My hospitalized patient needs help after discharge: Navigating post-acute care options

Abstract

Importance:

Roughly 25-40% of hospitalized adults are discharged to receive post-acute care (PAC) either at home through home health (HH) or in skilled nursing facilities (SNFs), inpatient rehabilitation facilities (IRFs), or long-term acute care hospitals (LTACHs). This Narrative Review provides an overview of various PAC settings to help hospital-based clinicians collaborate most effectively with patients, caregivers, and an interdisciplinary care team to promote high-quality PAC and facilitate better PAC transitions.

Observations:

PAC settings vary in their eligibility requirements and in the intensity and complexity of services they provide. HH provides intermittent in-home support for homebound individuals with skilled needs. SNFs provide housing, daily nursing care, rehabilitative services, and medical supervision, although the medical care provided is often a dramatic step down from hospital-based acute care. IRFs provide intensive rehabilitation, most commonly for individuals with specific diagnoses, such as stroke, hip fracture, or spinal cord injury. LTACHs care for medically complex patients who require prolonged hospital-level care for services such as ventilator weaning or complex wound management. Discharge planning to PAC should be guided by clinical needs, caregiver support, and values and preferences of patients and caregivers; yet it is often opaque and influenced by external factors such as insurance coverage and geography. Clinicians should proactively recognize PAC needs, advocate for specific PAC settings for targeted situations, and assist in selecting an appropriate PAC facility, especially for patients with insurance restrictions. Finally, PAC admissions frequently represent a pivotal time period for individuals with serious illness, during which a subsequent poor prognosis becomes clear. Therefore, PAC stays represent an opportunity for clinicians to revisit prognosis with patients and families, engage in goals of care conversations, and clarify plans for the future.

Conclusions and Relevance:

To provide patient-centered care and help facilitate better transitions for patients and caregivers, hospital-based clinicians have a responsibility to understand the different types of PAC settings and actively participate in discharge planning.

Introduction

Clinicians often care for hospitalized patients who cannot return to their prior level of support in the community due to new functional impairments or complex post-hospital care needs. Following discharge, these patients may require post-acute care (PAC), broadly defined as medical and rehabilitative services intended to help individuals recuperate and rehabilitate. PAC can be provided at home through home health (HH) or in skilled nursing facilities (SNFs), inpatient rehabilitation facilities (IRFs), and long-term acute care hospitals (LTACHs). A key criterion for PAC eligibility is the need for “skilled” nursing and/or rehabilitative services as determined by the treating physician.^1–3 Payers require that these services, due to their nature and inherent complexity, can only be provided by qualified health professionals (e.g., wound care or intravenous infusions by nurses) and are reasonable and necessary for the treatment of a patient’s illness or injury. Yet, clinicians often play a passive role in PAC planning; many report a lack of knowledge around PAC capabilities, quality, and constraints.^4–6

The epidemiology of PAC in the United States is best understood for traditional Medicare (or fee-for-service). Among hospitalized Medicare beneficiaries, roughly 40% were discharged to PAC in 2023: 18% to HH, 17% to SNF, 5% to IRFs, and 1% to LTACHs,⁷ accounting for approximately $60 billion of Medicare spending annually.⁷ PAC accounts for up to three-quarters of total geographic variation in Medicare spending, suggesting that discharge decisions are often driven by local practice norms, rather than patient need, and offering opportunities to improve PAC quality.^8,9

Hospital-based physicians, nurse practitioners, and physician assistants play an important role in PAC discharge planning due to their in-depth understanding of a patient’s complex medical needs. A better understanding of the qualifications and services provided can help clinicians engage in a more helpful role in the PAC discharge planning process. This article provides an overview of PAC settings with the goal of helping clinicians collaborate most effectively with patients, caregivers, and interdisciplinary care teams to promote transition to high-quality PAC.

Methods

Descriptive information and general eligibility for PAC settings were obtained by reviewing publications from the Centers for Medicare & Medicaid Services (CMS) and reports from the Medicare Payment Advisory Commission. We searched PubMed for English-language studies published between January 2010 and October 2025 using the search terms post-acute care, skilled nursing facility, inpatient rehabilitation facility, and long-term acute care hospital. Relevant articles were reviewed, and additional studies were identified through reference list searches. Finally, we drew on the collective expertise of our authorship group across the PAC spectrum, discussions with interdisciplinary colleagues in hospital medicine, geriatrics, and primary care, and input from experts in PAC.

Goals of this review

PAC settings vary in their eligibility requirements, intensity and complexity of services provided, and costs (Table 1).^7,10,11 Here, we present a general summary of the most common types of PAC facilities, followed by a comparison of the supporting evidence for each PAC type. Descriptions of elements of PAC care are based on the benefits covered by traditional Medicare, which generally inform other payers’ coverage policies. Finally, we review best practices for clinicians to actively discuss PAC options with their patients, helping to orchestrate transitions of care to PAC for eligible individuals.

