Ethnicity and Health Care in Cervical Cancer Survival: Comparisons between a Filipino Resident Population, Filipino-Americans, and Caucasians

Few studies have assessed and compared cervical cancer survival between developed and developing countries, or between ethnic groups within a country. Fewer still have addressed how much of the international or interracial survival differences can be attributed to ethnicity or health care. To determine the role of ethnicity and health care, 5-year survival of patients with cervical cancer was compared between patients in the Philippines and Filipino-Americans, who have the same ethnicity and between Filipino-Americans and Caucasians, who have the same health care system. Cervical cancer databases from the Manila and Rizal Cancer Registries and Surveillance, Epidemiology, and End Results 13 were used. Age-adjusted 5-year survival estimates were computed and compared between the three patient groups. Using Cox proportional hazards modeling, potential determinants of survival differences were examined. Overall 5-year relative survival was similar in Filipino-Americans (68.8%) and Caucasians (66.6%), but was lower for Philippine residents (42.9%). Although late stage at diagnosis explained a large proportion of the survival differences between Philippine residents and Filipino-Americans, excess mortality prevailed after adjustment for stage, age, and morphology in multivariate analysis [relative risk (RR), 2.07; 95% confidence interval (0), 1.68-2.55]. Excess mortality decreased, but persisted, when treatments were included in the multivariate models (RR, 1.78; 95% Cl, 1.41-2.23). A moderate, marginally significant excess mortality was found among Caucasians compared with Filipino-Americans (adjusted RR, 1.22; 95% Cl, 1.01-1.47). The differences in cervical cancer survival between patients in the Philippines and in the United States highlight the importance of enhanced health care and access to diagnostic and treatment facilities in the Philippines. (Cancer Epidemiol Biomarkers Prev 2009;18(8):2228-34)