Are We Falling Short? Evaluating the Accuracy of Common Clinical Fall Risk Assessments in Stroke Survivors: A Systematic Review and Meta-Analysis

Read the full article See related articles

Listed in

This article is not in any list yet, why not save it to one of your lists.
Log in to save this article

Abstract

Stroke survivors experience a significant risk of falls, with 73% falling within the first year post stroke. Clinical practice guidelines currently use more than 27 assessment tools to evaluate fall risk in this population. However, there are conflicting findings regarding their diagnostic accuracy in correctly identifying those at risk of falling. This systematic review and meta-analysis evaluates the sensitivity and specificity of the Timed Up and Go and the Berg Balance Scale, which are the most commonly used clinical assessments for identifying fall risk among stroke survivors. Our data search included Web of Science, PubMed, and Ovid Medline. This protocol was registered in PROSPERO (CRD420251004460) prior to data extraction. Fifteen studies, comprising 1,492 stroke survivors, were included in the meta-analysis. We found a negative association between these tests and their ability to accurately identify fall risk in this population (OR = 0.469, 95% CI [-0.230, 1.169], however, this was not statistically significant ( p = 0.188). No heterogeneity was observed across studies (τ 2 = 0.000; I² = 0%). We found considerable variability in cut-off values across protocols, without significant moderating effects of these thresholds on their diagnostic accuracy. No publication bias was detected according to the Egger’s weighted test (t (20) = 0.024, p = 0.981) and the rank correlation test (τ = - 0.030, p = 0.867). Future research should focus on developing or implementing an objective stroke-specific fall risk assessment with appropriate cut-off values that better capture the underlying mechanisms of fall risk in stroke survivors to improve fall prevention strategies and rehabilitation care.

Article activity feed