New York University Grossman School of Medicine, New York, NY, USA.
Correspondence: Ariana.CobaClementel@nyulangone.org
Conflict of interest statement: The authors have no conflicts of interest relevant to this article to disclose.
Release documentation: All images were generated by the authors in “R Core Team. (2024). R: A Language and Environment for Statistical Computing. R Foundation for Statistical Computing, Vienna, Austria.” All in-text information cited according to journal standards.
Keywords: medical respite; homelessness; discharge; health care stewardship
Medical respite programs provide short-term housing and clinical support for unhoused patients who are too ill to recover safely on the streets or in shelters but no longer require hospitalization. This perspective reviews the clinical, financial, and systems-level benefits of medical respite, including reduced readmissions, fewer emergency department visits, improved outpatient follow-up, and lower hospital costs. Despite these benefits, access remains limited by insufficient funding, variable eligibility criteria, administrative complexity, and low provider awareness. Using New York City and other state programs as examples, the authors argue that expanding medical respite requires increased investment, streamlined referral pathways, and stronger clinician advocacy. Medical respite represents a patient-centered and cost-effective approach to safe discharge planning, health care stewardship, and improved care for unhoused patients.
Medical respite (MR) programs originated in Boston and Washington, D.C., in 1985 to provide shelter and resources to unhoused patients who were too sick to recover on the streets but not sick enough to remain hospitalized.1,2 Without respite, patients are discharged to the streets or to shelters, where environmental exposure, infection risk, and poor follow-up impair recovery — often leading to deterioration, readmission, or death. Recognizing this burden on patients and hospitals, many states followed Boston’s lead in implementing MR programs. However, access and utilization remain limited nationwide. This article examines the benefits, limitations, and prevalence of MR programs across the United States and argues that improving clinician awareness is a critical first step toward expanding access.
For decades, studies have consistently demonstrated the clinical benefits of MR for unhoused patients. Homeless individuals have an average of six emergency department visits annually, as compared with 1.6 visits among housed counterparts.3 Thirty-day revisit rates are 5.7 times as high in the emergency department and 1.9 times as high for inpatient admissions, reflecting poor post-hospital recovery.4 Although the cause of frequent readmissions is often multifactorial, unhoused patients are more likely to receive discharge plans that leave them vulnerable immediately after hospitalization. A Connecticut study of 98 unhoused patients found that 11% spent their first postdischarge night on the street, 66% went to shelters, 26% were discharged after dark, and 59% lacked a safe discharge plan.5 MR directly addresses these gaps.
Studies consistently show that MR improves outcomes across various metrics. MR reduces 90-day readmissions by up to 50%, lowers emergency department visits by 1.2 visits per patient, and decreases hospitalizations by 29%.6,7 When readmission occurs, MR paired with permanent supportive management reduces the length of stay by 2.7 days.7 MR also increases scheduled outpatient visits by 192.6%, representing a major step toward sustained care.8 Beyond clinical metrics, MR addresses social determinants of health: a California study found that 59% of participants secured stable housing or reunited with relatives after discharge.9 Benefits extend across conditions, including postsurgical recovery, traumatic brain injury, HIV care, and outpatient antibiotic therapy.10−14 Notably, one study found no significant outcome differences between patients with and those without substance use disorders, countering assumptions that substance use disorders limit treatment success because of greater complexity of care.15
Research similarly shows that MR is financially advantageous for hospitals. On average, unhoused patients stay 4.1 days longer, costing $4094 more than housed patients.16 In Florida and Connecticut, inpatient care costs for unhoused patients exceeded reimbursement by 48% and 26%, respectively.17 Discharge delays driven by housing instability significantly increase costs. In Minnesota, unhoused patients accounted for 42.9% of nonmedical discharge delays during a study period.18
Accordingly, MR delivers substantial financial relief by reducing emergency department visits, admissions, and length of stay.13,19−21 One analysis found that hospitals saved $1.81 for every $1 invested in MR, largely by avoiding inpatient costs.17 In 2016, intravenous antibiotics were identified as a major driver of prolonged hospitalization among unhoused patients. For those discharged to MR for outpatient parenteral antimicrobial therapy (OPAT), completion rates reached 87%, saving institutions an average of $25,000 per case.14 These findings position MR as a cost-effective, patient-centered solution for the unhoused.
Despite its benefits, MR faces significant limitations nationwide. To understand these obstacles, it is helpful to examine utilization in New York City (NYC), home to the largest unhoused population in the nation.22 In 2022, NYC Health + Hospitals (H+H) served more than 64,000 unhoused patients, including 1100 who had used MR since its inception.23 At that time, H+H maintained 51 MR beds, increasing to 75 by April 2025. According to the National Institute for Medical Respite Care (NIMRC), three additional New York programs provide a combined 150 beds.24 This shortage of beds reflects similar, although less severe, capacity limitations in other states (Table 1) and underscores significant gaps in funding across the country.
