Intraobserver and interobserver agreement of International Classification of Diseases, Ninth Revision codes in classifying shoulder instability

HypothesisThe purpose of this study was to investigate the intraobserver and interobserver reliability of the International Classification of Diseases, Ninth Revision (ICD-9) system when applied to the classification of shoulder instability.Materials and methodsFrom December 2005 through February 2007, 50 patients with shoulder instability were evaluated and classified by an attending physician using one of the 16 ICD-9 codes for shoulder instability. Patients were reassessed after two weeks by the original physician and two additional shoulder specialists. 42 patients completed the study. These data were then analyzed to assess intraobserver and interobserver reliability.ResultsIntraobserver agreement for ICD-9 codes was 50% (κ=0.25, fair). Interobserver agreement was 23% (κ=0.002, poor).DiscussionThe ICD-9 coding system is the recognized standard for classifying disease states and is used for large epidemiologic studies. The poor agreement demonstrated in this study suggests that the ICD-9 coding system has poor agreement and as such is not a precise method to classify shoulder instability.ConclusionShoulder instability cannot reliably be classified using the ICD-9 coding system. Until a more reliable system is developed, epidemiologic studies of shoulder instability that use ICD-9 codes may be difficult to interpret.Level of evidenceLevel 1; Testing a previously developed classification system.