Table 1:

Overview of post-acute care settings

	Home health (HH)	Skilled nursing facility (SNF)	Inpatient rehabilitation facility (IRF)	Long-term acute care hospital (LTACH)
Synonyms	Home health care; home care; home rehabilitation; home health agency; Visiting Nurse Associations or Services (VNA/VNS)	Sub-acute rehabilitation (SAR), short-term rehab, rehab	Inpatient rehabilitation unit; acute rehabilitation unit; rehabilitation hospital	Long-term care hospital; transitional care hospital; specialty hospital
Typical patient profile and services provided	Homebound patients with a need for skilled nursing or therapy (e.g., PT, OT, SLP, wound care, medication management). Can provide part-time aides for personal care (e.g., dressing, bathing) or home services (e.g., light cleaning) on a short-term basis only if a skilled service is also provided.	Patients with short-term skilled nursing or therapy needs following acute illness or surgery (e.g., wound care, intravenous antibiotics, PT, OT, SLP). Most SNFs are dual certified as nursing homes, which provide long-term residential care.	Patients who may benefit from intensive rehabilitation therapy. Facilities typically have at least 60% of their patients with 1 of 13 qualifying diagnoses (e.g., stroke, spinal cord injury, amputation, hip fracture, traumatic brain injury, or major trauma).a Can be located within a hospital as a separate unit or as a free-standing facility.	Patients with complex medical conditions that require extended intensive care (e.g., ventilator support, complex wound care, dialysis, infusions). Can be located within a hospital as a separate unit or as a free-standing facility.
General eligibility requirements b	Need for intermittent skilled nursing or therapy services for homebound patients (e.g., considerable and taxing effort to leave home; need for assistive devices)	Need for daily skilled nursing care or rehabilitation services. For traditional Medicare beneficiaries, individuals must have a qualifying hospital stay of at least 3 consecutive days (days under “observation” status do not count)	Need for ongoing intensive rehabilitation in at least 2 modalities (one of which must be PT or OT) with an ability to participate in 3 hours of therapy per day for 5 days per week, and requires supervision by a rehabilitation physicianc	Extended inpatient care (weeks or months) for individuals who typically survive an ICU stay of ≥3 days with ongoing need for intensive medical therapies (ventilation, dialysis, infusions, complex wound care, TPN)
Frequency of services b	~2–3 visits per week of an hour durationd	~60–90 minutes of therapy daily, which may include PT, OT, and/or SLP. Daily nursing care. Physician oversight may be limited.	≥3 hours of combined therapy daily for 5 days weekly, which may include PT, OT, and/or SLP.c Care is supervised by rehabilitation physicians with at least 3 face-to-face visits in the first week.	Daily multidisciplinary medical and therapy services, including daily physician oversight
Nurse to patient ratio e	1:1	1:15–20	1:5–10	1:5–10
Number of facilities/agencies in the US	~12,000 home health agencies	~15,000 SNFs	~1,200 IRFs	~350 LTACHs
Average length of stay f	58 days	29 days	13 days	28 days
Rate of successful community discharge g	81%	51%	67%	23%
Cost for traditional Medicare beneficiaries	No copayments or deductible	Days 1-20: $0 Days 21-100: co-payment ($209.50 each day in 2025) Days 101 and beyond: Beneficiary pays all costs	Days 1-60: $0 after meeting Part A deductible Days 61-90: co-payment ($419 each day in 2025) Days 91 and beyond: higher co-payment ($838 each day in 2025)h	Days 1-60: $0 after meeting Part A deductible Days 61-90: co-payment ($419 each day in 2025) Days 91 and beyond: higher co-payment ($838 each day in 2025)h
Estimated cost to traditional Medicare program i	$100 - $300 per in-person visit	$500 - $1500 per day	$1000 - $2000 per day	$1500 - $3000 per day

Abbreviations: ICU, intensive care unit; OT, occupational therapy; PT, physical therapy; SLP, speech-language pathology; TPN, total parenteral nutrition; US, United States

^aTo be classified for payment under Medicare’s IRF prospective payment system, at least 60% of an IRFs total inpatient population must require treatment for 1 or more of 13 conditions, including stroke, spinal cord injury, congenital deformity, amputation, major multiple trauma, hip fracture, brain injury, neurological disorders (e.g., multiple sclerosis), burns, knee or hip joint replacement meeting specific criteria (e.g., bilateral knee replacement, age 85 or older), and various conditions if significant functional impairments (e.g., rheumatologic conditions such as polyarticular rheumatoid arthritis, systemic vasculitides with joint inflammation, and severe osteoarthritis involving multiple major weight bearing joints).

