting Op. at 50-51, that the Secretary has in any way "conce[ded]" or "stipulate[ed]" that the district court's injunction satisfies the PLRA's narrowness, necessity, and non-intrusiveness requirements. See Cason v. Seckinger, 231 F.3d 777, 785 n.8 (11th Cir. 2000). In the district court, the Secretary vigorously (if largely unsuccessfully) challenged all 15 forms of additional relief that the plaintiffs sought on summary judgment. The fact that the Secretary has narrowed his merits appeal to the question whether the Eighth Amendment requires DAA treatment of all F0- and F1-level inmates does not mean that he must forgo his separate argument that the district court's injunction—even as it pertains to other issues—violates the PLRA.

Under 18 U.S.C. § 3626(a)(1)(A), a "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." As we have explained, § 3626(a)(1)(A) "require[s] particularized findings that each requirement imposed by the preliminary injunction satisfies each of the need-narrowness-intrusiveness criteria." United States v. Sec'y, Fla. Dep't of Corr., 778 F.3d 1223, 1228 (11th Cir. 2015). And, as we have also explained, the particularity required by § 3626(a)(1)(A) is the same as that required by § 3626(b)(3), which applies when a district court declines to lift an injunction. See id. ("We see no reason why the term 'finds' in § 3626(a)(1) does not require the same particularity as the term 'findings' in § 3626(b)(3).").

In Cason v. Seckinger, we considered § 3626(b)(3) and concluded that "summary conclusion[s]" regarding the required PLRA requirements are "seriously deficient." 231 F.3d 777, 785 (11th Cir. 2000). We explained that the PLRA "requir[es] particularized findings, on a provision-by-provision basis," meaning that "[p]articularized findings, analysis, and explanations should be made as to the application of each criteria to each requirement imposed." Id.

The district court's one-sentence, boilerplate paragraph here regarding PLRA compliance was "seriously deficient." Id. Even where, as in this case, a district court issues a lengthy order that carefully considers each requested form of relief on the merits, it cannot just append a rote, catch-all assertion that "the permanent injunction satisfies the Prison Litigation Reform Act . . . because it is narrowly drawn, extends no further than necessary to effect the changes this Court concludes are constitutionally required, and is the least intrusive means of effecting such changes." Hoffer, 382 F. Supp. 3d at 1315. As we held in Cason, "[i]t is not enough to simply state in conclusory fashion that the requirements of the [injunction] satisfy" the PLRA's narrowness, necessity, and non-intrusiveness standards. 231 F.3d at 785. Rather, our precedent makes clear that if a district court's injunction grants 15 separate forms of relief, the court must make—and explain—15 separate PLRA-related findings.

Contrary to our dissenting colleague's contention, Thomas v. Bryant, 614 F.3d 1288 (11th Cir. 2010), does not suggest otherwise. See Dissenting Op. at 52-57. Nor could it, as the court there expressly declined to consider the question whether the injunction at issue satisfied the PLRA's particularized-findings requirement. See 614 F.3d at 1323 n.33 ("[T]he defendants do not specifically argue that the district court erred by not making the required particularized findings that the proposed injunction satisfies the PLRA's need-narrowness-intrusiveness requirement. . . . Because this issue is abandoned, we need not decide whether the district court's order satisfied our circuit's interpretation of § 3626(a)(1)(A).").
In response to our dissenting colleague's objection that enforcement of the PLRA's explicit findings requirement is "formalistic," see Dissenting Op. at 57, we can say only (1) that the Act requires what it requires and (2) that the statute's formalism serves one of its principal (and undisputed) purposes—namely, to require district courts to dot their i's and cross their t's before issuing or maintaining injunctions that interfere with prison administration. Cf. Harrington v. Richter, 562 U.S. 86, 102 (2011) (observing that if 28 U.S.C. § 2254's stringent standard for habeas corpus relief "is difficult to meet, that is because it was meant to be").
One final point: The PLRA's narrowness, necessity, and nonintrusiveness requirements also defeat the plaintiffs' alternative argument that a district court can grant an injunction broader than necessary to correct an Eighth Amendment violation. Whatever "substantial flexibility" district courts may enjoy to fashion Eighth Amendment relief, Brown v. Plata, 563 U.S. 493, 538 (2011), it does not countenance violation of the PLRA's explicit requirements.

IV

For the foregoing reasons, we reverse the district court's order inasmuch as it mandates DAA treatment of all F0- and F1-level HCV-positive inmates and remand to the district court with instructions to award summary judgment to the Secretary on that issue. We likewise vacate the district court's permanent injunction to the extent that it requires such treatment. We vacate the remainder of the order and injunction and remand to the district court so that it can make the findings required by the PLRA.

Order REVERSED IN PART, injunction VACATED IN PART, and case REMANDED. MARTIN, Circuit Judge, dissenting:

People who are incarcerated are stripped of "virtually every means of self-protection" and have no "access to outside aid." Farmer v. Brennan, 511 U.S. 825, 833, 114 S. Ct. 1970, 1977 (1994). Recognizing this, the Supreme Court told us decades ago that "prison officials must ensure that inmates receive adequate . . . medical care, and must take reasonable measures to guarantee the safety of the inmates." Id. at 832, 114 S. Ct. at 1976 (quotation marks omitted); see also Estelle v. Gamble, 429 U.S. 97, 103, 97 S. Ct. 285, 290 (1976) ("[E]lementary principles establish the government's obligation to provide medical care for those whom it is punishing by incarceration. An inmate must rely on prison authorities to treat his medical needs; if the authorities fail to do so, those needs will not be met.").

