P1-11-07: Impact of Reduction in Cost-Sharing on Screening Mammography Utilization among Rural U.S. Women.

Abstract

Abstract Background: Early detection of breast cancer through mammography screening leads to earlier stage at diagnosis and improved survival. For reasons that are poorly understood, in recent years the rate of screening has demonstrated periods of decline, and screening has proved to be less common in rural compared to urban areas of the U.S. In 2006, the National Rural Electric Cooperative Association (NRECA) which provides health care to over 100,000 electrical workers and their families in primarily rural areas of the U.S. eliminated copayments for screening mammography in an effort to boost screening rates. We conducted a population based analysis of screening utilization to determine the impact of this policy initiative. Methods: Using the NRECA insurance database, all women aged 40 to 64 with no prior history of breast cancer or DCIS (based on ICD-9 codes) were identified and we evaluated claims data on annual screening mammography utilization (SMU) between 1999 and 2009 stratified by age group in 5 year intervals. Changes in SMU over time were assessed focusing on the periods before and after the policy change in January 2006. We also evaluated diagnosis of breast cancer and receipt of mastectomy and chemotherapy as a potential proxy for more advanced disease at diagnosis. Descriptive statistics were estimated and the mammography rate was fitted on years using the identity link (proc genmod in SAS). In order to test the impact of the 2006 change in cost-sharing on the trend in mammography rate, we introduced change point terms in slope and intercept to the linear model. Chi-squared test for 2×2 tables was used to compare SMU rates between two consecutive years for each age group. All p-values are two-sided. Results: During this period, a mean of 20,825 women aged 40 to 64 each year received health insurance through NRECA. SMU increased from 38.1% in 1999 to 49.5% in 2009. Analyzing SMU before and after the change in cost sharing policy demonstrates a significant change in the rate of screening at the 2006 intercept (p = 0.0275) although the slope of year to year change in screening rate did not change. In stratified analysis there was a significant change in SMU between 2005 and 2006 for all age groups. In the 4 years prior to the NRECA change in policy, 554 women were diagnosed with breast cancer and 40 underwent mastectomy and chemotherapy. In comparison, from 2006 to 2009, only 20 women out of 540 with newly diagnosed breast cancer underwent such therapy (7.2% prior to policy change vs. 3.7% following, P = 0.01). Conclusion: The impact of health plan benefits changes can be evaluated among a primarily rural population of women aged 40 to 64 using the novel NRECA database. Annual SMU remained low, but improved following elimination in copayments suggesting that financial barriers impact screening. Multiple factors may explain changes in SMU and treatment intensity over time. However, cost-sharing for high value health care services may have unintended negative consequences. Further evaluation of this database is planned to evaluate biannual screening rates, correlation with sociodemographic factors, impact of recent controversy over screening guidelines and additional barriers to screening utilization in this rural population. Citation Information: Cancer Res 2011;71(24 Suppl):Abstract nr P1-11-07.