This piece originally appeared in Luke O’Neil’s “Hell World“ newsletter.
For the homeless, COVID-19 posed something of a paradox.
With the help of federal funding, cities and states across the country covered the cost of hotel and motel rooms for their homeless populations, in a bid to reduce shelter crowding and give homeless people a safe place to quarantine. For the first time in decades, protecting the homeless was a priority; thus, the paradox.
These disease-prevention efforts not only controlled the spread of COVID, they also highlighted a desperate need for a key homeless support resource: medical respite shelters. A single medical respite shelter can have a greater positive impact on the homeless, hospitals and the shelter system as a whole than a brand-new traditional shelter with ten times as many beds. Yet in most cities the strategy for combating homelessness begins and ends with building more traditional shelters; as a result, there are only roughly 80 medical respite facilities in the entire country. And this gap is killing the homeless.
In order to fully appreciate the benefits of medical respite shelters, it is necessary to understand just how significant a threat homelessness poses to an individual’s health. Numerous studies have shown that the homeless have poorer long-term health outcomes relative to the housed population, and the average life expectancy for a homeless person is around 50 years—30 years shorter than a housed person.
Homelessness is a precarious existence. Homeless people are routinely subjected to verbal harassment, physical and sexual assault, and theft. An unsheltered homeless man died last week after he was set on fire while he slept in the stairwell of a public housing building in New York City. The de facto criminalization of street homelessness means unsheltered homeless people have, on average, 21 encounters with police per six-month period, which often leads to increased rates of incarceration and in turn makes the prospect of exiting homelessness even more dim.
These are the most attention-grabbing examples of the perils of homelessness. But the true danger, the thing that kills more homeless people than anything else, is far less sensational: a lack of space and time to rest and recover from illness. The sheer banality of this danger seems all the more outrageous in the wake of COVID-19, because as we learned during the pandemic, we have the resources to change this—we just choose not to use them.
—-
“Homelessness can happen to anybody,” says Dr. John McAdam. “They just need support for a little while, and then most will move on and become self-sufficient.”
McAdam spent 30 years providing health care to the homeless. His career began in 1983 at ground zero for homeless health care: St. Vincent’s Hospital in Greenwich Village. At St. Vincent’s, McAdam worked in the Community Medicine department under Dr. Philip Brickner, who is credited with almost singlehandedly establishing a new healthcare model geared specifically towards the homeless. “This whole [homeless health care] movement basically arose from Dr. Brickner,” McAdam said.
As McAdam explained the evolution of homeless health care from 1983 to now, I was struck by the sheer number of logistical obstacles that homeless people are expected to navigate as they recover from illness or injury. “In 1983, when a homeless person got discharged from the hospital, they’d be handed prescriptions,” McAdam said. But he estimates that 90% of his patients had no health insurance and couldn’t afford the cost of the prescriptions, so “we used to purchase stock medications and fill their prescriptions ourselves.” Shelter residents often missed their follow-up appointments, not because they didn’t want to go, but because they couldn’t make the trip on their own and nobody was available to escort them.
“Things have evolved over the last 30 years, I think overall for the better,” McAdam said. The Affordable Care Act made it much easier for the homeless to get health insurance, and an increased investment in case managers and social workers meant homeless people had more help navigating both the healthcare system and the shelter system. Nevertheless, he acknowledged, “we have a long way to go.”
Homelessness is itself a complicating factor when it comes to healthcare. The homeless often have complex healthcare needs; separating the knot of symptoms into their individual underlying conditions is a time-consuming and labyrinthine process, and modern emergency rooms are no place for such undertakings. Even if a provider has the time, patience and skill to properly identify the problem, unless they regularly work with the homeless, their treatment plans often fail to account for the unique limitations homelessness imposes upon the patient—which means they are unlikely to succeed.
Imagine that a homeless person goes to the hospital complaining of swelling in their legs and feet, and they receive the best possible care that could be expected under the circumstances. The provider takes the time to properly assess the patient and correctly concludes that the patient might be suffering from hypertension. The provider knows enough about homelessness and social determinants of health that they decide not to waste the patient’s time recommending lifestyle changes like “get some more fruits and vegetables in your diet” or “watch your sodium intake” or “try to limit your stress levels” because they recognize that the patient likely has no control over their diet, and their stressors are both inescapable and orders of magnitude more complex than those of the average patient. The provider wants to fix the problem fast, so they prescribe a diuretic—a water pill—to help the patient get rid of excess water and salt in their body and reduce the swelling in their legs and feet. The provider has the prescription filled at the hospital pharmacy, which saves the patient having to go somewhere else to pick it up. The provider discharges the patient with instructions to make sure they put their feet up.
But this solution—like so many others offered to homeless people—relies on incorrect assumptions about what is and is not possible when you’re homeless. Diuretics cause frequent urination, and homeless people often do not have ready access to a bathroom. And as McAdam pointed out, the homeless often can’t rely even on something as simple as the ability to put their feet up—even if they’re in a shelter. “In New York City, drop-in shelters are open 24/7, but there’s nowhere to lay down, so they have to sit up in chairs. And human beings weren’t meant to sleep sitting up in chairs.”
