COVID-19 is spreading rapidly in U.S. jails and prisons, but testing of inmates and staff remains spotty and many confirmed cases are going unreported. The resulting lack of data has deep implications for the fight against the virus, because prison outbreaks can move easily to surrounding communities.
When COVID-19 began tearing through Detroit’s county jail system in March, authorities had no diagnostic tests to gauge its spread. But the toll became clear as deaths mounted. First, one of the sheriff’s jail commanders died; then, a deputy in a medical unit.
“Working in the Wayne County Jail has now become a DEATH sentence!” the head of the deputy sheriffs’ union, Reginald Crawford, wrote on Facebook as the losses mounted.
By mid-April, the jail system’s medical director and one of its doctors also had died from COVID-19, the disease caused by the new coronavirus. The virus was everywhere, but jail officials had little sense of who was infected and spreading it. Testing of inmates and staff – needed to determine who should be quarantined to slow transmission – was just getting started. In the weeks since, more than 200 staff and inmates have tested positive.
COVID-19 has spread rapidly behind bars in Detroit and across the nation, according to an analysis of data gathered by Reuters from 20 county jail systems, 10 state prison systems and the U.S. Bureau of Prisons, which runs federal penitentiaries. But scant testing and inconsistent reporting from state and local authorities have frustrated efforts to track or contain its spread, particularly in local jails. And figures compiled by the U.S. government appear to undercount the number of infections dramatically in correctional settings, Reuters found.
Reuters documented well over three times the CDC’s tally of COVID-19 infections – about 17,300 – in its far more modest survey of local, state and federal corrections facilities conducted about two weeks later. The Reuters survey encompassed jails and prisons holding only 13% of the more than 2 million people behind bars nationwide. Among state prisons doing mass testing of all inmates, Reuters found, some are seeing infection rates up to 65%.
The CDC tally “is dramatically low,” said Aaron Littman, a teaching fellow specializing in prison law and policy at the law school of the University of California, Los Angeles. “We don’t have a particularly good handle” on COVID-19 infections in many correctional and detention facilities, “and in some places we have no handle at all.”
Problems with unreliable data aren’t unique to corrections. Epidemiologists say the incidence of COVID-19 in the general U.S. population also is unclear due to limited testing, especially in the pandemic’s early days. And the CDC acknowledged in its report that its infection count for jails and prisons was similarly hampered by spotty data and “not representative” of the disease’s true prevalence in those facilities.
But uneven testing for COVID-19 in correctional settings and erratic reporting of confirmed cases have profound implications for health officials and policy makers tracking its spread, because epidemiologists see jails and prisons as key pathways of transmission.
Evidence of “a crisis”
The United States has more people behind bars than any other nation, a total incarcerated population of more than 2.2 million as of 2018, including nearly 1.5 million in state and federal prisons and just under 740,000 in local jails, according to the U.S. Bureau of Justice Statistics.
Jails generally keep inmates for short stays: arrestees awaiting trial or people serving short sentences. The churn of these inmates raises the risk of infections among both the inmates themselves and jail staff, who can carry the virus to and from the community.
“It’s like a bad dream”
Prisons, which hold convicted criminals on longer sentences, also are fertile ground for the virus. While inmates come and go far less frequently, the pathogen can be carried in from the community by a single contagious staffer, spread quickly in crowded cell blocks, and be re-introduced to the community by other, newly infected workers.
Reuters collected data from 37 state prison facilities across the country that have done mass testing for COVID-19 among all inmates, including those with no symptoms, and found more than 10,000 confirmed cases among the 44,000 tested. There were 91 deaths from the disease at those facilities, which span 10 states.
In contrast, federal prisons, which typically limit testing to inmates with obvious symptoms, reported confirmed infections in fewer than 4,200 of their total inmate population of about 150,000, with 52 deaths.
The situation in the nation’s 2,800 local jails is even more opaque. Many don’t report their COVID-19 cases publicly, and there is no national tracking of their infection numbers.
Reuters surveyed the 20 U.S. counties with the largest jails, holding an average total of about 73,000 inmates, and found nearly 2,700 confirmed COVID-19 cases – a figure that has risen nearly 30-fold over the past six weeks. While some of that increase is a result of increased testing during that time, it still reflects an almost certain undercount, because testing remains limited in many of those facilities.
The surge in jail infections comes amid a chorus of concerns from judges, oversight agencies, corrections officers, defense lawyers and civil rights groups that most local lockups are ill-equipped to control the virus, which has killed at least 310,600 people worldwide. Unlike state and federal prisons, typically equipped to provide health care for long-term inmates, jails often have little medical capacity.
In health care, jail inmates “are the last and the least and the lost,” said Dr. Thomas Pangburn, chief medical officer for Wellpath LLC, the medical contractor in Wayne County’s jails and hundreds of others nationwide. Many jails have been overlooked in the race to secure COVID-19 test kits and medical supplies for hospitals and nursing homes, he said, but “we have the most vulnerable population in a very confined space meant for correctional housing – and not for medical care.”
In many jails and prisons, the toll of COVID-19 on corrections officers and other staff approaches that of inmates – and here, too, the numbers reported to the CDC by state and local authorities appear to be a vast undercount.
