2056. Identifying Future Very High Cost Patients Using Decision Theory
Kenneth C Pietz, PhD, Houston Center for Quality of Care and Utilization Studies, Houston VAMC and Baylor College of Medicine, MM Byrne, Center for Bioethics and Health Law and Department of Medicine, University of Pittsburg, M Thompson,
Houston Center for Quality of Care and Utilization Studies, Houston VAMC, HA Nelson,
Houston Center for Quality of Care and Utilization Studies, Houston VAMC, S Sookanan,
Houston Center for Quality of Care and Utilization Studies, Houston VAMC, LA Petersen,
Houston Center for Quality of Care and Utilization Studies, Houston VAMC and Baylor College of Medicine
Objectives: Assess the application of a decision rule to identify future high cost users of medical care.
Methods: Using VA administrative databases, we identified 2,896,473 veterans who used VA medical care services in fiscal year 2000 and were alive on October 1, 2000. We defined very high cost (VHC) as greater than the 99th percentile of cost for the following year. Diagnostic Cost Groups (DCGs) were used for risk adjustment. A decision rule was developed, using costs of misclassification. The ratio of the cost of failing to identify a true VHC patient to the cost of falsely identifying one who is not VHC (the loss ratio) must be specified. The decision rule minimizes the expected loss.
Results: The decision rule divided the population into potential VHC and non-VHC patients, as a function of the loss ratio. For example, the ratio must be at least 3 in order to classify any patients as VHC. If the ratio is 10, all patients in DCG 15 and above are classified as VHC.
Conclusions: The proposed method can be used by health plan administrators such as the VA to identify future VHC patients, provided that the loss ratio is specified. This methodology can provide insight even if costs of misclassification are not known exactly, by trying multiple values of the ratio to determine the effect on VHC classification.
Impact: More research is needed into the application of decision theory to identify potential high-cost patients.