CLINICAL QUESTION: Is a hybrid hospital at home program as safe as traditional inpatient care?
BACKGROUND: Hospital at home (H@H) programs are expanding in the U.S. in an effort to ease hospital overcrowding, reduce costs, and improve the patient experience. Early small studies have suggested H@H models are as safe as traditional inpatient care; most of these studies relied on in-person physician visits. As telemedicine has advanced and expanded, hybrid models with exclusively virtual physician encounters combined with in-home clinical support have emerged, thus far with limited safety data.
STUDY DESIGN: A pragmatic, randomized, controlled, noninferiority trial
SETTING: Clinicians identified eligible patients admitted or in the emergency department being considered for admission at one of three Mayo Clinic hospitals in Arizona, Florida, and Wisconsin between July 10, 2023, and October 31, 2023.
SYNOPSIS: Investigators randomized 1,150 adults requiring acute hospital-level care to either a hybrid H@H model or standard brick-and-mortar inpatient care. In the H@H model, all physician encounters were virtual, supported by in-home nurses and paramedics using digital monitoring tools. The no-crossover cohort, after excluding patients who became ineligible or rejected randomization, had 686 patients (192 intervention, 494 control), while the no-crossover, propensity-score matched cohort included 374 matched pairs admitted to the intervention (187) or control (187) group.
The composite primary outcome of 30-day all-cause mortality or unplanned readmission occurred in 17.3% of H@H patients and 19.8% of brick-and-mortar patients in the randomized no-crossover cohort (OR, 0.85; 95% CI, 0.63 to 1.14; P=.28)and in 15.5% versus 18.7% in the propensity-score matched cohort (OR, 0.80; 95% CI, 0.46 to 1.37, P=0.41), meeting the noninferiority threshold. Thirty-day readmissions and mortality were similar, and no deaths occurred in the H@H group.
Patient-reported comfort was significantly higher in the H@H group (randomized 84% versus 61%, P=.001; propensity-score matched 90% versus 61%, P=.005). Patient satisfaction and perceived safety were high in both settings.
LIMITATIONS: The study was conducted within a single integrated health system, which may limit generalizability. A high rate of patient crossover and post-randomization exclusion reduced study power and introduced potential selection bias. Patient-reported outcomes are subject to selection bias, recall bias, and response bias. Additionally, the underrepresentation of Medicaid and minority patients highlights equity concerns that may affect access to home-based acute care models.
BOTTOM LINE: A hybrid hospital-at-home model using entirely virtual physician visits combined with in-home nursing support was as safe as traditional inpatient care with higher rates of patient comfort, supporting its use to expand hospital capacity safely and enhance patient-centered care.
CITATION: Maniaci MJ, et al. Safety in a hybrid hospital-at-home program versus traditional inpatient care: A pragmatic randomized controlled trial. J Hosp Med. 2025;20(11):1174-1184. doi: 10.1002/jhm.70076.
Dr. Karaaslan is a hospitalist and instructor in medicine at the Columbia University Medical Center in New York.