The Hospital Comes Home: Taiwan’s Pilot Trial Brings Acute Wards Into Living Rooms

For more than a century, the hospital ward has been the default destination for anyone seriously ill. But a growing body of research suggests that for many older patients, the hospital itself may be the riskiest part of the illness. Now, a pilot trial from Taiwan offers some of the strongest evidence yet that hospital-level acute care can be delivered safely and effectively in patients’ own homes, even when the patient never passes through a hospital door. The study, published in the Journal of General Internal Medicine, tested a model known as direct home admission, in which frail older adults with acute infections were admitted to a virtual hospital ward operated out of their living rooms, bypassing the emergency department entirely.

The trial was conducted between October 2022 and July 2023 and enrolled 144 homebound older adults who required hospital-level treatment for acute infections. Rather than a single medical center, the study spanned five hospitals and thirteen community clinics across Taiwan, reflecting a deliberate effort to test the model within the country’s existing Integrated Home-Based Medical Care system rather than in a purpose-built research environment. Participants were randomly allocated in a non-randomized controlled design: 83 received hospital-at-home care, while 61 received standard inpatient care in a brick-and-mortar hospital. The researchers were careful to note that this was a pilot study, designed to generate the preliminary evidence needed before Taiwan scales the model nationally.

The population studied is precisely the group for whom hospitalization carries the greatest hazards. Participants were generally frail, chronically ill, and functionally dependent, meaning they were homebound even before they became acutely unwell. For such patients, a trip to the emergency department can trigger a cascade of harms: delirium from the unfamiliar environment, falls, pressure injuries, loss of muscle mass from bed rest, exposure to multidrug-resistant organisms, and the disorientation of being cared for by rotating teams of strangers. Hospital-at-home models, which have been pioneered in the United States, Israel, Australia, and elsewhere, aim to deliver the clinical intensity of a ward while sparing patients these collateral damages.

The Taiwanese model centered on direct home admission is distinctive because most patients were admitted straight from the community, identified by the family physicians and home-care nurses who already knew them, rather than being triaged through an emergency department. A small proportion did enter through an emergency department pathway that cleared them for home treatment. Once enrolled, patients in the hospital-at-home group received a package of interventions that mirrors what a medical ward provides: interprofessional visits from physicians and nurses, intravenous pharmacotherapy administered at the bedside, point-of-care diagnostics such as blood tests performed on site, and remote patient monitoring to track vital signs between visits. The control group received conventional inpatient care.

The primary outcomes focused on the questions any clinician would ask before sending a frail patient home with an intravenous line. Could the illness be resolved without hospital transfer? Would patients deteriorate and need rescue? Would they die during the treatment episode or within 30 days? The answers were reassuring. Remission or complete treatment was achieved in 86.7 percent of the hospital-at-home group compared with 95.1 percent of the hospitalized group, a difference that was not statistically significant. Death during the treatment episode was rare in both arms, at 2.4 percent versus 3.3 percent. Nine of the 83 hospital-at-home patients, or 10.8 percent, required clinical escalation to a hospital, a figure the researchers viewed as an expected safety valve rather than a failure of the model.

Where the home model clearly outperformed the ward was in efficiency and experience. The hospital-at-home group had a significantly shorter length of stay, averaging 9.72 days compared with 11.85 days for conventional inpatients, a difference that reached statistical significance. Shorter treatment episodes matter for more than convenience: they free up scarce hospital beds, reduce the exposure of immunocompromised patients to hospital-acquired pathogens, and lower the cumulative cost of an episode of care. In a health system like Taiwan’s, where an aging population is straining inpatient capacity, shaving two days off the average admission for this population could translate into substantial system-level relief.

The patient experience data were equally striking. Patients and caregivers in the home group reported significantly higher satisfaction and convenience, and significantly greater willingness to choose the model again, compared with those treated in hospitals. This finding aligns with a consistent pattern in the international literature. Qualitative evaluations of hospital-at-home programs elsewhere have found that patients value sleeping in their own beds, eating their own food, and having family nearby without visiting-hour restrictions, while caregivers appreciate being integrated into the care team rather than relegated to visitors. For functionally dependent older adults, avoiding the transfer to an unfamiliar ward may also protect cognition and preserve the routines that anchor daily life.

The technological backbone of the model deserves attention, because it is what makes ward-level care in a living room clinically credible. Intravenous antibiotics and fluids, historically the defining feature of inpatient care, can now be administered safely at home by visiting nurses with standardized protocols. Point-of-care laboratory diagnostics allow clinicians to monitor inflammatory markers, renal function, and other parameters without transporting the patient. Remote patient monitoring devices stream vital signs to the clinical team, enabling early detection of deterioration. Together, these tools compress the distance between the patient and the hospital to the length of a data connection, while the interprofessional home visits preserve the hands-on assessment that remote care alone cannot provide.

The study’s design has limitations that the authors themselves acknowledge. It was a non-randomized pilot, so patients were not assigned by chance, and unmeasured differences between the groups could have influenced the outcomes. The sample size of 144 was sufficient to detect large signals but not subtle ones, and the confidence intervals around the mortality and remission comparisons are correspondingly wide. Selection is also a consideration: only patients judged suitable for home treatment by their clinicians were offered the model, so the results apply to a carefully screened population rather than to all acutely ill older adults. Still, the trial adhered to the CONSORT extension for non-randomized studies, was approved by an institutional review board, and obtained written informed consent from all participants or their legal representatives, lending rigor to what is explicitly a proof-of-concept exercise.

What makes the Taiwanese trial consequential is its policy trajectory. The findings directly informed the development of a national acute care at home program, meaning this was not an academic curiosity but the evidence base for a system-wide rollout. Taiwan is unusually well positioned for such a model because of its mature home-based medical care infrastructure, in which community clinics and home-care nursing agencies already deliver longitudinal care to homebound patients. The direct home admission pathway leverages that existing network: the clinician who knows the patient best becomes the admission gateway, and the home-care team becomes the ward staff. If the national program performs as the pilot suggests, Taiwan will join a small group of health systems, including the United States under its Medicare Acute Hospital Care at Home program, that have moved hospital-at-home from experiment to standard option. For the growing number of countries confronting aging populations and hospital capacity crises, the message from this trial is provocative and practical: for a substantial share of frail older patients, the safest hospital bed may be the one they already sleep in.

Subject of Research: Hospital-at-home acute care via direct home admission for frail older adults in Taiwan

Article Title: Hospital-Level Acute Care at Home via Direct Home Admission in Taiwan: A Non-randomized Controlled Pilot Trial Before National Implementation

Article References: Wang, Y.-C., Liao, J.-Y., Yang, F.-C., Huang, S.-W., Tseng, W.-Z., Chang, H.-C., Huang, C.-H., Yu, S.-J., Chen, Y.-C., Huang, T.-H., Chen, H.-Y., Lin, C.-C. C., & Chen, P.-J. (2026). Hospital-Level Acute Care at Home via Direct Home Admission in Taiwan: A Non-randomized Controlled Pilot Trial Before National Implementation. Journal of General Internal Medicine. https://doi.org/10.1007/s11606-026-10828-2

Image Credits: AI Generated

DOI: 10.1007/s11606-026-10828-2

Keywords: hospital-at-home, direct home admission, acute care at home, frail older adults, home-based medical care, Taiwan, pilot trial, remote patient monitoring, geriatrics, health policy, patient satisfaction, length of stay