BackgroundIntermediate care is intended to reduce hospital admissions and facilitate early discharge. In Norway, a model was developed with transfer to intermediate care shortly after hospital admission. Efficacy and safety of this model have not been studied previously.In a parallel-group randomized controlled trial, patients over 70 years living at home before admission were eligible if clinically stable, without need for surgical treatment and deemed suited for intermediate care by attending physician. Intervention group patients were transferred to a nursing home unit with increased staff and multidisciplinary assessment, for a maximum stay of three weeks. Patients in the control group received usual care in hospital. Blinding to group assignment was not possible.The primary outcome was number of days living at home in a follow-up period of 365 days. Secondary outcomes were mortality, hospital admissions, need for residential care and home care services. Data were obtained from patient records and registers.Results376 patients were included, 74 % female and mean age 84 years. There was no significant differences between intervention (n = 190) and control group (n = 186) for number of days living at home (253.7 vs 256.5, p = 0.80) or days in hospital (10.4 vs 10.5, p = 0.748). Intervention group patients spent less time in nursing home (40.6 days vs. 55.0, p = 0.046), and more patients lived independently without home health care services (31.6 % vs 19.9 %, p = 0.007).For orthopaedic patients (n = 128), mortality was higher in the intervention group; 15 intervention patients and 7 controls died (25.1 % vs 10.3 %, p = 0.049). There was no significant difference in one-year mortality for medical patients (n = 150) or the total study population.ConclusionsThis model of rapid transfer to intermediate care did not significantly influence number of days living at home during one year follow-up, but reduced demand for nursing home care and need for home health care services. In post-hoc analysis mortality was increased for orthopedic patients.Trial registrationThe trial was registered 26. July 2013 at Current Controlled Trials and assigned with registration number ISRCTN21608185.

No abstract