Are there regional tendencies toward controversial screening practices? A study of prostate and breast cancer screening in a Medicare population

作者：Raffin Eric^*; Onega Tracy; Bynum Julie; Austin Andrea; Carmichael Donald; Bronner Kristen; Goodney Philip; Hyams Elias S

来源：Cancer Epidemiology, 2017, 50: 68-75.

DOI：10.1016/j.canep.2017.07.015

• Prostate cancer
• Breast cancer
• PSA
• Screening
• Mammography

Introduction: Prostate and breast cancer screening in older patients continue to be controversial. Balancing the desire for early detection with avoidance of over-diagnosis has led to competing and contradictory guidelines for both practices. Despite similarities, it is not known how these screening practices are related at the regional level. In this study, we examined how screening PSA and mammography are related within healthcare regions, and, to better understand what may be driving these practices, whether they are associated with local intensity of care.
Methods: We performed a retrospective cross-sectional study of fee-for-service Medicare beneficiaries in 2012. For each of 306 hospital referral regions (HRRs), we calculated rates of PSA screening for men aged >= 68 years, as well as rates of screening mammography for women aged >= 75 years, adjusted for age and race. Additionally, we determined regional rates of "healthcare intensity", including spending on tests and procedures, and intensity of end-of-life care. Pearson correlations of adjusted rates were calculated within HRRs.
Results: The mean adjusted rate of PSA screening was 22%. The mean age of screened and unscreened patients was 75.0 and 77.4 years, respectively (p < 0.0001). The mean adjusted rate of screening mammography was 23%; mean ages of screened and non-screened women were 79.95 and 83.67, respectively (p < 0.0001). HRR-level PSA screening rates were independent of screening mammography rates (r = 0.06, p = 0.31). PSA screening rates were associated with spending on testing and procedures (r = 0.42, p < 0.0001) and various measures of intensity of EOL care (e.g. r = 0.40, p < 0.0001 for mechanical ventilator use). Screening mammography had low correlation with both health care spending and EOL care intensity measures (all r-values < 0.3).
Conclusions: Regional rates of PSA screening rates were independent of screening mammography, thus these practices appear to be driven by different factors. Unlike mammography, PSA screening was associated with local enthusiasm for testing and treatment. Efforts to reduce over-testing should contemplate these practices differently, and future research should examine the factors motivating these screening practices.

• 出版日期2017-10