OBJECTIVES

To perform an umbrella review of systematic reviews with meta-analyses (MAs) examining the effectiveness of comprehensive geriatric assessment (CGA) delivered within community settings to general populations of community-dwelling older people against various health outcomes.

DESIGN

Umbrella review of MAs of randomized controlled trials (RCTs).

SETTING AND PARTICIPANTS

Systematic reviews with MAs examining associations between CGA conducted within the community and any health outcome, where participants were community-dwelling older people with a minimum mean age of 60 years or where at least 50% of study participants were aged ≥60 years. Studies focusing on residential care, hospitals, post-hospital care, outpatient clinics, emergency department, or patients with specific conditions were excluded.

METHODS

We examined CGA effectiveness against 12 outcomes: not living at home, nursing home admission, activities of daily living (ADLs) and instrumental ADLs (IADLs), physical function, falls, self-reported health status, quality of life, frailty, mental health, hospital admission, and mortality, searching the MEDLINE/PubMed, Cochrane Library, CINAHL, Embase databases from January 1, 1999, to August 10, 2022. AMSTAR-2 was used to assess the quality of included systematic reviews, including risk of bias.

RESULTS

We identified 10 MAs. Only not living at home (combined mortality and nursing home admission) demonstrated concordance between effect direction, significance, and magnitude. Significant effects were more typically observed in earlier rather than later studies.

CONCLUSION AND IMPLICATIONS

Given the widespread adoption of CGA as a component of usual care within geriatric medicine, the lack of strong evidence demonstrating the protective effects of CGA may be indicative of a cohort effect. If so, future RCTs examining CGA effectiveness are unlikely to demonstrate significant findings. Future studies of CGA in the community should focus on implementation and adherence to key components.

TRIAL REGISTRATION

Study protocol registered in PROSPERO 2020 CRD42020169680.