Purpose: To outline the development and implementation of the Canadian Intercollegiate Sport Injury Registry (CISIR), to examine its validity, including the data collection forms, the recording of athlete exposure, and the mechanism of injury, and to determine the ability of the CISIR to meet its stated objectives of assessing rates and risk of injury.
Design: Prospective cohort study.
Setting: Canadian intercollegiate athletics.
Subjects: 344 varsity football players from five western Canadian universities. ASSESSMENT OF RISK FACTORS AND OUTCOME MEASURES: Three data collection instruments were developed to capture the principle types of information forming the cornerstones of the CISIR: a medical form for preseason assessment of risk factors, a weekly exposure sheet (WES) for the documentation of daily individual athlete participation, and an individual injury report form (IIRF) for collection of injury-related information. Design and implementation input was provided by therapists and physicians through initial meetings, pilot testing, site visits, questionnaire, and final consensus meeting. The completeness of injury reporting was assessed through cross-referencing with participation time loss data. An item analysis was conducted on the principal elements of the IIRF. The categorization of participation itself was also examined, as was the diagnostic agreement between the therapists and physicians involved in data collection. The recorded mechanism of injury was compared with that noted through a video analysis for game-related injuries. Lastly, a test analysis was conducted to extract data and compute rates and risks of injury.
Results: This developmental phase was successful, with 99.7% subject enrollment, high therapist satisfaction, and good flow of data. A relational database, incorporating dual-entry data verification, was designed and functioned well. The collection process revealed that 100% of the WESs were submitted, and the data therein was 99.7% complete. The injuries resulting in participation time loss were recorded on an IIRF 97.9% of the time. The exposure (participation) codes were thought to be overly precise, and a simplification of these categories is suggested. The diagnostic agreement between physicians and therapists was 70%. It was possible to validate game exposures, but no standard was identified to permit validation of the categories of exposure. Likewise, the mechanism of injury as recorded by the therapists was thought to be more precise than the video analysis. After two modifications in the table structure of the relational database, it was possible to extract data relating to rates and risks of injury.
Conclusions: This study demonstrated a high degree of validity for many elements of the CISIR. One limitation was that no reference standard existed for some components, limiting some aspects of validity assessment. With the suggested revisions, the CISIR represents the current standard in athletic injury reporting in terms of individual injury risk assessment. This system will be used in the future to explore the prediction and prevention of sport injuries.