Our Court's ruling today reverses the District Court's ruling on summary judgment that found the Secretary of the Florida Department of Corrections (the "Secretary") to be deliberately indifferent to the medical needs of the Plaintiffs, Florida prisoners in the early stages of the chronic hepatitis C virus ("cHCV"). I do not share my colleagues' view of this case, and I would have affirmed the District Court. Beyond that, I am concerned that recent decisions of this Court will undermine the rights of our incarcerated citizens to maintain their health and safety while they serve their sentences.

HCV is a viral infection, which is spread by exposure to blood or blood products. "The most common way of contracting HCV is through intravenous drug use, but a person can also get infected through tattooing or blood transfusions." R. Doc. 153 at 2. I am concerned that a reader of the majority opinion might understand it to say this disease spreads only among people who engage in risky or deplorable behavior. See Maj. Op. at 2.

See, e.g., Swain v. Junior, 961 F.3d 1276, 1286-89, 1294 (11th Cir. 2020) (overturning the District Court's conclusion that prison officials were deliberately indifferent to serious risk of harm posed by COVID-19 and vacating preliminary injunction); Keohane v. Fla. Dep't of Corr. Sec'y, 952 F.3d 1257, 1264-65, 1277 (11th Cir. 2020) (overturning the District Court's conclusion that prison officials were deliberately indifferent to a prisoner's gender dysphoria because prisoners aren't constitutionally entitled to medical care "that is great, or even very good" (quotation marks omitted)); Marbury v. Warden, 936 F.3d 1227, 1238 (11th Cir. 2019) (per curiam) (declining to hold a prison official was deliberately indifferent to an attack on a prisoner because to do so based on "only some risk of harm" would erode "the 'subtle distinction' between deliberate indifference and mere negligence").

I take issue with four main points in the majority's deliberate indifference holding. First, the majority says that the disagreement between the parties' respective experts is "a simple difference in medical opinion." Maj. Op. at 17 (quotation marks omitted). In contrast, my reading of the expert testimony reveals more agreement than disagreement. The expert offered by the Secretary affirmatively acknowledged that all cHCV-positive prisoners need treatment. Second, the majority concludes the Secretary "isn't refusing or denying medical care to any HCV-positive inmate." Id. at 15 (emphases omitted). Yet it appears to me that refusing treatment is precisely what the Secretary is doing—at least up to a certain point. It is not until an cHCV-positive prisoner reaches the F2 stage that the Secretary provides treatment. This delay in treatment ignores the progression of the disease and the underlying damage that cHCV-positive prisoners experience in the meantime. Third, I think the majority gets it wrong when it says the District Court shifted the burden of proof from Plaintiffs to the Secretary. See id. at 19-22. As someone who once wrote opinions in a busy district court, I say the majority's judgment rests too heavily on a single sentence in the District Court's order. In directing so much attention to a summary sentence, the majority opinion fails to acknowledge the evidence of harm Plaintiffs did present to demonstrate that treatment of F0- and F1-stage patients is medically necessary. Fourth, I do not agree with the discussion in Section II.C.2 relating to the Secretary's cost arguments. My review of the record reveals no evidence of any reason, other than cost, for delaying treatment to F0- and F1-stage patients. Cost may be considered in determining whether a prison official is deliberately indifferent. However, in light of the Secretary's years-long delay in providing treatment for cost reasons, and the evidence showing the standard of care here is to treat everyone with cHCV, I would affirm the District Court's finding of deliberate indifference.

HCV attacks the liver and, in particular, can cause liver scarring called "fibrosis." Fibrosis can be measured, or staged, on a five-step scale in ascending order of severity, from F0 (no fibrosis) to F1 (mild fibrosis) to F2 (moderate fibrosis) to F3 (severe fibrosis) to F4 (cirrhosis).

And generally, I do not join the majority's strict application of the Prison Litigation Reform Act's narrowness-need-intrusiveness requirements. It is evident when reviewing the entirety of District Court's order that the court made proper findings. I view the majority's decision to remand for the District Court to further explain its findings as elevating form over substance. I also think it ignores the findings the District Court has already made.

I.

For years, the Florida Department of Corrections ("FDOC") denied Carl Hoffer, Ronald McPherson, and Roland Molina, along with thousands of others, proper medical treatment for their cHCV. The Secretary does not dispute this fact. In an attempt to remedy this wrong, the District Court ordered the Secretary to provide treatment to Florida prisoners suffering from cHCV. Everyone agreed that the proper standard of care was to treat cHCV-positive prisoners with direct acting antivirals, or DAAs. So that's what the District Court ordered—at least, with respect to prisoners at the more advanced stages of cHCV, who needed to be prioritized in the treatment queue at the preliminary injunction stage. See R. 465 at 24 (explaining that Secretary's expert "never testified that F0 and F1 inmates should never be treated," and that his statement that "it's safe to wait for a bit" for the treatment of F0- and F1-stage patients "was simply made in the context of a preliminary-injunction hearing and faced with the need to triage thousands of inmates").

Sadly, Mr. Hoffer succumbed to his illness before seeing the resolution of his case.