—-
Homeless advocates and healthcare providers have known for decades that it’s not the getting sick that imperils the homeless, it’s the inability to get better. In New York City, someone who is ill or convalescing and is fortunate enough to be in a long-term shelter with beds can receive a “bed rest” pass exempting them from the requirement that residents leave the shelter during the day. However, the typical limit for a bed rest pass is 3 consecutive days; after that, the resident is considered “not shelter appropriate” (i.e., too sick to be in a shelter) and they’re supposed to be sent to a hospital.
But what if you were just discharged from the hospital? Or what if your illness or injury is significant, but not quite significant enough to be admitted to the hospital? “Hospitals are all about quick turnover—they want to get people out of the hospital,” said McAdam. “Our patients can be not very pleasant, and a lot of them have advanced diseases when they show up, so ERs tend to stabilize them and send them back.”
Boston Health Care for the Homeless Program (BHCHP) founder Dr. James O’Connell saw the same problems in Massachusetts. “In 1985, most homeless people were too sick to handle being discharged [from the hospital] back to the shelter,” O’Connell told me. “They had to leave the shelter in the morning and weren’t allowed back in until 3 or 4 in the afternoon, and they were much too sick to be out on the street.” Using funds made available by then-Governor Dukakis, BHCHP established (“by fiat,” in O’Connell’s words) the nation’s first medical respite program for the homeless in 1985: 25 beds in Jamaica Plain’s Lemuel Shattuck Shelter.
“The concept was just to take [homeless] people coming out of the hospital, or who were in another shelter and were sick but maybe didn’t need the level of care a hospital would give but sure needed the kind of care we might normally get at home with a lot of family support: meals, visiting nurses, home health aides, things like that,” said O’Connell. “That was the first, simple concept of medical respite care.” In 1993, BHCHP opened the Barbara McInnis House, the nation’s first free-standing respite care facility, followed by the Stacy Kirkpatrick House, a step-down respite care facility, in 2016.
Despite the expansion of BHCHP’s medical respite program, there are still nowhere near enough beds to accommodate every homeless person who needs one. “We have 124 beds in total, and for every one bed that opens up, we get about 20 calls from shelters or hospitals asking if one of their residents or patients who needs respite care can take it,” O’Connell said. And as profitability continues to play an increasingly outsized role in hospital operations, it seems inevitable that things will only get worse.
“Hospitals used to keep people for a much longer period of time,” said O’Connell. “In 1985, for example, if you had cardiac surgery in one of the major hospitals in Boston, you would stay for 4 to 6 weeks in the hospital after your surgery. If you had cancer and you needed chemotherapy, we would admit you to the hospital for a week every month. If you had heart failure or a heart attack, you used to be admitted for 3 weeks. So you would do your acute care, your post-acute care, and part of your recovery in the hospital. And then several things happened.”
O’Connell believes that one of the big changes was the implementation of a new payment model known as diagnostic-related groups (DRGs). DRGs pay hospitals a predetermined amount based on the individual’s diagnosis, which means hospitals are incentivized to discharge patients as quickly as possible in order to keep as much of that payment for themselves as they can. Ironically, advancements in medicine have exacerbated the problem: many surgeries that once required lengthy stays in the hospital are now performed as same-day outpatient procedures, even though the recovery time remains unchanged. It still takes 4 to 6 weeks to recover from cardiac surgery, but in 2020, the average length of an in-hospital stay for cardiac surgery was just 5 to 7 days. Chemotherapy is now solely an outpatient procedure.
In New York City, the situation seems even more dire for the homeless. The city’s Department of Homeless Services (DHS) has strict guidelines that stipulate when hospitals are allowed to discharge patients back to a shelter, but there is no mechanism in place to enforce those guidelines in the moment. If an ambulance shows up at a shelter to drop off a patient who was just discharged, non-medical shelter staff aren’t qualified to assess whether the patient is appropriate or not, and medical staff know that refusing to accept a resident who’s not healthy enough to return to the shelter will just lead to a back-and-forth that only harms the patient. As a result, hospitals can—and often do—disregard those guidelines without repercussions.
Some of New York City’s shelters have medical clinics attached, but those clinics are meant to provide primary care: physicals, ongoing management of medical conditions, and the like. Shelter clinics lack the necessary equipment and staff to deliver post-acute or sub-acute care to residents just out of the hospital, or to residents who aren’t quite sick enough to go to the hospital but are too sick to leave the shelter during the day. The housed population can access visiting nurse services (VNS) or a home health aide to help them manage their daily activities or change dressings while they recover from surgery or illness, but that’s not an option for a shelter resident: anyone unwell enough to need a visiting nurse is, per DHS guidelines, likely medically inappropriate for a shelter setting. The only other option is skilled nursing facilities, but as O’Connell noted, shelter residents often don’t feel comfortable in these facilities—they often impose strict restrictions on what residents are allowed to do, where they’re allowed to go, even what they can eat and when. Besides, there aren’t enough skilled nursing facilities to meet the needs of the people who currently qualify for a spot, and they are nowhere near equipped to take on thousands of new patients with complex medical needs.