^bInformation about general eligibility requirements and services provided was collected from the Medicare Benefit Policy Manual and most directly applies to traditional Medicare beneficiaries.

^cFor determining admission eligibility to IRFs, intensity of therapy services could also be demonstrated by participation in therapy for at least 15 hours over a 7-day period in some circumstances, whereby some days are less than 3 hours of therapy (e.g., temporary fatigue due to radiation or chemotherapy).

^dAmong traditional Medicare beneficiaries who received home health following a hospitalization in 2021, there were an average of 9 skilled nursing visits, 9.8 therapy visits, 0.1 medical social work visits, and 1.0 home health aide visits per home health stay. The average stay among individuals receiving post-hospital home health was 57.9 days in 2021.

^eThese numbers are based on general experience. The Centers for Medicare & Medicaid Services will typically report staffing numbers for various facilities in hours per resident day rather than staffing ratios. In SNFs, nurses can either be registered nurses or licensed practical/vocational nurses. In IRFs and LTACHs, nurses are generally registered nurses. Nursing staff ratio for ICU-level patients in LTACHs is generally 1:2.

^fThe average length of stay for various settings was obtained from the Medicare Payment Advisory Commission reports. For home health, the average length of stay was 57.9 days for post-hospital home health stays and 93.9 days for community-admitted home health stays in 2021.

^gThe “successful discharge to the community” metric is a risk-adjusted rate of traditional beneficiaries who were discharged to the community after a PAC stay and did not have an unplanned re-hospitalization and/or die in the 31 days following discharge. The numbers for SNF, IRF, and HH reflect data from 2022–2023. The number for LTACH reflects data from 2020 as outlined in the 2022 Medicare Payment Advisory Commission report.

^hFor traditional Medicare beneficiaries, the payment structure for IRFs and LTACHs is the same as that of acute care hospitals. For IRFs and LTACHs, payment in 2025 for days 91 and beyond is $838 each day for each lifetime reserve day. An individual has up to a maximum of 60 reserve days over their lifetime. Additionally, a beneficiary does not have to pay a deductible for IRF or LTACH care if Medicare has already been charged a deductible for care in the hospitalization within the same benefit period.

ⁱAverage cost per day was a rough estimate related to the base rate reimbursement and average length of stay for traditional Medicare beneficiaries based on the Medicare Payment Advisory Commission Payment Basics overview documents. For HH, the average traditional Medicare payment per in-person visit was $237 in 2023. For SNFs, the Medicare daily base rate in 2025 is around $500. IRFs receive a base payment of $18,907 for fiscal year 2025. LTACHs receive a base payment of $49,383.26 for fiscal year 2025. Payments are subsequently adjusted based on geographic factors (e.g., area wages) and case mix.

Different types of PAC settings

Home Health

For hospitalized individuals with intermittent skilled needs after discharge and adequate support at home, HH provides nursing services (e.g., simple wound care or medication management) and short-term rehabilitative services (e.g., physical/occupational therapy, speech-language pathology) beginning within 48 hours of discharge for about 2-3 times weekly for up to 60 days per episode (divided into two 30-day periods). On average, beneficiaries receive 8-9 visits total every 30 days.⁷ To qualify, the recipient must be “homebound” (defined as requiring a considerable and taxing effort with significant assistance to leave home), have skilled nursing or rehabilitation needs (such as wound care or gait training), and be under a physician’s plan of care (which includes their primary care physician). If indicated, HH can be extended with a repeat assessment every 60 days. Nearly all patients live in a region with access to HH.⁷

If a skilled service is provided, HH can also include medical social services (e.g., counseling) and intermittent HH aide services (e.g., ADL assistance). HH aides may provide personal care services (e.g., light cleaning) during visits when they provide health-related services. However, due to its intermittent nature, HH is often insufficient to meet even short-term caregiving needs of individuals with ADL dependence or severe cognitive impairment.

Skilled Nursing Facilities

For hospitalized individuals with daily skilled needs (e.g., daily wound care, gait training), SNFs provide housing, daily nursing care, supervised medication administration, rehabilitative services (~60-90 minutes of therapy daily),^12,13 ADL support, and medical supervision. Importantly, traditional Medicare beneficiaries must have a preceding hospitalization of at least 3 consecutive days, not inclusive of days under Medicare’s “observation” status.¹⁴ However, patients remain eligible for SNF care up to 30 days after discharge, allowing a trial of HH if appropriate.²

The medical care provided is significantly lower-intensity than what is typically provided in an acute care hospital: one nurse is assigned to 15-20 patients, and only 70% of initial physician evaluations occur within the first four days.¹⁵ Thus, patients should have a stable medical plan for recovery akin to patients discharged home (e.g., not requiring active dose titration or close physician monitoring), and clinicians should plan to provide a 3-day supply of scheduled medications (e.g., opioids) during the transition. SNF-level care is also indicated to maintain a patient’s current condition or prevent further deterioration, and cannot be denied based on the absence of potential for improvement.¹⁶ SNF-level care also cannot be denied based on certain medical conditions or treatment needs, such as patients with opioid use disorder who have experienced inequities due to stigmatizing beliefs and barriers in medication access (e.g., methadone).^17,18

