Adding Confidence to our Injury Burden Estimates: Is Bootstrapping the Solution?

Injury burden is a composite measure of injury incidence and mean severity that can be used to understand the overall impact of injuries and help identify priority areas for injury prevention.1 Injury burden has been used within rugby union epidemiological studies since the early 2000s,2 but it is now recognised and recommended within other sports, including the most recent International Olympic Committee consensus statement for the recording and reporting of epidemiological data on injury and illness.3 Injury burden is normally reported as athlete days absence per 1000 athlete-hours and is derived from the product of injury incidence (expressed as injuries sustained/1000 athlete-hours) and severity (expressed as the mean severity of injury in days).1 While the value of injury burden as an output measure from injury surveillance studies is evident, there appears to be some confusion in the literature regarding its calculation. For instance, some authors have used median severity to calculate injury burden rather than mean severity, as discussed in a recent critical review.4 In addition, there appears to be no clear guidance within the sports medicine literature regarding the most appropriate way to calculate confidence intervals (CIs) for this metric.