A Low-Cost Approach to Prospective Identification of Impending High Cost Outcomes

Objectives. The overall objective of this study was to define and evaluate patterns of use of medical services in the care of patients with chronic illness that represent circumstances which, if modified, might lead to reduction in risk of acute-level care. Methods. This was a retrospective observational study. The study population consisted of Kaiser Permanente enrollees at four sites during January 1993 through June 1995, who were 20 to 64 years of age and had two of three chronic diseases (diabetes, circulatory disorders, obstructive pulmonary disorders). Using logistic regression, the effect of primary care visit patterns and therapeutically risky drug combinations on likelihood of hospital admission in a subsequent 3-month period is adjusted for age, gender, and disease state in the prior 12-month period. Results. Enrollees with visits to three or more different primary care physicians were 46% more likely to be admitted than expected (P < 0.01) according to their age, gender, and disease state, and those with therapeutically risky drug combinations were 34% more likely to be admitted (P < 0.01). Conclusions. The risk adjustment models evaluated in this study defined care processes associated with increased risk of subsequent acute-level services. Those processes may represent nascent acute disease states or suboptimal organization of care delivery. The results of these models can be used to inform changes in organization and delivery of outpatient care that might improve patient outcomes.