Mobility interventions have been shown to mitigate functional decline in various clinical populations; however, the effects of mobility programs in older hospitalized patients are unclear. The objective of this study was to determine the effects of unstructured mobility programs on physical activity, physical function, length of stay (LOS), and quality of life (QOL) in older (≥60 years) general medicine inpatients.
In this systematic review and meta-analysis, we systematically searched MEDLINE, Embase, CINAHL, and AMED databases from inception to March 2020, plus hand screening references of relevant studies.
We included randomized controlled trials (RCTs) and quasi-experimental studies assessing the effects of mobility programs compared to usual care in older adults admitted to general medicine units.
Teams of 2 reviewers independently extracted data, assessed risk of bias, and evaluated quality of evidence. Where study population, intervention, and outcomes were similar, results from RCTs were combined by meta-analysis.
Three RCTs and 10 quasi-experimental studies met eligibility criteria. Interventions mainly included ambulation and staff, patient, or caregiver education. Meta-analyses showed that mobility interventions had a moderate effect on physical activity [step count standardized mean difference 0.60, 95% confidence interval (CI) 0.23-0.97] and a nonsignificant effect on LOS (mean difference -0.36, 95% CI -1.92 to 1.21), both favoring mobility. Narrative synthesis showed consistent evidence for improvement in physical function, potential decrease in LOS, and no increase in adverse events with mobility interventions.
Unstructured mobility interventions in general medicine units may improve older hospitalized patients’ physical activity and physical function; however, the quality of evidence was low. More RCTs are needed to evaluate the effectiveness of mobility interventions, particularly on LOS and QOL.

Copyright © 2020 AMDA – The Society for Post-Acute and Long-Term Care Medicine. Published by Elsevier Inc. All rights reserved.