Table 2

 Recommendations when using classification systems in injury research

(1) The classification system should be one that:(i) caters for all possible diagnoses (ii) has clear rules regarding what codes to use in particular circumstances (iii) is easy to use and navigate (iv) has a high intraobserver and interobserver reliability
(2) The research group should:(i) provide clear instructions to those clinicians assigning the diagnoses so that they provide written diagnoses that are able to be coded (ii) consider having the coding done by the sports practitioner assigning the diagnosis so that there is no chance for misinterpretation of the diagnosis when being coded (iii) have all coders knowledgeable in sports medicine diagnoses and/or able to communicate with the diagnosing practitioner if a query arises (iv) have all coders well trained in the coding system or under close supervision by an experienced coder (v) have established rules and guidelines about what aspect of the diagnosis to prioritise if the exact code is unavailable, e.g. location v pathology (vi) have a system whereby group consensus can occur with any code where the coder is unsure regarding the best code to use