Identifying and characterizing one-time and never users of mammography screening services among Medicare eligible women

Abstract Background. Despite eligibility for a screening mammogram once every 2 years from 1991 to 1997, only a small percentage of Medicare women utilized this benefit. We examined mammography use among 388,707 North Carolina Medicare women from 1994 to 1997 to identify characteristics of one-time and never users of mammography. Methods. Data were obtained from North Carolina Medicare mammography claims and enrollment files from 1994 to 1997. Women ages 65+ as of 01/01/1994, continuously enrolled in Medicare from 1994 to 1997, and alive as of 12/31/1997 were included in the sample ( n = 388,707). Mammogram use was categorized as never, once, or at least twice during 1994/1995 and 1996/1997. Women with at least one mammography claim during 1994/1995 and at least one mammography claim during 1996/1997 were called repeat users, women with one mammography claim during the 4 years were labeled one-time users, and women with zero mammography claims during the 4 years were termed never users. Multivariate logistic regression analyses were conducted to determine associations between characteristics and mammography frequency. Results. Biennial mammography claims data rates were 35.3% in 1994/1995 and 41.8% in 1996/1997. Compared with all other users, one-time users ( n = 108,899) were more likely to be ages 65–74 (vs 75–84 and 85+), live in an urban versus rural county, and be eligible for Medicare only versus Medicare and Medicaid. Never users ( n = 184,545) were more likely to be ages 85+, be non-Caucasian, live in a rural county, and be eligible for both Medicare and Medicaid versus Medicare. Conclusions. These results demonstrate different demographic characteristics for one-time and never mammography users. This approach of using multiple years of claims data to segment the targeted population provides the opportunity to tailor interventions to subgroups.