Geriatric assessment for older adults admitted to the emergency department: A systematic review and meta-analysis

Abstract

Older adults are the most frequent users of emergency services. Comprehensive geriatric assessment (CGA) can help identify high-risk older adults at an early stage. We conducted a systematic review and meta-analysis to identify and evaluate CGA tools used in the emergency department (ED), analyze their predictive validity for adverse outcomes and recommend tools for this particular situation. We systematically searched Medline, Web of Science and CENTRAL for eligible articles published in peer-reviewed journals that observed patients ≥65years admitted to the ED, used at least one assessment tool and reported adverse outcomes of interest. We performed a descriptive analysis and a bivariate meta-analysis of the diagnostic accuracy and predictive validity of the assessment tools for the chosen adverse outcomes. 28 eligible studies were included. The pooled sensitivity (95% CI) of the assessment tools for predicting mortality within short (28-90days) and long (180-365days) periods after the first ED visit was 0.77 (0.61-0.89) and 0.79 (0.46-0.96), respectively, with specificity (95% CI) values of 0.45 (0.32-0.59) and 0.37 (0.14-0.65). These findings indicate that the tools used in the included studies had modest predictive accuracy for mortality and were more appropriate for identifying individuals at high risk of readmission in the short term than in the long term. Early use of assessment tools in the ED might improve clinical decision making and reduce negative outcomes for older adults.