A Medicare reimbursement change in 2019 incentivizes SNFs to accept a wider variety of patients with needs beyond rehabilitation who require more intensive but stable medical care.^10,12,19 For example, SNFs may now accommodate patients with more complex needs, including total parenteral nutrition, noninvasive ventilatory support for chronic respiratory failure, tracheostomy care, or frequent transportation for dialysis or radiation therapy.¹⁹ In some regions, SNFs may even accept patients for ventilation weaning in the subacute phase of respiratory failure.²⁰

Inpatient Rehabilitation Facilities

IRFs provide more intensive therapy than SNFs, with an emphasis on specific diagnoses (e.g., stroke, hip fracture, or spinal cord injury). Based on CMS medical necessity criteria, the expectation is for patients admitted to IRFs to significantly benefit from intensive rehabilitation programs, making measurable improvements within a reasonable time duration.¹ This typically involves 3 hours of combined therapy per day for 5 days per week (“3-hour rule”).

In contrast to SNFs, IRFs are more often separate units within a hospital, have lower nurse-to-patient ratios (e.g., one nurse to 5-10 patients), and medical care is supervised by a non-hospital-based physician with specialized training in inpatient rehabilitation (e.g., physiatrist). Physicians are required to assess new patients within 24 hours and at least 3 times in the first week.¹ IRF lengths of stay are typically ≤2 weeks.

Long-term Acute Care Hospitals

LTACHs care for medically complex patients who require prolonged hospital-level care. While there are no explicit eligibility criteria, due to Medicare reimbursement policies, LTACHs typically accept patients who have survived a critical illness with an intensive care unit stay of ≥3 days and maintain an average length of stay of 25 days.²¹ While best known for ventilation weaning, only about 25% of patients are on prolonged mechanical ventilation.^22,23 LTACHs also manage patients with complex wound care needs, dialysis, multi-organ failure, and/or multiple intravenous infusions. LTACHs are more akin to step-down units in hospitals with comparable nursing ratios and daily physician care, differing in their greater focus on the recovery phase of a prolonged illness involving multidisciplinary therapies.

Clinicians’ role in the PAC selection process

The PAC discharge process is often rushed due to external pressures to expedite discharges, ill-informed due to a lack of knowledge and clear frameworks for decision-making, and disempowering to patients and caregivers who feel left out and unprepared for the hospital-to-PAC transition.^6,24–28 To optimize the process, discharge planning should be an interdisciplinary team-based approach that leverages each team member’s unique skillsets and expertise and incorporates the values and preferences of patients and caregivers.^29,30 Traditionally, clinicians often assume a passive role in this process, in part due to lack of a knowledge base about PAC.^4,6,24,26 However, clinicians can improve PAC decision planning by synthesizing their knowledge of a patient’s medical needs and prognosis with the assessments of a patient’s rehabilitative and social needs from therapists, case managers, social workers, and nurses. Providing guidance to patients and caregivers around PAC decisions can be additionally challenging^31,32 because of differing values and preferences among patients and caregivers (e.g., strong desire to avoid admission to a facility),³³ quality and availability of PAC services,³⁴ and insurance coverage determinations.^10,11 Herein, we outline key considerations for selecting the most appropriate PAC setting (Figure 1).

Figure 1:

Discharge planning for hospitalized patients with increased care needs

Abbreviations: ALF, assisted living facility; IRF, inpatient rehabilitation facility; IV, intravenous; LTACH, long-term acute care hospital; OT, occupational therapy; PT, physical therapy; SLP, speech-language pathology; SNF, skilled nursing facility; TBI, traumatic brain injury

a Skilled nursing and/or rehabilitative services are generally defined as services requiring the skills of qualified health professionals such as registered nurses, physical therapists, occupational therapists, and speech-language pathologists. The services generally involve a level of complexity that can be performed safely and/or effectively only by or under the general supervision of qualified health professionals. This would include, for example, wound care or intravenous medication administration by nurses or gait training by physical therapists following a total hip arthroplasty.

b Medicare does not pay for the cost of long-term nursing homes or residential care facilities (e.g., assisted living facilities or board and care homes). Due to the oftentimes short length of stay within hospitals, individuals are infrequently discharged directly from the hospital to one of these facilities.

c The Centers for Medicare & Medicaid Services has issued direction that claim denials for IRFs should not be based solely on a threshold of therapy time not being met (i.e., the “three hour rule” should not be solely used to determine eligibility). Clinical reviews should be based on the individual circumstances of each case.

d These numbers are for traditional Medicare beneficiaries. In 2022, 41.6% of hospitalized traditional Medicare beneficiaries were discharged to a post-acute setting. Among all inpatient discharges, 18.6% were discharged to home with home health, 17.4% were discharged to SNFs, 4.7% were discharged to IRFs, and 0.8% were discharged to LTACHs.

