Consumer turnover in service utilization patterns: implications for capitated payment

Risk-adjusted capitation rates based on a group's expected cost to a provider have been proposed to create incentives for including people with severe mental disabilities in managed healthcare systems. The most appropriate methods for classifying people according to levels of risk, however, have yet to be determined. This research employed cluster analysis statistical techniques to identify patterns of services delivered to a sample of 4,346 consumers in Ohio. Findings about the extent of client movement in and out of service clusters over time suggest caution in the implementation of risk-adjusted capitation plans, particularly those based primarily on past service utilization.