The majority does not view the Secretary as having conceded that DAA treatment is required for F0- and F1-stage patients. See Maj. Op. at 12 n.3. This view is apparently based on the Secretary's "clarifi[cation]" that "there has been no showing that delaying treatment with DAAs for inmates with fibrosis scores of F0 or F1, until they reach F2, will cause harm to those inmates." Id. (quotation marks omitted and alteration adopted). To support this position, the Secretary points to the purported lack or insignificance of symptoms experienced by F0- and F1-stage patients. But I cannot square either the Secretary's arguments or the majority's conclusion with the concessions the Secretary made in this case. While it may be true that some of the symptoms experienced by F0- and F1-stage patients, alone, may not rise to the level of a serious medical need requiring treatment, the Secretary has acknowledged that cHCV—which presents various symptoms that do not necessarily correlate with the seriousness of the underlining condition and risk to the health of the patient—is itself a serious medical condition. See R. 270 at 17 ("[T]he Court has concluded that cHCV constitutes a serious medical need. Defendant does not dispute this.").

This record reflects that the Secretary agreed that cHCV-positive patients must be treated with DAAs, until Plaintiffs asked the District Court to order the Secretary to provide treatment with DAAs to every cHCV-positive prisoner, as opposed to just the F2-, F3-, and F4-stage patients covered by the preliminary injunction. It was only then, when the Secretary realized the cost of these treatments, that this dispute arose. The Secretary draws the line—in my view, arbitrarily and without support from this record—between prisoners at less-progressed stages of cHCV and prisoners at stages that have progressed further. See R. 270 at 8 (The Secretary says "the present-day standard of care is to treat cHCV patients with DAAs . . . . There has been no showing that delaying treatment with DAAs for inmates with fibrosis scores of F0 or F1, until they reach F2, will cause harm to those inmates . . . ."). In other words, despite acknowledging that all cHCV-positive prisoners should be treated with DAAs, and that cHCV—which all Plaintiffs and class members have—is a serious medical need to which the Secretary was deliberately indifferent, the Secretary is now drawing lines between the stages of symptoms of the underlying disease.

See R. 465 at 7-8 ("This Court previously found that chronic HCV is a serious medical need. Defendant does not dispute that finding. . . . This Court also previously found that Defendant was deliberately indifferent to Plaintiffs' serious medical needs. Defendant does not dispute that either." (citations omitted)).

To its credit, the majority refuses to engage in the arbitrary line-drawing for which the Secretary advocates. See Maj. Op. at 12-13. Instead, I gather that the majority agrees with the Secretary's ultimate outcome because treating prisoners at the less-progressed stages of cHCV is the "best obtainable" care, rather than the "minimally adequate care" they are owed. Id. at 27 (quotation marks omitted). But the majority opinion's holding that the Secretary was not deliberately indifferent overlooks one important fact: the District Court found, and the Secretary did not contest, that the only reason cHCV patients were not treated was due to lack of funding. The District Court found—twice, in fact—that the Secretary admitted the failure to treat prisoners with cHCV, a serious medical need, was based solely on cost. And now, the Secretary draws lines between what stages of cHCV should receive treatment on the same rationale: cost. So even though the District Court found that the Secretary was deliberately indifferent by not providing treatment for the underlying disease that the Plaintiffs and class members all have, the Secretary now argues that he was not deliberately indifferent to F0- and F1-stage patients. It is on this basis that the Secretary seeks to limit the injunction entered by the District Court.

The majority finds it unnecessary to rule on this argument made by the Secretary. But in making its ruling, I fear the majority is rejecting facts found by the District Court and instead providing its own interpretation of the facts in the record. See Maj. Op. at 12-13. By reversing the District Court's order, the majority effectively denies treatment to prisoners who "plainly proved an unsafe, life-threatening [medical] condition . . . on the ground that nothing yet had happened to them." See Helling v. McKinney, 509 U.S. 25, 33, 113 S. Ct. 2475, 2481 (1993). Because this holding offends contemporary standards of decency and is also contrary to the Secretary's own expert testimony on the proper standard of care, I cannot join it. See id. at 32, 113 S. Ct. at 2480 (citing Estelle, 429 U.S. at 103-04, 97 S. Ct. at 290-91). A. THE PARTIES' EXPERTS AGREE THAT F0- AND F1-STAGE PATIENTS MUST BE TREATED.

The majority says that the parties' differing expert testimony reflects "a simple difference in medical opinion." Maj. Op. at 17 (quotation marks omitted). I don't think this view properly characterizes the testimony of the Secretary's expert. Indeed, neither Dr. Daniel Dewsnup, the Secretary's expert, nor Dr. Margaret Koziel, Plaintiffs' expert, dispute that the standard of care is to treat all cHCV-positive prisoners with DAAs.

Dr. Koziel gave a statement explaining that "[a]lthough the standard of care is to treat all persons with chronic HCV with DAA drugs, the choice of the regimen is dictated by . . . the progression (staging) of fibrosis and/or cirrhosis." She explained that "[t]he HCV guidelines call for all individuals with Hepatitis C to be treated [with DAAs] unless there is a contraindication to treatment." She reiterated that "all" patients, including those at the F0- and F1-stage, "need to be treated." However, when determining how "to set up a priority system for treatment," Dr. Koziel said that she would treat those with more advanced fibrosis and cirrhosis, which typically fall within the F3 and F4 stages, and then keep treating the next, less-advanced cHCV stages in descending order, from the F2 to F1 to F0. Dr. Koziel put it this way: "So if you have a list of 5,000 patients, it is reasonable to think I can't see 5,000 patients tomorrow. Let's find[,] of those 5,000[,] the individuals who are sickest and treat those first because they are most[] like[ly] to suffer the complications of liver disease."

