Health Outcomes of Post‐Hospital In‐Home Team Care: Secondary Analysis of a Swedish Trial

Publisher Website
Google Scholar
Abstract
To determine patient and treatment-related factors predictive of health outcomes. Secondary analysis of a randomized trial with 6-month follow-up. After using bivariate and three-way analysis in the total sample to screen outcome predictors and interactions among baseline variables, multivariate logistic regression was used to model outcomes. A county general hospital in central Stockholm, and patients' homes nearby. Hospital inpatients stable for discharge from acute care, having at least one chronic condition, and dependent in 1 to 5 Katz activities of daily life (ADLs) were included. Subjects (mean age = 81.1 years) were randomized to "team" (n = 150) or "usual care" (n = 99). Team patients were eligible for in-home primary care by an interdisciplinary team that included a physician, physical therapist, and 24-hour nursing services and geriatric consultation where necessary. "Usual-care" patients received standard district nurse-administered services at home upon hospital discharge. Demographic, functional status, and medical characteristics were measured at randomization. Outcomes included survival and higher ADL, instrumental ADL (IADL), and outdoor ambulation scores. Multiple medical, social, behavioral, and functional factors were associated with outcomes. Primary cardiac disease, number of prescription drugs, alcohol abstinence, and baseline mental status all impacted 6-month survival. Controlling for other factors, team care improved the likelihood of ambulation independent of personal assistance at follow-up (P = .027), treating an estimated 10 patients per 1 benefiting. Further, rehabilitative in-home team care neutralized mortality and functional risk factors (low number of baseline contacts and coresidence) apparent in usual care. Heterogeneous clinical populations of older patients contain many prevalent characteristics important to outcomes. Secondary analysis of trials including interactions identifies treatable and untreatable risks, what program components may be effective, and who benefits.