Coding sports injury surveillance data: has version 10 of the Orchard Sports Injury Classification System improved the classification of sports medicine diagnoses?

Abstract

To compare versions 8 and 10 of the Orchard Sports Injury Classification System (OSICS) to determine whether the revised version of OSICS has improved its use in a sports medicine setting, and to assess the inter-rater reliability of OSICS-10. Injury surveillance data, gathered over a 2 year period in professional football, cricket and rugby union to produce 335 diagnoses, were coded with both OSICS-8 and OSICS-10. Code-diagnosis agreement was assessed for OSICS-8 in terms of whether a diagnosis was codeable or noncodeable, and for OSICS-10 by evaluating the highest available OSICS-10 tier of coding. Eight clinicians coded a list of 20 diagnoses, comprising a range of pathologies to all gross anatomical regions, which were compared to assess inter-rater reliability. All diagnoses could be assigned an appropriate code with OSICS-10, compared with 87% of diagnoses that could be assigned an OSICS-8 code. Contusions comprised almost half of OSICS-8 noncodeable diagnoses. OSICS-10 tier 2 codes accounted for 20% of diagnoses coded with the updated system. Of these 20%, almost half contained a more detailed diagnosis that did not have an available OSICS-10 tier 3 or 4 code. Inter-rater reliability increased with decreasing diagnostic detail, with an overall level shown to be moderate (k = 0.56). OSICS-10 is a more encompassing system than OSICS-8 to use in classifying sports medicine diagnoses, and has a moderate level of inter-rater reliability. Further minor revision may be required to address lack of detail in some strain, effusion and contusion codes.