These "HCV Guidelines" refer to the report published by the American Association for the Study of Liver Disease and Infectious Disease Society of America, entitled "HCV Guidance: Recommendations for Testing, Managing, and Treating Hepatitis C," and available at https://www.hcvguidelines.org/contents.

Dr. Dewsnup's testimony seems to align with that of Dr. Koziel. Dr. Dewsnup agreed that the HCV Guidelines recommend "treating everyone with chronic hepatitis C" with DAAs, and that this recommendation represents the standard of care. He noted that "in the correctional setting," the standard of care is "a goal that we should strive for." He recognized that in practice, due to the sheer number of prisoners, patients must be triaged and then prioritized "to treat the sickest first." Dr. Dewsnup admitted "this is the reality" of things, and he would treat less-advanced cHCV-positive prisoners "as soon as we can." He suggested prioritizing treatment as follows: people with decompensated cirrhosis "need to be prioritized for urgent, immediate treatment"; people with cirrhosis (F4-stage) "need to be treated very soon"; people at the F3- and F2-stages are slotted next; and "it's safe to wait for a bit" to treat the F0- and F1-stage patients, but they would need to be treated "eventually."

The majority views Dr. Dewsnup's testimony as meaning that treatment will eventually be required as a matter of disease progression, such that "many inmates who are currently at F0 or F1 will, if they remain incarcerated long enough, 'eventually' progress to F2 and need DAA treatment." See Maj. Op. at 19-21. This conclusion is not supported by the record. Dr. Dewsnup never testified that F0- and F1-stage patients should never be treated. And, while it might be true that many F0- and F1-stage patients may progress to the F2 stage slowly, neither expert said withholding treatment until a patient's condition deteriorates is appropriate. And, importantly, the experts did not recommend treating only F0- and F1-stage patients sentenced to longer prison terms.

Based on this testimony, the experts' opinions do not differ in the ways the majority seems to think. See Maj. Op. at 17-18. Rather, both experts agree that the standard of care requires treating all cHCV-positive prisoners and, while F0- and F1-stage patients need not be treated immediately, the goal is to treat them with DAAs. B. THE MAJORITY OPINION FAILS TO ADDRESS EVIDENCE SHOWING F0- AND F1-STAGE PATIENTS ARE AT RISK DURING THE PERIOD DAA TREATMENT IS DELAYED.

I agree, based on the expert testimony described above, that the FDOC's current policy of monitoring F0- and F1-stage patients is appropriate for some period of time. Notably, however, neither expert testified that treatment with DAAs could be delayed indefinitely for the F0- and F1-stage patients. The question thus becomes whether delaying DAA treatment for F0- and F1-stage patients is "tolerable"; and that "depends on the nature of the medical need and the reason for the delay." Farrow v. West, 320 F.3d 1235, 1247 (11th Cir. 2003) (quotation marks omitted). Courts have recognized that if symptoms of a disease get progressively worse, this factors into whether the delay constitutes gross negligence. See Goebert v. Lee County, 510 F.3d 1312, 1327 (11th Cir. 2007) ("Where the prisoner has suffered increased physical injury due to the delay, we have consistently considered: (1) the seriousness of the medical need; (2) whether the delay worsened the medical condition; and (3) the reason for the delay."). Again, the Secretary admits both that cHCV is a serious medical need and the reason for the delay is the cost of treatment. See Ancata v. Prison Health Servs., Inc., 769 F.2d 700, 704 (11th Cir. 1985) (concluding prison was deliberately indifferent when "necessary medical treatment ha[d] been delayed for non-medical reasons"). We therefore must examine the risks posed to F0- and F1-stage patients when their treatment is delayed, and whether, during the delay, Plaintiffs' conditions worsen such that treatment with DAAs is medically "necessary." See id.

First, no one here disputes that the "principal consequence" of HCV is "infection of the liver, which causes inflammation that in turn may result in scarring of the liver (fibrosis)." R. 270 at 4 (quotation marks omitted); see also R. 153 at 2. Neither is there any dispute that a patient's symptoms do not necessarily match up with the extent of liver scarring they are experiencing, or that a patient's symptoms correlate to their risk of liver failure. A person "can be completely asymptomatic and present with cirrhosis." R. 270 at 6 (quotation marks omitted); see also R. 153 at 5. Consistent with this, Dr. Dewsnup testified that he has measured the progression of F2- and F3-stage patients' symptoms from none "to end-stage liver disease." This record thus shows that symptoms are not indicative of the progression of the virus and the underlying damage to the liver. This matters because the Secretary is proposing delaying treatment until a patient reaches a more advanced stage. But because there is no dispute that once a patient reaches this higher F2 stage, scarring of the liver is already present, the Secretary is in essence waiting for F0- and F1-stage patients' condition to deteriorate to begin treatment. R. Doc. 270 at 4-5 (describing the F2 stage as "moderate fibrosis" and defining fibrosis as "scarring of the liver"); Doc. 153: 2-3 (same).

