Background The usefulness of the Swedish Parachute Association (SFF) national registry of skydiving injuries may benefit from a description of its base population and an evaluation of its sensitivity and specificity.

Methods Cross-sectional analysis based on data collected through a web-based questionnaire (SKYNET), from all persons renewing a Swedish skydiving license 2008 (n=1049). Sensitivity was measured as the proportion of injury events fulfilling the reporting criterion that were actually reported, by comparing the injury events reported to the SFF during the skydiving seasons 2006 and 2007 with the injury events retrospectively reported to the web-based questionnaire for the same time period. Specificity was measured as the proportion of false positives for the same time period. Factors affecting the likelihood of reporting were evaluated using logistic regression.

Results Response rate 100%. Overall sensitivity 2006/2007 was 0.37 (95% CI 0.24 to 0.51). For self-stated non-minor injuries, the sensitivity was 0.67 (95% CI 0.43 to 0.85). No significant effect on reporting was found for gender, age, license level, years in the sport, total number of jumps or club affiliation. The specificity was 0.91 (95% CI 0.83 to 0.95). Descriptive statistics of the Swedish skydiving population show gender differences.

Conclusions The low sensitivity will yield false low incidence calculations, but risk comparisons related to the candidate predictor variables appear reasonably valid. The false positive reporting warrants assortment of incoming data.