Recognizing the need for PAC

Clinicians should recognize PAC needs as early as possible to engage in discussions about discharge preferences and planning, while remaining flexible to change course if sufficient recovery enables return to the community. Given the complexity of factors involved (Figure 2), few validated decision aids or tools exist for predicting PAC discharge.³⁵ The Activity Measure for Post-Acute Care (AM-PAC) “6-Clicks” is the most widely disseminated measure to assess independence. While various cut-points may help predict home vs. PAC discharge, AM-PAC may have less utility in predicting SNF vs. IRF discharge.^35–38

Figure 2:

Factors influencing discharge planning to post-acute care in the United States

Abbreviations: IRF, inpatient rehabilitation facility; PAC, post-acute care; SNF, skilled nursing facility; TPN, total parenteral nutrition

Comparison of overlapping PAC settings

Given overlap in services provided, a patient may be eligible for multiple PAC settings. Existing evidence comparing outcomes among PAC settings is heterogeneous and limited to observational studies (Table 2).^39–49 However, clinicians may advocate for specific PAC settings for targeted conditions (e.g., IRFs for post-stroke rehabilitation and LTACHs for prolonged mechanical ventilation). Given that the “right” PAC setting is highly individualized, clinicians should avoid interjecting their own values (e.g., patient safety) ahead of patient values (e.g., independence).⁵⁰ Instead, hospital-based clinicians should elicit the goals and priorities of patients and caregivers, discuss tradeoffs, and provide education around the capabilities and constraints of PAC.⁵¹ For patients whose PAC needs can be met across different settings, we recommend the less intensive and least restrictive PAC setting when evidence suggests equivalent outcomes (e.g., HH vs. SNF following elective joint replacement).^41,49 Below, we outline common scenarios of PAC overlap.

Table 2:

Representative studies comparing outcomes between post-acute care settings

Setting	Reference	Specific comparison	Methods	Key findings
HH vs. SNF	Werner et al. (2019)39	Any discharge to HH vs. SNF	Medicare beneficiaries from 2010-2016 (n=17,235,854 discharges). Instrumental variable of distance to nearest HHA or SNF	• HH associated with 5.6% higher absolute 30-day re-admission rate vs SNF discharge (95% CI 0.8-10.3%) • No difference in mortality or functional status • Lower Medicare payment for HH
	Burke et al. (2021)42	HH vs. SNF among individuals with dementia	Medicare beneficiaries from 2015-2016 (n=977,946). Instrumental variable of distance to nearest HHA or SNF	• No differences in 30-day readmission and mortality • 70% of individuals with dementia were discharged to SNF
HH vs. SNF/IRF for lower extremity joint replacement	Burke et al. (2020)41	Home vs. institutional PAC (most commonly SNF) for lower extremity joint replacement	Data from Pennsylvania between 2011-2013 and 2016-2018 (n=189,949). Difference-in-differences approach with propensity score matching to exploit differences in financial incentives during the study time period	• Matched patients discharged home had lower 30-day readmission (difference −2.9%; 95% CI −4.2% to −1.6%) and 90-day readmission (difference −3.9%; 95% CI −5.8% to −20%) compared to those discharged to institutional PAC • No measures of function and mobility
	Dummit et al. (2016)44	Home vs. institutional PAC (most commonly SNF) for lower extremity joint replacement	Claims data from 176 hospitals that participated in the Bundled Payments for Care Improvement (BPCI) initiative from 2011-2015 (n=29,441 episodes in baseline period and n=31,700 episodes in intervention period). Difference-in-differences approach	• Decline in institutional PAC use for the BPCI population was 5.7% greater than the comparison population without differences in quality measures (30 and 90-day unplanned readmissions, emergency department visits, and mortality) • No measures of function and mobility
IRF vs. SNF for stroke/hip fracture	Hong et al. (2019)40	IRF vs. SNF among individuals hospitalized for stroke	Medicare data from 2012–2014 (n=99,185). Multivariable regression, propensity score, and instrumental variable analyses (multiple IVs including distance to nearest IRF vs. SNF and number of stroke patients discharged to IRF in the region).	• Mean difference in functional scores (scored from 0-100) was higher for patients at IRF than SNF (5-10 points higher for mobility and 8-12 points higher for self-care, where a 10-point difference represents the difference between maximal assistance and needing supervision)
	Simmonds et al. (2022)61,62	IRF vs. SNF among individuals hospitalized for stroke	Medicare beneficiaries from 2011–2024 (n=60,529). 1:1 matched propensity score analysis	• Compared to SNF, IRF discharge was associated with an increased rate of successful community discharge (absolute difference 21% (95% CI 20–22%)) and lower 1-year all-cause mortality (absolute difference −11% (95% CI −12% to −11%)). • No measures of function and mobility
	Lake et al. (2025)48	IRF vs. SNF among individuals hospitalized for stroke and hip fracture	Traditional Medicare beneficiaries admitted to 55 hospitals that closed their IRF units between 2009 to 2017 (n=10,761 for stroke cohort; n=13,963 for hip fracture cohort). Quasi-natural experimental design with cross-temporal propensity score matching and differences-in-differences model	• Relative to IRF, SNF discharge was not associated with a significant difference in hospital readmission rates or successful community discharge rates • Relative to IRF, SNF discharge was associated with an increase in 90-day mortality for stroke (+6.5%, 95% CI 1.5-11.4%) and hip fracture (+5.8%, 95% CI 2.5-9.0%) patients from baseline rates of 9.2% and 6.0%, respectively • No measures of self-care and mobility
LTACH discharge	Kahn et al. (2013)45	Transfer to LTACH vs. all other options	Medicare beneficiaries from 2002-2006 (n=234,799). Instrumental variable analysis using distance to the nearest LTACH and number of LTACHs in the region.	• Patients transferred to LTACHs had similar survival, lower hospital-related costs, and higher total Medicare payments vs patients not transferred.
	Einav et al. (2018)47	Transfer to LTACH vs. all other options	Medicare beneficiaries from 1998-2014. Instrumental variable analysis using LTACH entry into a regional market	• After a new LTACH enters a market, there was a 9% increase in the probability of LTACH discharge (tripling of transfers) and a decrease in hospital length of stay by 9 days, mostly replacing SNF discharge. • For patients newly transferred to LTACH who would not have been prior to market entry: no difference in 90-day mortality or returning home within 90 days, with an increase in out-of-pocket costs by $2,420, and total spending by $32,003.

