查看文章

Administrative claims datasets have great potential for health services researchers who wish to evaluate patient care on a large scale across providers, but categorizing patients’ primary health conditions from these data can be challenging. The goal of this work is to describe and evaluate a methodology to assign workers compensation claimants to meaningful groups within back and shoulder injuries using claims data. Claims data from a large multi-state workers compensation insurance dataset were used to assign eligible claimants to condition and subcondition groups using available ICD9 codes. Assignments were evaluated against body part indicators, severity indicators, resource utilization, and specific clinical interventions. Of the 575,967 claimants who met inclusion criteria, 54,066 claimants were designated as shoulder injuries and 118,772 were designated as back injuries. Within back and …