Abstract
OBJECTIVES: To examine associations between falls and subsequent motor vehicle crashes (MVCs), crash-related injuries, driving performance, and driving behavior. DESIGN: Systematic review and metaanalysis. PARTICIPANTS: Observational studies including drivers aged 55 and older or with a mean age of 65 and older. MEASUREMENTS: Two authors independently extracted study and participant characteristics, exposures, and outcomes and assessed risk of bias. Pooled risk estimates for MVCs and MVC-related injuries were calculated using random-effects models. Other results were synthesized narratively. RESULTS: From 3,286 potentially eligible records, 15 studies (N = 27-17,349 subjects) met inclusion criteria. Risk of bias was low to moderate, except for cross-sectional studies (n = 3), which all had a high potential for bias. A fall history was associated with a significantly greater risk of subsequent MVC (summary risk estimate = 1.40, 95% confidence interval (CI) = 1.20-1.63; I(2 ) = 28%, N = 5 studies). One study found a significantly greater risk of MVC-related hospitalizations and deaths after a fall (hazard ratio = 3.12, 95% CI = 1.71-5.69). Evidence was inconclusive regarding an association between falls and driving cessation and showed no association between falls and driving performance or behavior. CONCLUSION: Falls in older adults appear to be a risk marker for subsequent MVCs and MVC-related injury. Given the nature of the evidence, which is limited to observational studies, the identified associations may also result at least partly from confounding or bias. Further research is needed to clarify the mechanisms linking falls to crash risk and to develop effective interventions to ensure driving safety in older adults with a history of falls.