Organ donor index: a benchmark for comparing hospital organ donor rates.

OBJECTIVE: To develop organ donor indices to assess donor rates of individual hospitals. DESIGN: Data from hospital databases were retrospectively reviewed for patient separation ICD-9-CM codes (i.e., diagnostic codes from the International classification of diseases, 9th revision, clinical modification) to identify and categories actual and potential organ donors. Organ donor indices for groups of codes and for individual hospitals were determined by dividing the number of actual donors by the total number of patients who died with the same separation ICD-9-CM codes. SETTING: The three South Australian adult tertiary hospitals in 1988-1995. PATIENTS: The 154 actual organ donors, and all patients aged less than 71 years who died with the same groups of ICD-9-CM codes as the organ donors. RESULTS: Organ donors could be classified by three groups of ICD-9-CM codes specifying diseases or pathological processes that could result in brain death. These groups were head injury (44.2% of donors), cerebrovascular accident (CVA) (42.2%), and eight "other" codes (13.6%). Differences between the head injury donor indices for the three hospitals were not significant (Hospital A, 19.1%; Hospital B, 24%; Hospital C, 21%), but there were significant interhospital differences in donor indices for the CVA group (A, 11.2%; B, 5.7%; C, 5.1%; P < 0.05) and the "other" group (A, 3.6%; B, 0.7%; C, 0.3%; P < 0.001). CONCLUSIONS: ICD-9-CM codes can be used to describe organ donors and hospital populations from which potential organ donors may be found. The casemix-controlled organ donor indices can be used to compare the organ donor rates of individual hospitals and to examine reasons for low rates (other than purely casemix variation).