Mammography underutilization among older women in Connecticut

The primary goals were to examine mammography use rates among older women in Connecticut and to determine if there was significant variation among different areas and racial groups in the state. The secondary goal was to examine what impact the initiation of Medicare reimbursement for mammography screening has had on mammography use.

Statewide use rates were determined by retrospective Medicare Part B mammography claims analysis. Small area analysis methodology (SAA) was used to identify mammography rates for 23 hospital service areas (HSAs), representing all of the catchment areas for Connecticut's acute care hospitals.

Female Medicare beneficiaries 65 years and older with Part B coverage residing in Connecticut during the study period.

The main outcome (the use of at least one mammogram) was calculated for the calendar years 1991, 1992, and 1993. Mean annual use rates in 1993 were generated for the 23 HSAs and the different racial groups in Connecticut. To examine the effect that Medicare reimbursement for screening mammograms has had on mammography use, rates were calculated for women who met Medicare reimbursement criteria in 1991 through 1993. The rates in 1992 and 1993 were then compared with those in 1991, when the reimbursement program was first initiated.

The mean statewide annual rates among women aged 65 years and older were 23.4% (1991), 24.5% (1992), and 24.9% (1993). The mammography use rates among black women 65 years and older were significantly lower than their white peers in 1991 (18.8% black vs 23.8% white, P < .001), 1992 (20.6% vs 24.7%, P < .001), and 1993 (22.0% vs 25.1%, P < .001). Significant variation was identified among hospital service areas (HSAs) within the state for each time interval studied. The use rates among women aged 65 years and older who were eligible for Medicare screening mammography reimbursement increased significantly from 14.6% in 1991, when Medicare reimbursement for screening mammograms was first initiated, to 18.9% in 1992 (P < .001). The rates in 1993 (17.4%) also increased from the baseline year 1991 (P < .001). However, the observed increases since 1991 have been limited in magnitude.

Low mammography use persists among older women in Connecticut and, in particular, among older black women. The initiation of Medicare reimbursement for screening mammograms in 1991 has had some impact on mammography use although its effects are still limited. Through the use of small area analysis methodology, significant underutilization of mammography in localized areas of the state was identified. These findings have facilitated local outreach interventions. Additional research is needed to understand if health service barriers are contributing to the local variation in rates observed in this study.