Objective: To compare the most commonly used and proposed injury definitions for surveillance systems in team sports and attempt to assess their suitability for consensus definitions in terms of reliability and functionality.
Data sources: The PubMed and SportDiscus databases were searched for papers on team sports that discussed consensus definitions or compared various definitions of injury.
Data synthesis: A continuum between the most broad "tissue damage" definition and the most narrow "match time loss only" definition was developed.
Results: A "match time loss only" injury definition can be reliably and accurately applied but only captures a small percentage of the total pool of all "tissue damage" injuries. There are some inherent biases in using a match time loss only definition (late season matches, matches with unequal breaks between games), but these are clearly visible. All other definitions improve the volume of data captured but suffer serious theoretical and/or practical flaws with respect to accuracy and reliability. No study using a broad definition has demonstrated good reliability to date (eg, using 2 independent recorders at the same team).
Conclusion: A "match time loss only" injury definition is the most accurate and reliable of those commonly used in team sports. Other injury definitions are broader and may be more appropriate for individual team and specific injury studies. However, a match time loss definition is the most accurate and reliable tool for comparing injury rates at different teams and between different seasons within teams. Hence, we recommend this as the basis for the injury definition in a consensus statement.