Sociodemographic predictors of delayed- versus early-stage cervical cancer in California

Purpose: We sought to evaluate and distinguish roles of sociodemographic predictors for delayed- versus early-stage cervical cancer. Methods: Demographic variables for 13,624 cervical cancers having complete data for age at diagnosis (4 categories), race and ethnicity (4 categories), socioeconomic status (SES) quintiles, and marital status (3 categories) were extracted from the California Cancer Registry database for the period 1996 to 2005 and analyzed using multiple logistic regression as predictors of delayed- versus early-stage diagnosis. Results: Fifty-eight percent of cervical cancers were among women younger than 50years, compared with 46% of delayed-stage cases. Independent odds of delayed- versus early-stage cervical cancer were higher for older age categories within each race and ethnic group. Declining odds of delayed- versus early-stage diagnosis were evident for increasing SES quintiles among Asian or other (trend P=015), non-Hispanic black (P=024), Hispanic (P=001), and non-Hispanic white (P=001) women. Odds of delayed- versus early-stage cervical cancer were highest among unmarried compared with married women. Conclusions: Our findings support evidence that older age, low SES, and unmarried status predict delayed-stage cervical cancer diagnosis in each of the four major race and ethic groups. The two lowest SES quintiles independently identified larger percentages of delayed-stage cervical cancers in each of the race and ethnicity groups assessed, particularly among Hispanic and non-Hispanic black women.