Why methods matter in a meta-analysis: a reappraisal showed inconclusive injury preventive effect of Nordic hamstring exercise

Abstract

Objectives: The Nordic hamstring exercise (NHE) has been strongly recommended to reduce hamstring injuries in previous meta-analyses (50% reduction in risk of injury). To underline the importance and impact of adopting appropriate methodology for evidence synthesis, we revisited the study selection, reanalyzed and updated the findings of the most recent meta-analysis. Study Design and Setting: Only randomized control trials (RCT) using NHE as one of the prevention arms were selected. Summary effects for risk ratios (RR) for original studies included in the earlier meta-analysis, and new studies identified (update), were re-estimated under the random-effects model and presented with 95% confidence intervals (CI) and prediction intervals (PI). Tentative recommendations were provided according to the Grading of Recommendations Assessment, Development and Evaluation. Results: Only five RCTs out of the 15 studies included in the earlier meta-analysis randomized to NHE. Our update revealed one additional RCT. The point estimate (RR) for the five RCTs previously considered RCTs was 0.56 (95% CI, 0.20–1.52; 95% PI, 0.06–5.14, parametric, and 0.13–1.80, nonparametric). After the update, the RR was 0.59 (95% CI, 0.27–1.29; 95% PI 0.10–3.29, parametric, and 0.17–1.52, nonparametric). Conclusion: Contrary to the conclusions of a recent meta-analysis, as well as earlier meta-analyses, by using more appropriate methodology, the evidence underpinning the protective effect of NHE so far remains inconclusive and mostly derived from high risk of bias RCTs. At best, only conditional recommendation can be provided (for soccer) and future RCTs are warranted.