Applying Conjoint Analysis To Predict Inpatients' Preference For Medical Care

Object: To provide a new technique?conjoint analysis for decision- making in respect of hospital management and medical care market research. To explore some methdological issues in the application of conjoint analysis. Subjects: Two hundred and flinty inpatients were randomly selected in proportion as 15% from 21 departments of two general hospitals in Guangzhou. Method: The inpatients were interviewed with a questionnaire including five factors: drug cost, bed expenses, charging information, service after discharge, informed decision-making. Conjoint analysis was applied to predict patients?preference for medical care. Resu I t: The part-worth model had the highest validity with Pearson抯 correlation coefficient 0.997 among all conjoint analysis models. The satisfied reliability was based on high agreement between scoring scale and rating scale. The scheme of medical care for inpatient was suggested as drug with medican price, double room with RMB 50-70 per day, cost list delivered every 3-5 days, seeing doctors for free within a certain period after discharge and informed-decision desired. Conc I us i on: Conjoint analysis is a valid and reliable technique to assess and predict inpatients?preference for medical care. The model should be established at individual level other than aggregate level. There were obvious contradiction between supply and demand in medical care provided by these two hospitals. The present medical service of the hospitals should be adjusted to meet inpatients?need according to the study.