Risk factors and risk statistics for sports injuries.

BACKGROUND Risk factors for sports injuries include characteristics and behaviors of athletes and characteristics of sports and the environment that are associated with some measure of risk of injury. OBJECTIVE To introduce risk statistics to clinicians evaluating studies of sports injuries. METHODS Plain-language review of risk statistics and their practical application to sports injuries. RESULTS The various measures of injury incidence are injury risk (proportion of athletes injured in a given period of training, playing, or other exposure time), injury rate (number of injuries per unit of exposure time), odds of injury (probability injury will happen divided by probability injury will not happen), injury hazard (instantaneous proportion injured per unit of time or mean injury count per unit of time), and mean time or mean number of playing exposures to injury. Effects of risk factors are estimated as values of effect statistics representing differences or ratios of one or more of these measures between groups defined by the risk factor. Values of some ratios and their sampling uncertainty (confidence limits) are estimated with specialized procedures: odds ratios with logistic regression, rate ratios with Poisson regression, and hazard ratios with proportional hazards (Cox) regression. Injury risks and mean time to injury in each group can also be estimated and can give a better sense of the effect of a risk factor. Risk factors identified in nonexperimental cohort and case-control studies are not always causes of injury; data from randomized controlled trials provide stronger evidence of causality. CONCLUSION Expressing risk statistics as meaningful numbers should help clinicians make better use of sports injury studies.