The relationship of HMOs, health insurance, and delivery systems to breast cancer outcomes

BACKGROUND. The current climate of anger and frustration with managed care has heightened interest in the quality of health care provided by managed-care plans, particularly health maintenance organizations (HMOs). This breast cancer outcomes study, investigating relationships of health insurance and delivery systems to stage at diagnosis, treatment selected, and survival, is based in a heavily penetrated, highly competitive HMO market. METHODS. Data for 1,788 residents of northern California younger than 65 years of age at diagnosis (1987-1993) were provided by a population-based cancer registry. Patient insurance included fee-for-service (FFS), group-model HMO, nongroup HMO, publicly insured, and uninsured. Diagnosis and treatment occurred in 73 hospitals (large, medium/moderately small, or very small community, rural, teaching, or HMO-owned hospitals). Regression models examined relationships of insurance and hospital type to 3 outcomes (stage, treatment, and survival), controlling for age, ethnicity, education, neighborhood occupational class, and time period. RESULTS. Early diagnosis was as likely for group-model and nongroup-model HMO-insured patients as for the private FFS-insured patients. In 1987-1990, HMO-owned hospitals were leaders in treating 46% of early-stage breast cancers with breast-conserving surgery plus radiation (BCS+); by 1991-1993, the most significant increases in BCS+ use occurred at teaching and large community hospitals. Survival of group-model HMO, nongroup-model HMO, and FFS patients was not significantly different. Publicly insured/uninsured patients had more stage III/IV disease (OR=2.01, P = 0.006) and greater all-cause mortality (risk ratio 1.46, P = 0.015). CONCLUSIONS. Group-model and nongroup-model HMO patients are similar to FFS-insured patients in stage at diagnosis and survival outcomes. Treatment selection is related to hospital type rather than insurance coverage.