The CDC report documented nearly 2,800 COVID-19 cases among staff across all U.S. correctional facilities. But Reuters found more than 80% of that number – upwards of 2,300 infected jail and prison workers – in its far less comprehensive survey of just the federal prison system, a few dozen state prisons and the 20 counties with the biggest local jails.
In an effort to curb infection rates, many jails and prisons are releasing inmates to create more distance among those remaining behind bars. That has raised concerns about whether inmates, particularly in jails, are being screened for COVID-19 before returning to the community, where many can’t get medical care.
More than 37,000 state and federal prisoners have been released since March 31, according to U.S. government data and records collected from 41 state prison systems by the Vera Institute of Justice, a research group that seeks to reduce incarcerated populations. There is no national tracking of local jail releases, but in just the 20 counties surveyed by Reuters, at least 14,000 jail inmates have been let go.
Releasing inmates is critical “both in jails and surrounding communities, because of the role jails serve as vectors” for spreading the virus, said Udi Ofer, justice division director at the American Civil Liberties Union, which has filed dozens of “decarceration” suits and legal petitions. “It’s a crisis.”
Some groups have pushed back. Victims’ rights group Marsy’s Law, named after the murdered sister of billionaire Henry Nicholas, has criticized the releases, expressing concern that crime victims aren’t always notified when inmates are let out.
No tests, no protection
U.S. President Donald Trump declared the COVID-19 pandemic a national emergency on March 13. By that time, officials in the Wayne County Sheriff’s Office already were scrambling to address a looming outbreak in their three jails.
Days earlier, Sheriff Benny Napoleon and Chief Robert Dunlap, the jails supervisor, had laid plans to keep inmates more separated, cut public visits and quarantine new arrivals – rules that took effect just after Trump’s announcement. On March 19, the jail also began releasing low-level offenders, for the most part inmates with risky medical conditions.
Staff and inmates were already falling ill across the jail system, which typically houses a population of about 1,400. Donafay Collins, 63, a jail commander, was hospitalized with a COVID-19 infection that would kill him less than two weeks later – the first death among the four staffers claimed by the virus.
“It’s like a bad dream,” Chief Dunlap said in an interview with Reuters.
Meanwhile, getting diagnostic tests and protective equipment to track and manage the virus proved challenging, Dunlap said. Suppliers had little to offer, and just about everything they had was going to hospitals and emergency medical services.
Hunting for face masks, the sheriff’s office turned to Michigan Governor Gretchen Whitmer, who had been pressing the federal government to give states more supplies from federal stockpiles. On March 20, state officials sent the sheriff 7,500 N-95 masks provided by the U.S. Department of Homeland Security. The highly protective masks are used by staff handling sick inmates, Dunlap said. Basic surgical masks became available later for more routine use by both staff and inmates, he added.
Getting test kits proved even harder. As COVID-19 raced through Wayne County’s jails in March, corrections officers needing tests had to visit a local testing center or hospital, where they often were refused if they did not show specific symptoms, Dunlap said. It wasn’t until April 6 – the day the virus killed the jails’ medical director, Dr. Angelo Patsalis – that officers began getting regular tests through the Wayne State University Physician Group.
Getting the tests was “a matter of life and death,” said Crawford, the head of the deputy sheriffs’ union, in an interview.
For inmates, however, testing remained elusive.
In late March, the sheriff directed the jail’s medical contractor, Wellpath, to obtain test kits for inmates, but the company couldn’t get enough due to heavy demand, Dunlap said. “Wellpath, like every other provider around this county, couldn’t get them.” So, COVID-19 testing was limited to inmates with symptoms.
By April 30, the jail’s population had dropped to just 834 inmates – about 500 had been released – and only 89 had been tested for the new coronavirus. Of those tested, 29 were positive, just over 30%, according to the sheriff’s office. Among the sheriff’s 810-member staff, 196 had tested positive, or 23% – of whom 89 have returned to work.
On May 7, the jail expanded testing to all inmates under a grant from the Hudson Webber Foundation. That should “further mitigate the spread of the virus” inside and outside the jail, Dunlap said, and help identify infected inmates before release.
As in many states, Michigan’s prison system began universal testing earlier than the jails.
“If you don’t know where the problem is, you can’t fix it.”
On April 21, Michigan’s Department of Corrections began testing for coronavirus infections in large numbers of inmates even if they showed no sign of illness, said department spokesman Chris Gautz.
Demands for mass testing are growing. The ACLU and the Council of Prison Locals, representing 30,000 federal prison employees, called earlier this month for universal testing in all federal lockups.
But some public health experts are ambivalent on that approach. The CDC’s guidance for correctional facilities calls for quick COVID-19 testing of inmates who appear symptomatic, but it takes no position on universal testing.
The guidance reflects a belief among some public health experts that testing only symptomatic inmates and, in some scenarios, a sample of the rest may suffice for assessing the virus’ overall prevalence in a jail or prison, said Marc Stern, former medical director for the Washington State prison system and a faculty member at the University of Washington School of Public Health. Testing every asymptomatic inmate may not make sense if a jail lacks the capacity to isolate and trace the contacts of those who test positive – and also because not everyone who tests positive may be contagious.
