Prevalence of Nonrecommended Screening for Prostate Cancer and Breast Cancer in the United States: A Nationwide Survey Analysis

Existing guidelines acknowledge the risks of overdiagnosis and overtreatment associated with early detection of prostate cancer and breast cancer and recommend against screening for these tumors in individuals with limited life expectancy.^1,2 The cost to the US health care system related to overdiagnosis may be as high as $1.2 billion annually.³ That finding, combined with the aging population and an expected surge of older individuals with prostate cancer and breast cancer in upcoming years,⁴ indicates that a contemporary nationwide and state-by-state assessment of the prevalence of nonrecommended screening for prostate cancer and breast cancer is essential and timely.

All individuals responding to the Behavioral Risk Factors Surveillance System survey conducted between January and December 2012 who were 65 years or older and residing in the United States or District of Columbia were included. The median combined response rate was 45.2%, comparable with that in similar surveys.⁵ The primary outcome was nonrecommended screening, defined as receipt of prostate-specific antigen (PSA) testing or mammography in individuals with a life expectancy of less than 10 years.⁶ This cutoff was chosen based on previous reports.^1,2 Receipt of PSA testing was coded as performed if men answered yes to the question, “Have you ever had a PSA test?” and answered “Within the past year” to the question, “How long has it been since you had your last PSA test?” Receipt of mammography was coded as performed if women answered yes to the question, “Have you ever had a mammogram?” and “Within the past year” to the question, “How long has it been since you had your last mammogram?” Men who underwent PSA testing owing to a previous diagnosis of prostate cancer were excluded. Those previously diagnosed with breast cancer or prostate cancer were also excluded. A total of 149 514 individuals were identified (weighted n = 43 586 000).

For all point estimates, 95% CIs and P values were calculated using the Complex Samples Package for SPSS (SPSS, Inc). The prevalence of self-reported screening in the past year was estimated for 2012 and quantified for each state. With the state as the unit of analysis, Pearson correlation coefficient was used to test the association between nonrecommended screenings for prostate cancer and breast cancer. Complex sample multivariable logistic regression models assessed the odds of nonrecommended screening for prostate cancer and breast cancer.

All statistical analyses were performed using SPSS, version 21, and the R statistical package (R Foundation for Statistical Computing), with a 2-sided significance set at P < .05. Data analysis was conducted from July 27, 2014, to October 30, 2015. Henry Ford Hospital Institutional Review Board approval was waived in accordance with institutional regulation when dealing with deidentified, previously collected data.

Of 149 514 individuals 65 years or older (weighted n = 43 586 000), 76 419 (51.1%) had a PSA test or mammography in the past year; 23 532 (30.8%) of those individuals had a life expectancy of less than 10 years, corresponding to an overall rate of nonrecommended screening of 15.7% (23 532 of 149 514 individuals). Nonrecommended screening varied widely across the United States (Figure⁵ and Table⁵). In multivariable analysis, the state was independently associated with nonrecommended screening for both prostate cancer and breast cancer (P < .001 for both).

Overall, the rate of nonrecommended screening for prostate cancer and breast cancer was high, seen in 15.7% of individuals 65 years or older. Significant state-by-state variation in the rate of nonrecommended screening was observed. Moreover, nonrecommended screening at the state-by-state level had a correlation of 40% between the 2 forms of screening (P = .01). Therefore, a state with a high rate of nonrecommended screening for prostate cancer is likely to have a high rate of nonrecommended screening for breast cancer and vice versa. To our knowledge, our study is the first to report the extent and variation of nonrecommended screening across states, fluctuating from 11.6% (95% CI, 10.4%-13.1%) in Colorado to 20.2% (95% CI, 17.9%-22.7%) in Georgia. Efforts should be deployed to reduce nonrecommended screening in states with a high rate of nonrecommended screening. This effort may avoid significant harms to many individuals and improve the cost efficiency of screening initiatives. Limitations of the study include the possible overestimation of life expectancy owing to unmeasured comorbidities, as well as the potential inclusion of some men and women previously diagnosed, treated, or observed for prostate cancer and breast cancer.

Corresponding Author: Firas Abdollah, MD, Center for Outcomes Research, Analytics and Evaluation, Vattikuti Urology Institute, Henry Ford Health System, 2799 W Grand Blvd, Detroit, MI 48202 (firas.abdollah@gmail.com).

Published Online: January 21, 2016. doi:10.1001/jamaoncol.2015.5871.

Author Contributions: Dr Abdollah had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Study concept and design: Abdollah, Sammon, Choueiri, Menon, Weissman, Trinh.

Acquisition, analysis, or interpretation of data: Abdollah, Sun, Sammon, Choueiri, Trinh.

Drafting of the manuscript: Abdollah, Sun, Sammon, Trinh.

Critical revision of the manuscript for important intellectual content: All authors.

Statistical analysis: Abdollah, Sun, Sammon, Choueiri, Trinh.

Administrative, technical, or material support: Sammon, Choueiri, Menon, Weissman.

Study supervision: Choueiri, Menon, Weissman.

Conflict of Interest Disclosures: Dr Abdollah is a consultant for GenomeDx biosciences. Dr Trinh is supported by the Professor Walter Morris-Hale Distinguished Chair in Urologic Oncology at Brigham and Women’s Hospital. No other disclosures were reported.

Additional Contributions: Paul L. Nguyen, MD, Department of Radiation Oncology, Brigham and Women’s Hospital and Dana-Farber Cancer Institute, Harvard Medical School, and Adam S. Kibel, MD, and Mark A. Preston, MD, Division of Urologic Surgery and Center for Surgery and Public Health, Brigham and Women’s Hospital and Dana-Farber Cancer Institute, Harvard Medical School, assisted with study supervision. Julian Hanske, MD, Center for Surgery and Public Health, Brigham and Women’s Hospital, Harvard Medical School, assisted with data acquisition. Akshay Sood, MD, Center for Outcomes Research, Analytics and Evaluation, Vattikuti Urology Institute, Henry Ford Health System, assisted with drafting of the manuscript. None of these individuals were compensated for their contribution.