BACKGROUND: There is a need to develop other casemix classifications, apart from DRG for sub-acute and non-acute inpatient care payment mechanism in Thailand. OBJECTIVE: To develop a casemix classification for sub-acute and non-acute inpatient service. MATERIAL AND METHOD: The study began with developing a classification system, analyzing cost, assigning payment weights, and ended with testing the validity of this new casemix system. Coefficient of variation, reduction in variance, linear regression, and split-half cross-validation were employed. RESULTS: The casemix for sub-acute and non-acute inpatient services contained 98 groups. Two percent of them had a coefficient of variation of the cost of higher than 1.5. The reduction in variance of cost after the classification was 32%. Two classification variables (physical function and the rehabilitation impairment categories) were key determinants of the cost (adjusted R2 = 0.749, p = .001). Validity results of split-half cross-validation of sub-acute and non-acute inpatient service were high. CONCLUSION: The present study indicated that the casemix for sub-acute and non-acute inpatient services closely predicted the hospital resource use and should be further developed for payment of the inpatients sub-acute and non-acute phase.
BACKGROUND: There is a need to develop other casemix classifications, apart from DRG for sub-acute and non-acute inpatient care payment mechanism in Thailand. OBJECTIVE: To develop a casemix classification for sub-acute and non-acute inpatient service. MATERIAL AND METHOD: The study began with developing a classification system, analyzing cost, assigning payment weights, and ended with testing the validity of this new casemix system. Coefficient of variation, reduction in variance, linear regression, and split-half cross-validation were employed. RESULTS: The casemix for sub-acute and non-acute inpatient services contained 98 groups. Two percent of them had a coefficient of variation of the cost of higher than 1.5. The reduction in variance of cost after the classification was 32%. Two classification variables (physical function and the rehabilitation impairment categories) were key determinants of the cost (adjusted R2 = 0.749, p = .001). Validity results of split-half cross-validation of sub-acute and non-acute inpatient service were high. CONCLUSION: The present study indicated that the casemix for sub-acute and non-acute inpatient services closely predicted the hospital resource use and should be further developed for payment of the inpatients sub-acute and non-acute phase.