Abbreviations: CI, confidence interval; HH, home health; HHA, home health agency; IRF, inpatient rehabilitation facility; LTACH, long-term acute care hospital; SNF, skilled nursing facility

HH vs SNF

The most common scenario of PAC overlap involves the decision to return home with HH versus SNF admission.⁴⁹ Patients often prefer home-based PAC but may not have enough support. Caregivers favor SNFs when caregiving demands become excessive or when facing socioeconomic challenges.³³ Clinicians often use SNFs as “safety nets” as they are perceived to be safer.⁶ While SNF admissions may decrease rehospitalizations versus HH, they are more costly and may not improve functioning.³⁹

Individuals with cognitive impairment may recover better in a supportive home environment due to difficulties with the hospital-to-SNF transition.^52,53 This may be due to multiple reasons, including more time with family and/or friends which promotes orientation and maintenance of familiar routines. Given that mortality and readmissions are similar for individuals with dementia discharged to SNF vs. HH, returning home may be advantageous with appropriate supports in place.⁵⁴

Other scenarios favoring home discharge with adequate support involve the administration of prolonged intravenous antibiotics and enteral feedings, where HH nurses can provide intermittent teaching. Clinicians should set expectations pertaining to HH as described above, including limited assistance with ADLs, assurance of adequate home supports, potential for caregiver burden, and that post-discharge coordination will need to involve the patient’s primary care provider.⁵⁵

IRF vs SNF

For patients with predominantly intensive rehabilitative needs, the choice between IRF and SNF care is challenging. In the absence of high-quality studies, it is unclear which patients benefit the most from IRF-level care, which is costlier than SNFs due to more intensive therapies.^48,56–59 Stroke guidelines issue a Class I, Level of Evidence B recommendation to IRF care for stroke survivors who qualify for IRF services,^56,60 supported by observational studies showing greater successful community discharge and improved functioning.^61–63 Other guidelines related to spinal cord injury and traumatic brain injury broadly endorse post-acute care rehabilitation by specialized multidisciplinary teams, preferably in IRFs.^64–66

Adherence to the 3-hour rule is not necessarily associated with better outcomes, and Medicare contractors cannot deny reimbursement based solely upon any threshold of therapy time.^67,68 Therefore, for patients anticipated to benefit from IRFs, clinicians should collaborate with therapists to document rehabilitation potential and therapy indications.

