OBJECTIVE To determine the Fragility Index of hamstring injury risk factors, defined as the minimum number of participants who would need to change classification to make a hamstring injury risk factor statistically nonsignificant. DESIGN Retrospective secondary data analysis. METHODS Studies that investigated 1 or more risk factors for hamstring injury, and presented sufficient data to develop a 2 × 2 contingency table were included. A systematic literature search and reference screening of a recent hamstring injury systematic review were conducted to identify 78 articles. Relative risk and 95% confidence intervals were determined and then systematically recalculated by removing 1 observation from the high-risk injury count and adding it to the high-risk noninjury count. The Fragility Index for a risk factor was the number of observations required to be moved between groups until the relative risk was no longer significant. RESULTS The median Fragility Index of all hamstring injury risk factors was 3 (Q1-Q3 = 2-6). The Fragility Index for nonmodifiable risk factors was 3 (Q1-Q3 = 2-6) and 3 (Q1-Q3 = 2-5) for modifiable risk factors. Over 35% of all included hamstring injury risk factors had a Fragility Index of =2. CONCLUSION Most statistically significant hamstring injury risk factors are fragile associations. The interpretation of significant hamstring injury risk factors should consider a range of statistical metrics, and while the Fragility Index should never be considered in isolation, it is an intuitive measure to help assess the robustness of findings.
© Copyright 2024 Journal of Orthopaedic & Sports Physical Therapy. All rights reserved.