In Michigan’s prison system, however, officials say mass testing has been valuable.
“If you don’t know where the problem is, you can’t fix it,” spokesman Gautz said.
‘It’s crazy in there’
Charles Peterson, 78, began showing symptoms of COVID-19 a week after a parole violation landed him in Colorado’s Weld County Jail on March 11. By the time he was released on March 30, he was on the verge of dying from it.
Peterson declined quickly, two fellow inmates told Reuters. Coughing and disoriented, they said, he eventually struggled to stand and began losing control of his bladder and bowels.
Donovan Birch said he and other inmates alerted jail staff, but Peterson was left in the general population. Birch also became ill with COVID-19 symptoms after his exposure to Peterson, he said, but never was tested.
Peterson “needed help,” said Birch, who was jailed on a parole violation for trespassing charges. “I knew he was going to die if he didn’t get it.”
Instead, Peterson was released; two days later, he was dead. Official cause: “acute respiratory failure, viral pneumonia and COVID-19 infection.”
Peterson likely was a “superspreader,” according to an infectious disease expert who inspected the jail on behalf of inmates for a lawsuit they filed seeking better sanitary and safety measures. By early May, at least 10 of the jail’s roughly 480 inmates had tested positive for the virus – but just 22 had been tested. Eighteen deputies had also been infected, the jail said.
The inmates’ lawsuit claims Weld County Sheriff Steven Reams “willfully disregarded public health guidelines” by leaving three to four inmates to a cell, sharing sinks and toilets, as the virus spread. “Failing to prevent and mitigate the spread of COVID-19 endangers not only those within the institution,” the suit says, “but the entire community.” Reams declined to comment.
The case is among more than 100 lawsuits nationwide, many of them class-action cases, seeking mass releases of inmates or other measures to reduce overcrowding and infection risks in jails hit by the new coronavirus, according to the UCLA law school’s COVID-19 Behind Bars Data Project. Many of those cases, as well as hundreds more filed by individual inmates, argue that confinement in facilities with COVID-19 outbreaks violates the U.S. Constitution’s protections against cruel and unusual punishment.
Inmates have been issued masks since early April and have access to soap, hand sanitizer and cleaning supplies, said Weld County Sheriff’s spokesman Joe Moylan. He noted the jail has been on lockdown since April 1 – the day Peterson died – and inmates are rotated out of their cells in small groups to common areas that allow for social distancing. He declined to comment on the litigation and the specific cases of Peterson and Birch.
Peterson was released after Colorado’s Department of Corrections decided not to hold him for his parole violation, part of the effort to slow COVID-19 transmission in local jails by reducing inmate populations. Since March 1, the jail has reduced its population by more than 300 inmates; fewer than half its 954 beds are occupied.
Peterson’s parole violation involved failing to renew his sex offender registration while living at “Rock Found,” a re-entry home for convicts returning to the community. When he was let out of jail, a former Rock Found roommate brought him back to the home, cold, shivering, barely able to walk.
The program director called paramedics.
“I honestly could not believe that not a single person from the Weld County Jail had told anyone at Rock Found that they were releasing a seriously sick person into our care,” the director, Cheryl Cook, said in a statement filed in the inmates’ lawsuit.
Moylan, the sheriff’s spokesman, said Peterson was not tested for COVID-19 because he was not overtly symptomatic.
The conditions at the jail violated the constitutional rights of medically vulnerable inmates, a federal judge ruled May 11. He ordered the sheriff to socially distance those at risk, provide single cells when possible, and improve cleaning of communal spaces.
Many of the problems addressed by the judge were identified by the plaintiffs’ expert witness during two visits to the jail in April. He reported to the court that he found most inmates confined to group cells more than 22 hours a day with no handwashing options unless they were let out to a bathroom. Many complained of unsanitary conditions and said shared sinks and toilets were not cleaned between uses, the expert reported.
Ralph Brewer, 41, jailed for violating a restraining order, told Reuters he was directed to continue working in the kitchen after developing nausea and a bad cough. Staffing was short, he was told, so he had to work unless he had a fever.
“It really concerned me. We had no masks, just gloves,” Brewer said. He requested a doctor to check his lungs, he said, but nurses only gave him Tylenol, cough medicine and instructions to stay hydrated.
Brewer was released on April 3 and his daughter took him straight to an urgent care clinic. The doctor said he had COVID-19 symptoms – no tests were available – and told him to quarantine for 14 days, Brewer said. He recovered at his mother’s house.
“I was lucky to get out, but I’m worried about the people still in jail,” Brewer said. “It’s crazy in there.”
Editor’s Note: This story was updated to correct the name of the head of the deputy sheriffs’ union to Reginald Crawford
Additional reporting by Grant Smith
By Peter Eisler, Linda So, Ned Parker and Brad Heath
Data: Grant Smith
Photo editing: Corinne Perkins
Video: Linda So and Daniel Fastenberg
Art direction: Troy Dunkley
Edited by Jason Szep