The Secretary's approach does not come with a precise time frame. The delay in treatment could be one year, or it could be twenty years. Either way, the experts' testimony provides evidence that treating cHCV patients is medically necessary and, in the absence of treatment, cHCV-positive prisoners experience liver inflammation that may progress to liver scarring. The Secretary's proposed outcome, adopted in the majority opinion, banks on the chance that some F0- and F1-stage patients may not ever progress to F2 while they are incarcerated and that some may not progress to the most serious fibrosis stage—cirrhosis—until 20 years from now. But the fact that a disease may progress slowly does not mean that prison officials may refuse to treat it. The Supreme Court has said as much, warning that prison officials cannot ignore harm "that is sure or very likely to cause serious illness and needless suffering the next week or month or year." See Helling, 509 U.S. at 33, 113 S. Ct. at 2480 (reconciling standard for conditions of confinement with deliberate indifference standard).

This record shows that all cHCV patients—no matter the stage—are at risk of suffering liver damage. At least one other circuit has reached the same conclusion based on similar cHCV evidence. In Gordon v. Schilling, 937 F.3d 348 (4th Cir. 2019), the Fourth Circuit critiqued the prison's decision to simply monitor a plaintiff diagnosed with acute HCV. Id. at 359. That court recognized that even though the plaintiff's lab tests were "normal," he "could be suffering from ongoing liver damage." Id. The Fourth Circuit concluded "it is inconsistent with the Eighth Amendment for a prison official to withhold treatment from an inmate who suffers from a serious, chronic disease until the inmate's condition significantly deteriorates." Id. I agree and would hold the same here.

The majority also takes issue with District Court's statement that the Secretary "ha[d] not put forth any medical reason . . . why F0 and F1 inmates should not be treated." Maj. Op. at 19. The majority says the District Court improperly removed the burden of proof from Plaintiffs and put it on the Secretary. See id. at 19-22. To arrive at this judgment, the majority seizes on one sentence in the District Court's order and ignores that Plaintiffs had already established that cHCV is a "serious medical need" within the meaning of our precedent. The majority's approach also overlooks that Plaintiffs did provide evidence to demonstrate that treatment of F0- and F1-stage patients is no less medically necessary—though perhaps less urgent—than treatment for F2-, F3-, and F4-stage patients. I believe the District Court properly found that Plaintiffs presented evidence sufficient to show that treatment of F0- and F1-stage patients is necessary, and that withholding necessary treatment based solely on cost demonstrates the Secretary acted with more than gross negligence. In fact, testimony from the Secretary's own expert compels this conclusion.

When asked if "it's not that big of deal" for patients at less-progressed stages to wait for medical treatment, Dr. Dewsnup responded, "I wouldn't put it exactly that way . . . a medical way that we could put it [is] . . . , look, I've got to treat the sickest first. You're not the sickest, but we're watching you. We're looking at your labs. We are going to do it as soon as possible. . . . We'll treat you as soon as we can." He further stated that, "to the extent that [the HCV Guidelines] can be applied, if we had the money and we have the system capacity, I agree with them. However, in Stage 0 and Stage 1, . . . I think those people don't need to be treated immediately." This testimony does not, as the majority seems to believe, establish that the Secretary's expert thought that denying DAA treatment to F0- and F1-stage patients is a medically acceptable practice. I also disagree with the majority's comparison of F0- and F1-stage cHCV patients to people in the early stages of hearing loss. See Maj. Op. at 17. The majority reasons that withholding treatment to F0- and F1-stage patients is like accommodating an individual with early hearing loss by speaking up. Id. Setting aside the gravity of initial hearing loss, the majority's analogy fails because Plaintiffs have established, and the Secretary admits, that cHCV is a "serious medical need" deserving of treatment.

To summarize, a review of this record, drawing all inferences in favor of the District Court's factual and credibility determinations, see Fla. Int'l Univ. Bd. of Trs. v. Fla. Nat'l Univ., Inc., 830 F.3d 1242, 1253 (11th Cir. 2016), supports the District Court's finding of deliberate indifference. I would uphold the District Court's decision to require the Secretary provide F0- and F1-stage patients with DAA treatment. The Secretary is here seeking to avoid paying for DAA treatment for a portion of cHCV-positive prisoners it is charged with caring for. But the Secretary should be required to fund this treatment because delaying it until the F2 stage, when a patient's condition has certainly deteriorated from liver inflammation to scarring of the liver, is inconsistent with the Eighth Amendment. C. THE DISTRICT COURT PROPERLY APPLIED HARRIS V. THIGPEN.

Affirming the District Court's findings of fact is especially warranted in this case, where the Secretary sought to get a final judgment by, among other things, asking the court to convert the preliminary injunction into a permanent injunction. The Secretary disputed none of the findings of facts made by the District Court at the preliminary injunction stage. See Fla. Int'l Univ. Bd. of Trustees, 830 F.3d at 1253 (applying deferential standard of review to resolution of cross-motions for summary judgment when "the parties intended to submit the case to the district court for final resolution").

