The effect of a multifaceted physician office-based intervention on older women's mammography use.

BACKGROUND: In response to identified low mammography use among older women in three geographic areas in Connecticut, a physician office-based mammography intervention was initiated under the Health Care Financing Administration's Health Care Quality Improvement Program. OBJECTIVE: To evaluate the intervention's impact on older women's mammography use. DESIGN: A quasi-experimental design comparing mammography rates for women in the intervention program with a randomly selected control sample. SETTING: Community-based physician offices. PATIENTS: Female Medicare beneficiaries aged 65 to 74 years seen by participating and control physicians for at least one primary care visit in 1995 (baseline) and 1996 (follow-up). In the baseline period, 1720 women in the intervention sample and 2761 women in the control sample were included in the study. INTERVENTION: The recruitment strategies included the use of physician opinion leaders and modified academic detailing. The multifaceted intervention incorporated patient education, physician reminders, and audit-with-feedback MEASUREMENTS: Biennial mammography rates. Patient adherence to physician mammography referral was evaluated in a restricted cohort of women selected from the intervention sample. RESULTS: The mammography rate for the intervention sample increased from 62.7% (baseline) to 73.1% (follow-up), (P<.001), whereas the control sample's rate remained essentially unchanged (68.3 to 69.5%), (P = .34). The intervention patients were 48% more likely than controls to experience an increase in biennial mammography use (OR = 1.48; 95% CI, 1.22-1.79) after adjustment for patient race and income and physician gender, specialty, and age. The proportion of women who adhered to their physicians' mammography referral was 70.6%. CONCLUSIONS: These data demonstrate the effectiveness of a multifaceted intervention program administered in the setting of community physician practices. The relatively low rate of patients' acceptance of their physicians' mammography recommendations has identified the need to address more effectively older women's concerns about mammography screening.

The impact of a physician intervention program on older women's mammography use.

The Mammography Optimum Referral Effort (MORE) is a physician office-based intervention program initiated by the Connecticut Peer Review Organization (CPRO) to increase mammography use among older women in Connecticut. Three locales in the state were targeted for the MORE intervention based on identified low mammography rates in women aged 65 years and older. Thirty-seven physicians participated from March 1, 1996, to August 31, 1996. Annual mammography rates were derived by merging Medicare Part B mammography claims with a database from the Connecticut Tumor Registry. This strategy allowed us to exclude women with a prior history of breast cancer from the analysis, in order to estimate screening rates. The MORE intervention was associated with an absolute increase of 5.9%, which represents a relative increase of 15.4%, in annual mammography use. Our findings suggest that a multifaceted physician intervention is capable of increasing mammography use among older women.

Mammography underutilization among older women in Connecticut.

OBJECTIVES: 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. DESIGN: 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. PARTICIPANTS: Female Medicare beneficiaries 65 years and older with Part B coverage residing in Connecticut during the study period. MEASUREMENTS: 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. MAIN RESULTS: 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. CONCLUSIONS: 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.