Interventions to Support People With HIV Following Hospital Discharge: A Systematic Review

Abstract

Background. Individuals hospitalized with HIV-related complications face high post-discharge mortality and morbidity, particularly in resource-limited settings. This systematic review evaluated the impact of interventions to reduce post-hospital mortality, lower readmissions, and improve linkage to care. Methods. We searched the PubMed, Embase, and Cochrane databases up to 1 October 2024 for studies reporting outcomes of post-discharge interventions. Two independent reviewers performed study selection, extracted data, and assessed risk of bias. We pooled data using random effects meta-analysis. Results. We included 4 randomized controlled trials (conducted in Spain, South Africa, Tanzania, and the United States) and 6 observational studies (Canada, Thailand, Zambia, and the United States). Interventions included pre-discharge counseling, medication review, referral to care, and goal setting, as well as post-discharge follow-up via home visits, telephone calls, and support from social workers or community health workers. Pooled data from randomized controlled trials showed no difference between post-discharge interventions and usual care in mortality, but the estimate was imprecise (relative risk [RR], 0.98; 95% CI, .59–1.63). However, interventions may reduce readmissions (RR, 0.82; 95% CI, .52–1.30) and may slightly improve linkage/ retention in care (RR, 1.10; 95% CI, .95–1.27). Observational studies reported similar results, with no mortality effect but potential reductions in readmissions (RR, 0.77; 95% CI, .48–1.25) and improved linkage/retention (RR, 1.42; 95% CI, 1.11–1.81). Interventions were largely feasible, acceptable, and low cost. Conclusions. Interventions that include pre-discharge care planning and post-discharge follow-up, such as telephone contact and home visits, may improve linkage to care and reduce readmissions. However, interventions were not associated with reduced post-discharge mortality.