Finally, the majority gives costs too large a role in determining whether the Secretary is deliberately indifferent. See Maj. Op. at 22-28. The rule is this: nothing precludes a prison from considering cost, but "cost can never be an absolute defense to what the Constitution otherwise requires." Id. at 27. Based on what the Secretary has actually argued in this case, it is not necessary for the majority to go to the lengths it does to emphasize the role costs play in its analysis.

The Secretary argues that the District Court incorrectly applied Harris v. Thigpen, 941 F.2d 1495 (11th Cir. 1991), when it "agree[d]" with an advocacy statement from the Plaintiffs' brief, which used the term "per se deliberate indifference." In Harris, our Court rejected "any suggestion in the [district] court's reasoning that a state's comparative wealth might affect an HIV-infected prisoner's right to constitutionally adequate medical care." 941 F.2d at 1509. The Harris court said it did "not agree that financial considerations must be considered in determining the reasonableness of inmates' medical care to the extent that such a rationale could ever be used . . . to deny a prisoner the minimally adequate care to which he or she is entitled." Id. (quotation marks omitted). To the extent the majority recognizes Harris as standing for this principle, see Maj. Op. at 27, I agree.

Contrary to the Secretary's argument, the District Court's analysis was in keeping with Harris. Although the District Court reordered the Harris analysis, it said that evidence supported the finding that "F0 and F1 inmates face substantial suffering and harm." The Court then noted, as the Secretary claims is required by Harris, that FDOC did not provide "any medical reason" why, contrary to the experts' conclusions, F0 and F1 inmates should not be treated. See Appellant's Br. at 18 ("Properly understood, Harris establishes that a state's inability to pay for constitutionally required care is not a defense to an Eighth Amendment violation."). Thus, the Secretary's argument that there is no precedent supporting the idea that "there is a per se Eighth Amendment prohibition" against prison officials "considering cost" does not address the findings actually made by the District Court. See id. at 17.

The Secretary's cost justification raises the question of whether its actions amount to more than gross negligence. Again, the Secretary admits that the reason for delaying treatment of F0- and F1-stage patients is the cost of DAAs. See Ancata, 769 F.2d at 704 (concluding prison was deliberately indifferent when "necessary medical treatment has been delayed for non-medical reasons"). This record contains no other medical reason for delay.

The majority says that because Plaintiffs are not entitled to treatment "in the best of all possible worlds," the Secretary's actions do not rise to the level of "subjective recklessness as used in the criminal law." See Maj. Op. at 13-14 (quotation marks omitted). Even accepting the majority's formulation of the level of negligence required to show deliberate indifference, the Secretary's cost justification and lack of other evidence to support withholding treatment rises above what is required to prove an Eighth Amendment violation.

To the extent the Secretary suggests Dr. Dewsnup testified about medical reasons for delaying treatment, that view misconstrues the record. As discussed above, Dr. Dewsnup testified about the practical reasons for prioritizing patients according to capacity and resources. While such a consideration is appropriate with respect to fashioning the scope of the remedy, it is not a medical reason that precludes a deliberate indifference finding. See Thomas v. Bryant, 614 F.3d 1288, 1324 (11th Cir. 2010) ("[T]he injunction appropriately balances respect for the DOC's administration of its own affairs with the need to ensure that [plaintiff] is not further subjected to [constitutional violations].").

* * *

I understand this record to support the District Court's factual findings that the Secretary was deliberately indifferent to the serious medical needs of F0- and F1-stage cHCV patients. I would therefore affirm the District Court's summary judgment order.

II.

The Prison Litigation Reform Act of 1995 ("PLRA") "imposes limits on the scope and duration of preliminary and permanent injunctive relief." Nelson v. Campbell, 541 U.S. 637, 650, 124 S. Ct. 2117, 2126 (2004). These limits include a requirement that, before issuing such relief,

[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). The Secretary argues, and the majority holds, that the District Court's order contained only a "threadbare recitation" of the PLRA's narrowness-need-intrusiveness requirements, which is insufficient as a matter of law. See Appellant's Br. at 27; Maj. Op. at 29. I believe this holding is flawed in two respects. First, I do not agree that the Secretary has properly challenged any prospective relief other than the District Court's direction to treat F0- and F1-stage patients. Second, the record reflects that the District Court did indeed balance the concerns for redressing the Secretary's constitutional violation with the Secretary's authority to run the FDOC. I would therefore hold that the District Court met the PLRA requirements.

The first step in examining whether the District Court properly made the narrowness-need-intrusiveness findings is to determine what findings the Secretary is challenging. We do not require the District Court to make particularized findings "concerning any facts or factors about which there is no dispute" or which the Secretary has conceded. Cason v. Seckinger, 231 F.3d 777, 785 n.8 (11th Cir. 2000). Even though the majority opinion agrees that the Secretary doesn't challenge any of those additional items under the Eighth Amendment, it vacates the entire injunction and instructs the District Court to make the narrowness-need-intrusiveness findings on all relief ordered. See Maj. Op. at 29, 31. And it does so without acknowledging Plaintiffs' argument that the Secretary conceded the District Court's findings satisfied the PLRA. The majority should have resolved this dispute first. Because it did not, the majority opinion is based on a flawed premise, and its resulting narrowness-need-intrusiveness analysis overturns findings the Secretary has conceded.

