Acute Hospital Care at Home in The United States: The Early National Experience

BACKGROUND

Hospitalization is the standard of care for acute illness, but hospital care is often expensive, unsafe, and uncomfortable.(1) Acute hospital care at home (AHCaH) provides hospital-level care in patients’ homes as a substitute for brick-and-mortar care. Prior research demonstrates that compared with traditional inpatient hospital care, patients cared for in AHCaH have improved experiences, levels of physical activity, with lower mortality, rates of readmission, and discharge to skilled nursing facilities.(2, 3) However, there are little data describing the experience of AHCaH at the national level. In November 2020, the Centers for Medicare and Medicaid Services issued the AHCaH Waiver, creating a regulatory and payment pathway for hospitals to deliver AHCaH, with over 300 hospitals in 37 states approved.(4, 5)

OBJECTIVE

We report on the early national experience of the AHCaH waiver.

METHODS

Effective July 1, 2022, the National Uniform Billing Committee established occurrence span code 82, identifying patients who received AHCaH. We used 100% Medicare fee-for-service Part A claims reflecting care delivered between July 1, 2022 and June 30, 2023 to identify all patients with an inpatient admission for a medical diagnosis that included span code 82. We report their sociodemographic (directly from claims and extrapolated from 5-digit zip code) and clinical characteristics and clinical outcomes. To estimate patient complexity, we calculated the hierarchical condition category (HCC) score with one-year lookback. To estimate acuity, we calculated case mix index by summing the diagnosis related group (DRG) weights and dividing by number of patients. We report mortality and escalation (returning to the hospital for > one midnight) during hospitalization and skilled nursing facility use, mortality, and readmission 30-days post-discharge. We stratified these outcomes by disabled, dual-eligible, and Black or Latin@ patients. As many programs initially launched to serve patients with COVID, we initially stratified analyses by COVID as primary diagnosis. We did not find substantial differences and so describe findings for all patients. This study was approved by the Centers for Medicare and Medicaid Privacy Board and the Harvard Medical School Institutional Review Committee. Analyses were performed in SAS v7.15.

RESULTS

We identified 5,132 patients with a medical diagnosis who received AHCaH representing 5,551 admissions (out of 5,858 total patients and 6,345 total admissions). Fifty-four percent were female, 85.2% White, 41.7% >=80 years old, 13.8% dual-eligible, 18.1% disabled, and 1.7% lived in a rural area (Table 1). Mean household income was $83,932. AHCaH patients were medically complex: mean HCC score was 3.15 (standard deviation [SD], 2.06), 42.5% had heart failure, 43.3% had chronic obstructive pulmonary disease, 22.1% had cancer, and 16.1% had dementia. The 5 most common discharge diagnoses were heart failure, respiratory infection (including COVID), sepsis, kidney/urinary tract infection, and cellulitis (data not shown). Mean case mix index was 1.31 (SD, 0.49). Mean length of stay was 6.3 days (SD, 4.7). During hospitalization, escalation was 6.2% and mortality was 0.5%. At 30-days post-discharge, mortality was 3.2%, skilled nursing facility use was 2.6%, and readmission was 15.6%. These outcomes were similar in stratified analyses (Table 2).

Table 1.

Patient characteristics who received acute hospital care at home

Characteristic, n (%)	AHCaH (n=5,132)
Sex
Male	2,364 (46.1)
Female	2,768 (53.9)
Age
18–64	391 (7.6)
65–74	1,577 (30.7)
75–84	1,959 (38.3)
85+	1,205 (23.5)
Race and ethnicity
White	4,371 (85.2)
Black	346 (6.7)
Latin@	219 (4.3)
Asian	77 (1.5)
Other/Unknown	119 (2.3)
Mean household income, $ (SD)	83,932 (28,241)
Geography by RUCA
Metropolitan	4,807 (93.8)
Micropolitan	235 (4.6)
Small Town/Rural/Not Coded	89 (1.7)
Reason for Medicare
Old Age	4,136 (80.6)
Disability	930 (18.1)
ESRD	45 (0.9)
Disability & ESRD	21 (0.4)
Dual-Eligible	705 (13.8)
Census Region
Northeast	1,610 (31.4)
Midwest	747 (14.6)
South	2,439 (47.5)
West	335 (6.5)
HCC Score, mean (SD)	3.15 (2.06)
Case Mix Index, mean (SD)	1.31 (0.49)
Comorbidities
Heart failure	2,175 (42.5)
COPD	2,219 (43.3)
Asthma	1,306 (25.5)
Diabetes	2,375 (46.4)
Cancer	1,132 (22.1)
Stroke/TIA	977 (19.1)
Dementia	822 (16.1)
COVID Diagnosis (Primary)	564 (11.0)

HCC Score: Score used by the Centers for Medicare and Medicaid Services as part of a risk-adjustment model that identifies individuals with serious conditions and allows Medicare to project the future annual cost of care. A HCC score of 1.0 indicates average complexity and spending. A higher HCC score represents a patient with more complexity.

Case Mix Index (CMI): Calculated by summing the Medicare Severity-Diagnosis Related Group weight for each discharge and dividing the total by the number of discharges. CMI reflects the diversity, complexity, and resource needs of hospitalized patients. A higher CMI represents more complexity and resource needs.

Abbreviations: AHCaH, acute hospital care at home; COPD, chronic obstruction pulmonary disease; COVID, coronavirus disease; ESRD, end-stage renal disease; HCC, hierarchical condition category; RUCA, rural-urban community area; SD, standard deviation; TIA, transient ischemic attack

Table 2.

Acute Hospital Care at Home outcomes stratified by disadvantaged characteristics

	Overall	Disabled	Dual-eligible	Black or Latin@
n (discharges)	5,551	1,063	796	609
Mortality during admission	0.5%	0.3%	0.1%	0.2%
Mortality 30-days post-discharge	3.2%	2.1%	1.9%	1.5%
Skilled nursing facility use 30-days post-discharge	2.6%	2.4%	2.1%	2.1%
Readmission 30-days post-discharge	15.6%	18.3%	18.8%	18.4%

DISCUSSION

Early national experience providing AHCaH demonstrates a diverse group of medically complex patients received care with low rates of mortality (0.5% during hospitalization and 3.2% at 30 days), escalation (6.2%), skilled nursing facility use (2.6%), and readmission (15.6%). AHCaH was delivered to patients including those who were dual-eligible, disabled, or had dementia, with similar outcomes for socially vulnerable patients.

Our study has limitations. First, our study generalizes only to AHCaH within traditional Medicare. Second, we identified patients using a new span code, which may not identify all AHCaH patients. To verify the code’s validity, we estimated monthly volume with publicly available data (suggesting <3% missingness) and analyzed our own hospital-level data. Second, we lack a comparator. Constructing a comparator for a hospital alternative is challenging. Third, this analysis does not distinguish patients directly admitted to AHCaH from an emergency room versus those transferred to AHCaH after initial treatment in a brick-and-mortar hospital.

The current AHCaH waiver will expire in December 2024. Our data provide preliminary evidence on national uptake and suggest that AHCaH is an important care model to manage acute illness, including among socially vulnerable and medically complex patients. These data should help inform ongoing policy deliberations.

• Data availability

• Protocol: not available

• Statistical code: available to interested readers upon request by contacting David Levine at dmlevine@bwh.harvard.edu

• Data: available with an approved DUA from the Research Data Assistance Center