“The good changes in the overall healthcare system left homeless people in the lurch, and shelters are uniquely ill-equipped to deal with that,” said O’Connell. Naturally, the American healthcare system being what it is, BHCHP has been under increased pressure from hospitals to take homeless patients who are barely stable enough to transport, all so the hospital can make room for more profitable patients. So, in addition to providing the rehab and recuperative care that is the organization’s actual purview, BHCHP is now tasked with acute and post-acute care, pre- and post-operative care, even administering IV antibiotics. “When I go to the respite shelters, it’s like being on a hospital floor,” said O’Connell.
“All those people would have still been in the hospital in 1985.”
—-
The lack of medical respite facilities doesn’t just harm the homeless—though that harm alone should be reason enough to do something about it—but the hospital system and public health as well. Hospital readmission rates are higher for the homeless than the housed population, which puts more strain on the hospital system overall. Homeless patients who are forced to convalesce in a shelter take up beds that could otherwise have gone to a working homeless person who just needed a few days in the shelter to get back on their feet. And when I mentioned that the homeless are often the canaries in the coal mine when it comes to public health (e.g., the opioid epidemic), O’Connell agreed: “We’ve had Hepatitis A outbreaks, meningococcal outbreaks and flu outbreaks, [but] our respite has a wing to isolate people and contain the outbreak. There are many great public health uses for respite care.”
Everyone directly involved in homeless health care recognizes the value of medical respite shelters for the homeless—even the hospitals that contribute to the problem. “BHCHP receives a lot of funding from teaching hospitals who want safe places for their patients to get their care,” O’Connell told me. The main stumbling block, it seems, is everyone else.
Overall, New York City is arguably the most homeless-friendly city in the United States. It is one of just three places in the United States with a formal right-to-shelter law (along with Massachusetts and Washington D.C.) and the only one without restrictions on the right to shelter: Massachusetts’ law only covers families with children, and Washington D.C.’s only applies when the temperature is below 32ºF or above 95ºF. New York City also has an extensive homeless shelter and support system, and the state’s Medicaid expansion removed previously existing barriers to medical care for the homeless.
Yet even as New York City moves forward with plans for 90 new traditional shelters, there has been little—if any—focus on medical respite care for the homeless. “What little respite care there is in New York City is due to individual efforts rather than any uniform system,” McAdam said.
Expanding the shelter system to include medical respite care won’t happen overnight. O’Connell outlined the key obstacles that prevent this expansion from taking place, and the most significant is that the term “medical respite care” can mean different things to different people. “The lack of a uniform definition or parameters around what medical respite care is has made it difficult to marshal a national argument for respite care,” said O’Connell. Some of the haziness regarding what, exactly, constitutes medical respite care is beneficial. People in ideal health respond to illness and injury in different ways, and given the dizzying array of health conditions that commonly affect the homeless, a seemingly benign injury or illness can sometimes act as the catalyst for a cascade of far more serious medical problems.
Still, O’Connell is correct that any argument for respite care would need to establish a working definition of medical respite care and what it might entail, especially because it ties into another major hurdle: billing. CMS (the part of Health & Human Services that oversees Medicare & Medicaid) does not include medical respite care on its list of allowable payments, which means providers and facilities can’t bill Medicare or Medicaid for providing it. This forces respite facilities to focus more on the services they can bill for—like social services and recuperative care—than on the acute and post-acute care the homeless truly need. “I worry a lot that if Medicare and Medicaid don’t accept payment for actual respite care, [the need] will get worse,” said O’Connell.
These obstacles are very real, but it’s doubtful they are the reason there hasn’t been a significant mainstream push for medical respite shelters. More likely, the reason is that the majority of Americans picture homelessness the way it’s often portrayed in media: encampments, addiction, sleeping in public bathrooms or on subway trains, panhandling. Therefore, their strategies for combating homelessness often go no further than getting homeless people off the street. But in many areas, most homeless people are already off the street: the number of street homeless people in New York City only accounts for about 4% of the city’s total unhoused population.
Traditional shelters are most effective when they’re used as intended: as a temporary option for people who have a clear pathway out of homelessness. But many homeless people need more time and support than traditional shelters were meant to provide in order to find their pathway out of homeless, and an illness or injury can make things decidedly murkier. Without medical respite facilities, getting sick is the beginning of the end. O’Connell and McAdam both agree that it can’t hurt to build more traditional shelters, but our efforts can’t end there. The goal of homeless programs isn’t to put the homeless somewhere we don’t have to see them, but to help people exit homelessness, and that can’t happen with traditional shelters alone. O’Connell has 35 years’ worth of evidence on the benefits of medical respite care for the homeless, and our response to the pandemic provided a blueprint for how we might make it a reality. But outside of a small and dedicated group of people nobody seems all that interested in trying, so the homeless remain trapped in limbo.
“Some of my patients are the children of the men I treated back in the ‘80s and ‘90s. Their home life…we can’t imagine how they grew up,” said McAdam.
“It really is a self-perpetuating mess.”
A Hopeless Cynic 