LTACH vs SNF

Beyond mechanical ventilation weaning, the best available evidence supports LTACH care for survivors of critical illness with multiorgan involvement requiring ongoing complex intensive treatments for weeks to months.⁴⁶ Patients who were functionally independent and cognitively intact prior to hospitalization have superior outcomes post-LTACH stay to those who have functional and cognitive impairment.⁶⁹ Otherwise, LTACH usage may be driven more by local hospital and regional practice norms and differences in the availability of specialized expertise at SNFs, rather than patient illness severity or clinical complexity.²²

Providing guidance to patients on selecting PAC

After deciding on a specific PAC setting, patients and caregivers report that they receive little guidance when selecting a facility or HH agency, often merely receiving a list of names in alphabetical order.^28,70 Many select facilities based on prior experience, internet searches, personal referrals, or those nearest to their homes. Despite reporting that they would be willing to choose more distant but higher-quality facilities,^28,70 the evidence shows that many patients choose lower-quality PAC options by publicly available quality metrics from among their available choices.^25,71

One mistaken assumption is that a patient’s interdisciplinary team cannot legally recommend specific PAC providers. However, interdisciplinary teams can assist patients in selecting the highest-quality facility.⁷¹ Clinicians can direct them to CMS Care Compare, an online tool available in English or Spanish that displays quality metrics and ratings for all PAC providers, which can be printed and given to patients if desired for those without internet.⁷² This tool includes information on quality measures, such as the community discharge rate, and includes the Five-Star Quality Rating System for patients to compare HH agencies and SNFs on health inspections, staffing, and quality measures.^72,73 The rating system has important limitations, such as not including measures of patient satisfaction.⁷⁴ Still, interdisciplinary teams can improve discharge planning by helping patients and caregivers use and interpret the ratings to make informed decisions.

Selection of PAC for patients with insurance restrictions

Insurance coverage determinations greatly influence discharge planning. Unlike traditional Medicare, many other payers—including Medicare Advantage (MA), Medicaid managed care, and commercial insurance—actively “manage” PAC utilization by using prior authorization requirements and covering narrower facility networks,^75–78 resulting in longer hospital stays, shorter PAC durations, and lower use of more expensive PAC.^79–85 While effective at deterring overuse of unnecessary PAC care, these strategies can create friction in selecting the most clinically appropriate setting. Although more restrictive PAC use does not translate, on average, into worse healthcare outcomes, advocacy from the interdisciplinary team is crucial to facilitate more intensive PAC if clinically appropriate on a case-by-case basis.^{80–84,86,87}

Over half of Medicare beneficiaries are now enrolled in MA plans.⁸⁸ Medicare pays MA plan insurers a capitated amount to cover each beneficiary’s care, which incentivizes active management of the most expensive utilization, including PAC. MA plans came under increased federal scrutiny for their use of prior authorization, with denial rates among three of the largest MA plans in 2022 reaching 20% for SNFs, 70% for IRFs, and 85% for LTACHs.⁷⁵ The most commonly cited reasons to deny inpatient PAC, which is only publicly reported by CMS for second-level appeals, include intermittent instead of daily skilled needs for SNFs and lack of perceived rehabilitation and medical necessity for IRFs and LTACHs, respectively.⁸⁹ If indicated, all interdisciplinary team members should provide explicit documentation of the specific type of PAC recommended. If the prior authorization request is denied, clinicians should participate in appeals via peer-to-peer conversations. To support advocacy during these conversations, clinicians should be aware that Medicare issued a 2024 rule clarifying clinical eligibility to ensure “people with MA receive access to the same medically necessary care they would receive in traditional Medicare.”^90,91

Finally, while less is known about other payers, many of the same principles apply. However, PAC options may be very limited for Medicaid-only enrollees, which may prolong hospitalization while awaiting discharge to an approved facility.⁹²

Best practices for facilitating PAC transitions

Hospital-to-PAC transitions are prone to errors, in part due to the lack of accuracy and timeliness in discharge communication.^4,24,93–95 This contributes to patient care delays, adverse outcomes, and patient dissatisfaction. Programs exist to improve transitions (e.g., hospital-facility videoconferences, information exchange, and pharmacist medication reviews), but are not widely adopted due to many barriers.^{29,30,96–103} Clinicians can help facilitate better transitions through discharge summaries that provide accurate information for medication reconciliation (e.g., antibiotic durations, adjustments to anticoagulants, and other key medication changes), post-hospital medical needs (e.g., wound care, lab-work), specialist follow-ups, functional and mental status at time of discharge, and establish goals of care.^{98,102,104,105} This is particularly important for SNF discharges, as SNF clinicians may not be able to assess patients for several days. A hospital-facility “doc-to-doc” conversation also improves continuity and reduces adverse events.¹⁰⁶