To properly analyze the Secretary's claim, we first must determine what findings the Secretary has properly challenged. See Cason, 231 F.3d at 785 n.8. The Secretary claims it did not concede the District Court's findings satisfied the PLRA "for any of the additional relief" ordered by the court at summary judgment. Appellant's Reply Br. at 17 n.17. The problem with this assertion is that the Secretary does not specify what "additional relief" it is actually challenging. See Thomas, 614 F.3d at 1323 (upholding injunction because, in part, "[t]he defendants d[id] not articulate any particular aspect of the injunction that they claim extends beyond the violation found by the district court"). Indeed, the Secretary's arguments focus only on whether it should be required to treat F0- and F1-stage patients. It has not referenced its challenge to any other "additional relief." It is thus clear from the Secretary's briefing that its aim in this appeal is to get the F0 and F1 patient-directive reversed, and I would analyze this claim as challenging only the portion of the order requiring the Secretary to treat F0- and F1-stage patients.

I now move to the next step of the analysis. Our Court has read the PLRA "as requiring particularized findings, on a provision-by-provision basis, that each requirement imposed by the [district court] satisfies the need-narrowness-intrusiveness criteria." Cason, 231 F.3d at 785. We may look to the entire record to determine whether a district court made these specific findings. See Sec'y, Fla. Dep't of Corr., 778 F.3d at 1228 ("Neither the December 6 [preliminary injunction] order nor any of the court's later orders contained specific findings . . . ."). After reviewing the record, I believe the District Court made the findings required by the PLRA.

Cason applied § 3626(b)(3) of the PLRA, which describes the standards for terminating prospective relief. 231 F.3d at 785. However, this Court extended the Cason particularity requirement to the term "finds" in § 3626(a)(1), which sets forth the standards for granting prospective relief, i.e., the standard that must apply in this case. See United States v. Sec'y, Fla. Dep't of Corr., 778 F.3d 1223, 1228 (11th Cir. 2015); see also Johnson v. Breeden, 280 F.3d 1308, 1326 (11th Cir. 2002) (same). I question the viability of this extension. Based on my reading of the statute, § 3626(a)(1) does not on its face require "written findings" in the way that § 3626(b)(3) does. Beyond that, most other circuits only apply the narrowness-need-intrusiveness findings requirement when there is a motion to terminate prospective relief under § 3626(b)(3). See Gates v. Cook, 376 F.3d 323, 336 n.8 (5th Cir. 2004) (rejecting the conclusion that § 3626(a)(1) requires particularized written findings); Morales Feliciano v. Rullan, 378 F.3d 42, 54-55 (1st Cir. 2004) (denying motion to terminate because "[t]he need for continued prospective injunctive relief is patent"); Jones-El v. Berge, 374 F.3d 541, 545 (7th Cir. 2004) ("So long as the underlying consent decree remains valid—and the defendants here have not (yet) made a § 3626(b) motion to terminate or modify the decree—the district court must be able to enforce it."). However, we are bound to follow Cason under our prior precedent rule. See United States v. Archer, 531 F.3d 1347, 1352 (11th Cir. 2008) ("[A] prior panel's holding is binding on all subsequent panels unless and until it is overruled or undermined to the point of abrogation by the Supreme Court or by this court sitting en banc.").

As this Court has previously recognized, a district court is "keenly aware" of the PLRA narrowness-need-intrusiveness requirements when its order "seemingly has fashioned a narrow injunction" targeting the constitutional violation. Thomas, 614 F.3d at 1323. In Thomas, this Court found no flaw in the District Court's order because it discussed the balance between "redressing the constitutional violation and recognizing that it is primarily defendants' job, and not the Court's, to run the prison system." Thomas v. McNeil, 2009 WL 605306, at *1 (M.D. Fla. March 9, 2009) (unreported), aff'd sub nom. Thomas v. Bryant, 614 F.3d 1288 (11th Cir. 2010); see Thomas, 614 F.3d at 1323 (citing district court order). The District Court in Thomas explained that it asked the defendants to submit a proposal for injunctive relief and found their refusal "puzzling" because it thought the defendants "would be in the best position to suggest the least intrusive injunctive relief," and the Supreme Court has strongly encouraged soliciting the defendants' views on the scope of relief. 2009 WL 605306, at *1. Based on this discussion (and because the defendants did not point to any specific deficiency), our Court upheld the injunction imposed by the District Court in Thomas. Id. Like that court, the District Court in this case made the appropriate narrowness-need-intrusiveness findings.

As the majority points out, the Thomas court noted that the defendants did not argue that the PLRA findings were not particularized enough. Maj. Op. at 31 n.8. In Thomas, the defendants made conclusory arguments that the injunction was too broad: "The defendants do not articulate any particular aspect of the injunction that they claim extends beyond the violation found by the district court or is overly intrusive." 614 F.3d at 1323. A panel of this Court reviewed the scope of the injunction anyway and upheld it. Id. In contrast to the way the majority frames the Secretary's arguments here, see Maj. Op. at 29 n.7, the Secretary does not articulate any particular aspect of the injunction that is too broad. And this is the same as was done by the defendants in Thomas. I say the Secretary's framing of the argument matters. In light of its "failure to articulate any specific deficiency," together with the findings the District Court did make here (as explained below), the record does not support a finding that the District Court abused its discretion in fashioning the injunctive relief . See Thomas, 614 F.3d at 1323.