California’s outperformance of New York reflects expanded MR funding under the 2022 CalAIM Medicaid initiative.25 Still, California and New York face persistent shortages, underscoring the issue’s complexity. MR programs rely on Medicaid, federal and state subsidies, hospital budgets, and private funding. Yet the financial beneficiaries are often not the funders — an issue termed the “wrong pocket problem.”26 There is reduced incentive for hospitals and other players to invest when those paying are not the ones reaping direct financial benefits. Other strategies, such as hospital contract incentives and private insurance partnerships, also remain largely underused.
Funding is not the only barrier. As a relatively new public health initiative, MR lacks standardized program and admission requirements. NYC’s Department of Health (DoH), for example, recently advanced standardization. In August 2024, the DoH required programs to provide essential services: 24/7 room and board, including three meals and two snacks daily; care coordination, including transportation, telehealth, and benefit assistance; family support; wellness checks; counseling; and mental health therapy. Staff must be trained in first aid and basic life support (BLS).27 Eligibility criteria were also clarified: patients must be 18 years of age or older with a qualifying medical condition. Patients must be self-directed, able to perform activities of daily living (ADLs), and able to respond appropriately in emergencies, or they must have an established care team. Those posing safety risks are excluded. Therefore, careful assessment of a patient’s functional status may allow teams to coordinate physical and occupational therapy during admission, potentially reducing unnecessary exclusion of some of the most vulnerable patients who might otherwise benefit from medical respite. In addition, because qualifying conditions are not explicitly defined, clinicians retain significant discretion in determining eligibility.
In hospitals, clinicians and social workers initiate referrals by identifying candidates and drafting service plans that outline clinical, social, behavioral, interpersonal, and coordination needs, along with a discharge plan. This information must be submitted 72 hours before discharge.28 MR staff verify the information, meaning that coordination can require at least 4 days and substantial bedside advocacy. Once admitted, in states such as New York that fund MR through Medicaid waivers, patients must be enrolled in Medicaid for reimbursement and may stay up to 90 days per 12-month period.29
Even with standardized guidelines, creating MR programs and referring patients remains complex. This is again emblematic of many other states. A Duke-based review of affiliated hospitals in the Southeast found that only 8% of unhoused patients qualified for MR. Emergency department discharges without respite were declined for insufficient medical need (63%), ongoing acute care needs (16.3%), or leaving against medical advice (6.2%).30 Among inpatients, ineligibility stemmed from lack of medical necessity (42.5%), housed status (13.6%), or psychiatric instability (5.5%).30 Other states impose further barriers; Los Angeles requires active Medi-Cal or Medicare insurance.31 The impact of this procedural complexity is difficult to quantify, although at least two studies have cited low MR bed occupancy as evidence. An Ohio pilot study reported 77% occupancy from 2022 through 2023 despite high demand, and Washington State reported 59 to 60% utilization from 2023 through 2024.32,33 These rates indicate that funding and bed supply are not the sole obstacles to MR utilization.
MR coordination demands time, collaboration, and sustained advocacy. Integrating automated tools into electronic health records, simplifying referral paperwork, and shortening approval time could reduce administrative burden, but such solutions remain uncommon. Therefore, empowering all care team members to initiate MR discussions throughout patient admission may increase referrals. Medical students may be uniquely positioned to support this process, because they often have more time to conduct thorough social and discharge-planning assessments to identify patients who may benefit from MR early in the admission course.
MR is a powerful tool for physicians, social workers, students, and hospital leaders alike to improve care for unhoused patients and the systems that serve them. For patients, MR provides safe discharge planning that supports long-term clinical outcomes, housing stability, and quality of life, aligning with ethical principles of stewardship and nonmaleficence. For hospitals, MR generates cost savings and reduces systemwide strain. However, MR remains a limited resource nationwide, resulting in lengthy referrals and restricted admissions that leave vulnerable patients without access.
First and foremost, funding for MR must increase — a complex problem requiring multidisciplinary collaboration among hospitals, insurers, and relevant government agencies. Second, the referral process needs to be streamlined, an improvement that will take coordination among MR programs, clinicians, and hospitals. For these two steps to come to fruition, however, the extent of the demand for MR must be demonstrated. This third step begins with increased provider awareness of MR and advocacy on behalf of our unhoused patients. By better demonstrating the need for MR, clinicians can strengthen the case for expanded funding and simplified referral pathways. Cities such as New York have already begun scaling programs as a result, adding 25 beds in recognition of MR’s positive effect on outcomes and costs.34 Ultimately, greater clinician engagement can align the interests of unhoused patients, hospital systems, and third-party payers, all of whom benefit from expanded access to MR.