Comparison of Comorbidity Measurements to Control for Confounding in Health Outcomes Studies

摘要

ARTICLE SUMMARY: This study compares the performance of several co- morbidity indices and simple count measurements in the prediction of future health expenditures. OBJECTIVES: This study compares the performance of the Elixhauser and Charlson Indices with the Rx-Risk-V score and several simple count measurements including counts of prescriptions, physician's visits, hospital claims, unique prescription classes, and Diagnosis Clusters. STUDY DESIGN: Simple count measurements, the Elixhauser and Charlson Indices, and the Rx-Risk-V score were calculated one year prior to the filling of a new prescription for an anti-hypertensive medication for 20,378 members of a managed care organization. The primary outcome variable was the log transformed sum of prescription, physician, and hospital expenditures in the year following the prescription encounter. METHODS: In addition to descriptive statistics and spearman correlations between measurements, the predictive performance of each measurement was determined using linear regression models and corresponding R2 statistics.