PAC represents a pivotal time period

While its goals are often preserving and restoring health, PAC admissions are frequently a pivotal time period for individuals with serious illness. Many cycle between hospitals and PAC in a period of rapidly declining health.¹⁰⁷ For example, almost one-third of older adults receive care in SNFs for PAC in the last 6 months of life.¹⁰⁸ Among traditional Medicare beneficiaries discharged to SNFs, only 50% of community-dwelling patients successfully returned home, 20% are re-admitted to the hospital within 30 days, and 20% die within 6 months.⁷ Most long-term residential nursing home admissions (an outcome many older adults wish to avoid) are preceded by a hospitalization and short-stay SNF admission.¹⁰⁹ Even among individuals who return home, many experience prolonged functional dependence, such that only half are independently bathing even after a subsequent HH episode.¹¹⁰ Among patients admitted to LTACHs, the median survival is 8 months, and 37% die in an inpatient setting without ever returning home.¹¹¹

Therefore, PAC planning represents a window of opportunity for clinicians to revisit prognosis and engage in goals of care conversations.^112,113 These discussions are essential to establish the patient’s goals and preferences, which along with the prognosis for recovery, can strongly influence the choice of PAC settings, need for advanced directives, and future care planning. To estimate prognosis, clinicians can use several tools to augment their clinical judgment. For example, one approach is use of the surprise question (“Would you be surprised if this person died within the next year?”), a simple way to identify patients at higher 1-year mortality (sensitivity 72%, specificity 65% for hospitalized patients).¹¹⁴ Another approach involves using clinical prediction models.^115–117 Web-based tools, such as ePrognosis, can estimate 1-year survival among hospitalized adults.^118–121 More focused models have been developed among individuals discharged to SNF or with HH.^122–124

Another important consideration is prior health and functional status. For example, patients who were functionally independent and cognitively intact prior to LTACH hospitalization have higher survival and potential for recovery compared with those with pre-existing severe cognitive and/or functional impairment (58% vs 70% two-year mortality).⁶⁹ Similarly, cognitive dysfunction on admission to SNF strongly predicts adverse outcomes and mortality.⁴²

Prognosis discussions prior to a PAC admission can help set expectations around potential for recovery, guide referrals to palliative care, and unmask indications for hospice. Patients and caregivers often receive overly optimistic prognostic estimates.^4,125 Therefore, clinicians should set realistic expectations for recovery and avoid blanket statements, like “you’ll get ‘rehab’ for a couple of weeks to get stronger before you return home.” A variety of serious illness communication guides are available to assist with these discussions,^126–128 such as REMAP (Reframe, Expect emotion, Map out patient goals, Align with goals, and Propose a plan).¹²⁹

Given that prognosis may remain unclear at hospital discharge and that the PAC time period can serve as a pivotal time for clarification, opting for a time-limited trial period in a SNF or LTACH may be reasonable to assess recovery potential with a plan to revisit goals-of-care discussions during the PAC stay.¹³⁰ However, elicitation of patients’ values and goals in the hospital may reveal that hospice, either at home or in a facility, is the most appropriate PAC option. It is important that clinicians be aware that the Medicare hospice benefit provides little support for patients needing ADL assistance (e.g., often <2 hours of home aide services weekly).^7,131 Additionally, the Medicare hospice benefit does not cover any facility’s housing costs (e.g., nursing home) unless the patient also has Medicaid. Therefore, for patients with inadequate home support for hospice, a short-stay SNF admission may be the only Medicare-covered option, as patients are not allowed to use their Medicare hospice benefit and SNF benefit concurrently.¹³⁰ This may allow patients and caregivers time to logistically set up home resources prior to a home hospice transition.

Limitations and future directions

Our review is most informed by PAC coverage, utilization, and outcomes in traditional Medicare. Better transparency and research are needed to help inform decisions for individuals enrolled in MA, commercial insurance, and Medicaid plans. Our review is also focused on care in the United States because of major differences in the types of PAC provided and eligibility requirements in other countries.^132,133 Conceptually, many of the principles of this review apply, but operationally may differ. Additionally, the lack of high-quality comparative effectiveness studies limits evidence-based decisions around which PAC setting is most appropriate for every patient. Clinical trials are needed, such as those ongoing for HH.¹³⁴ Lastly, our pragmatic PAC considerations may change based on evolving regulations and insurance coverage determinations.

We recognize that hospitalization frequently unmasks the need for increased long-term services and supports (e.g., assistance with activities of daily living (ADLs), such as bathing and dressing). Our review focuses on short-term PAC for skilled rehabilitative needs following hospitalization, which differs from long-term services and supports (e.g., home and community-based services, nursing homes), medical respite programs (e.g., patients with dementia or advanced cancer, or those experiencing homelessness), and transitional care programs.^135–138

Conclusion

PAC use is rising, in part due to increased patient complexity and pressure to shorten hospital stays. To deliver high-quality person-centered care, hospital-based clinicians should understand the distinctions between PAC settings, take an active role in the selection of PAC, and engage in serious illness discussions with high-risk patients to facilitate better transitions for patients and caregivers.