PREDICTING HOSPITALIZATION AND FUNCTIONAL DECLINE IN OLDER HEALTH PLAN ENROLLEES - ARE ADMINISTRATIVE DATA AS ACCURATE AS SELF-REPORT

OBJECTIVE: To compare the predictive accuracy of two validated indices , one that uses self-reported variables and a second that uses variabl es derived from administrative data sources, to predict future hospita lization. To compare the predictive accuracy of these same two indices for predicting future functional decline. DESIGN: A longitudinal coho rt study with 4 years of follow-up. SETTING: A large staff model HMO i n western Washington State. PARTICIPANTS: HMO Enrollees 65 years and o lder (n = 2174) selected at random to participate in a health promotio n trial and who completed a baseline questionnaire. MEASUREMENT: Predi cted probabilities from the two indices were determined for study part icipants for each of two outcomes: hospitalization two or more times i n 4 years and functional decline in 4 years, measured by Restricted Ac tivity Days. The two indices included similar demographic charac teris tics, diagnoses, and utilization predictors. The probabilities from ea ch index were entered into a Receiver Operating Characteristic (ROC) c urve program to obtain the Area Under the Curve (AUG) for comparison o f predictive accuracy. RESULTS: For hospitalization, the AUC of the se lf-report and administrative indices were .696 and .694, respectively (difference between curves, P = .828). For functional decline, the AUC of the two indices were .714 and .691, respectively (difference betwe en curves, P = .144). CONCLUSIONS: Compared with a self-report index, the administrative index affords wider population coverage, freedom fr om nonresponse bias, lower cost, and similar predictive accuracy. A sc reening strategy utilizing administrative data sources may thus prove more valuable for identifying high risk older health plan enrollees fo r population-based interventions designed to improve their health stat us.