Here, after the District Court explained why the permanent injunction requirements were met, it summarized in one sentence that the injunction satisfied the PLRA "because it is narrowly drawn, extends no further than necessary to effect the changes this Court concludes are constitutionally required, and is the least intrusive means of effecting such changes." But in its explanation preceding that summary language, the District Court considered the balance of requiring treatment for F0- and F1-stage patients in light of the magnitude of the existing constitutional violations, the available remedial alternatives, and the prison's interest in administering its own affairs. Our Court has endorsed the very approach employed by the District Court here. See Thomas, 614 F.3d at 1323-24. Other courts have as well. See, e.g., Morales Feliciano, 378 F.3d at 54-55 (upholding injunction fashioned to remedy constitutional violations "substantial in both scope and degree" in light of the prison's "abject failure" to cure such violations).

First, the District Court acknowledged that the Secretary was taking steps to treat the prisoners, but that over 2,200 prisoners still required treatment. The District Court thus considered that "the level of improvement still falls well short of bringing serious violations into constitutional compliance." See Morales Feliciano, 378 F.3d at 55.

Next, the District Court considered the Secretary's past actions, as well as the fact that he "continues to oppose relief" the District Court found to be constitutionally required. See id. at 54 ("The constitutional violations . . . . have defied correction for more than two decades."). Based on these considerations, the District Court found the Secretary was deliberately indifferent. In addressing the additional relief Plaintiffs asked for, the District Court found that some forms of relief were not constitutionally required and denied them. The District Court therefore implicitly addressed the narrowness and intrusiveness of the injunction when it fashioned the scope of the injunction. And, in granting the Plaintiffs certain forms of relief, the District Court noted it was "mindful of its role not to supervise prisons but to enforce the constitutional rights of prisoners," and gave the Secretary its choice of two options in implementing a remedy. See id. at 55 (upholding choice of one of three possible remedies when second was ineffective and third was undesirable).

In ordering the Secretary to treat F0- and F1-stage patients, the District Court found that not treating those patients amounted to a constitutional violation. The Secretary's own expert never testified that F0- and F1-stage patients should never be treated. Rather, the state's expert recognized that such patients must be treated with DAAs "eventually." In light of that fact, tailoring the remedy came down to timing: how quickly should F0- and F1-stage patients be treated? Plaintiffs proposed they be treated within two years, and the District Court recognized that the Secretary never "propose[d] an alternative date (other than never)." The District Court then weighed Plaintiffs' proposed timeline against the Secretary's progress in treating prisoners to date—over 4,900 prisoners treated over a 15-month period—and found the two-year timeline to be reasonable.

Like in Thomas, the order in this case reflects that the District Court was "keenly aware" of the PLRA requirements. 614 F.3d at 1323. Rather than focusing on the District Court's summary paragraph citing to the PLRA, we should more fairly look to the record to determine if it made the required findings. See Sec'y, Fla. Dep't of Corr., 778 F.3d at 1228 (scanning the entire record for evidence of PLRA findings). This circuit, like others, has upheld orders of prospective relief where the narrowness-need-intrusiveness findings are interwoven with the analysis of the facts and legal issues of the underlying case. See Thomas, 614 F.3d at 1323. Based on this record, I would affirm the District Court's injunction. This District Court did not "simply state in conclusory fashion" that the F0- and F1-stage treatment required by the injunction satisfied the PLRA requirements. See Cason, 231 F.3d at 785. Rather, it made particularized findings and explained its reasons for arriving at those conclusions. See Thomas, 614 F.3d at 1323.

See, e.g., Morales Feliciano, 378 F.3d at 55 ("In 2003, the district court was keenly aware of its duty to ensure that prospective remediation complies with the PLRA's requirements." (citing Morales Feliciano v. Calderon Serra, 300 F. Supp. 2d 321, 331 (D.P.R. 2004)); see also Armstrong v. Schwarzenegger, 622 F.3d 1058, 1071 (9th Cir. 2010) ("[W]e have upheld as sufficient under the PLRA overall statements by the district court that the need-narrowness-intrusiveness standard has been met . . . .").

The majority's holding poses practical problems by placing a rigid requirement on district courts in this circuit. It says that "if a district court's injunction grants 15 separate forms of relief, the court must make—and explain—15 separate PLRA-related findings." Maj. Op. at 31. But district courts carry substantial caseloads and our district judges are quite busy. Imposing a formulaic requirement about how district judges must set out factual findings that are not, in my view, required by the statute itself, forces courts to prioritize formalistic concerns. I view the majority's holding as impractical and wasteful of precious judicial resources.

For example, for the year ending March 31, 2020, there were over 95,000 filings in the district courts in this circuit. --------

* * *

The evidence in this record shows cHCV is a progressive disease that inherently puts all patients at risk of liver complications. Even patients at the F0 and F1 stages are at a serious risk of harm without eventual treatment. The Secretary's challenge to the District Court's finding of deliberate indifference, in an attempt to ultimately limit the scope of the injunction, runs contrary to the medical testimony in this case and our precedent. Because the majority opinion effectively allows the Secretary to withhold treatment indefinitely, I dissent.

I also dissent from the majority's holding that the District Court did not make particularized findings required under the PLRA. Although the District Court did not formulaically reference the narrowness-need-intrusiveness criteria, it performed the factual analysis our circuit and others have upheld. The majority's holding to the contrary is impractically rigid and places an unnecessary burden on our hard-working district judges.