Costs and survival of patients with colorectal cancer in a health maintenance organization and a preferred provider organization

Background: Colorectal cancer is relatively frequent among adults of working age, yet few studies have examined treatment, outcomes, and costs for people under 65 years of age with this disease. Objective: The objective of this study was to compare the initial treatments, survival, cancer-related medical costs, and overall medical costs for working-aged persons with colorectal cancer in 2 large health insurance plans in Washington State, one a preferred provider organization (PPO) and the other a group model health maintenance organization (HMO). Study Population: This study consisted of patients, aged 20-64 years, diagnosed with colorectal cancer in both health plans from 1996 to 1998. For each cancer case, up to 5 control subjects, matched on age and sex, were selected for the analysis. Methods: We calculated unadjusted, attributable, and overall medical costs using the Kaplan-Meier sample average estimator. We calculated relative mortality rates using Cox regression. We used propensity scores to adjust overall costs and survival for potential confounding factors. Results: Two hundred ten persons in the PPO and 136 persons in the HMO, aged 20-64 years, were diagnosed with cancer over the observation period and included in this study. Patients in the PPO were more likely to have local excision of their tumor (16% compared with 11%) and were less likely to receive chemotherapy (48% compared with 60%). The overall medical costs for the cancer cases were $46,000 in the HMO and $46,400 in the PPO (95% confidence interval for the difference: -$19,300 to 20,100). The cancer-attributable medical costs over 2 years were $40,400 in the HMO and S44,300 in the PPO (95% confidence interval for the difference: -$17,400 to 25,200). Survival was similar in the 2 health plans: the hazard ratio was 0.89 for those enrolled in the PPO (95% confidence interval: 0.50 to 1.59). Adjustment for potential confounding factors altered the results little. Conclusions: There were differences in the initial treatment of the patients in each health plan, but costs and survival were not significantly different between the 2 plans.