Contemporary Nationwide Patterns of Self-reported Prostate-Specific Antigen Screening

Routine screening for prostate cancer using prostate-specific antigen (PSA) is a widely contested practice, and recommendations have recently changed dramatically. In October 2011, the US Preventative Services Task Force recommended against screening in any age group,¹ yet current nationwide patterns of PSA screening are largely unknown. We sought to elucidate contemporary PSA screening prevalence with a focus on heterogeneity among states and across age groups. We examine data from the 2012 Behavioral Risk Factor Surveillance System (BRFSS).

The BRFSS is the world’s largest continuously conducted health survey, a joint initiative of the Centers for Disease Control and Prevention and US states. Male respondents 50 years or older without a history of prostate cancer or prostate problem who reported PSA testing within the 12 months preceding the 2012 BRFSS survey were considered to have undergone screening. The 2012 survey was conducted between January 2, 2012, and February 12, 2013. Complex-samples logistic regression analysis incorporating age, race and/or ethnicity, education, income, residence location, insurance status, access to regular health care, and marital status was used to estimate an individual’s predicted probability of undergoing PSA screening. Individual probabilities were then normalized to the 2012 BRFSS screened population to derive state-specific estimates of screening prevalence.

In 2012, 114 544 unique responses from men 50 years or older were captured by the BRFSS, a weighted estimate of 46.24 million men, of whom 17.16 million (37.1%) reported undergoing PSA screening. Access to regular health care was most strongly associated with higher rates of screening (odds ratio [OR], 3.00 [95% CI, 2.69-3.34]). Additional predictors included income greater than $75 000 (OR, 1.91 [95% CI, 1.67-2.20]), college education (OR, 1.90 [95% CI, 1.70-2.12]), health insurance (OR, 1.83 [95% CI, 1.60-2.08]), and age 70 to 74 years (OR, 2.53 [95% CI, 2.29-2.79]) (see eTable in the Supplement). After adjustment for covariates, the estimated prevalence of self-reported PSA screening was highest in Hawaii (59.4%) and lowest in New Hampshire (24.5%) (Figure 1). The prevalence of PSA screening was highest in older men aged 65 to 69 years (48.4%) and 70 to 74 years (48.5%). Men aged 50 to 54 years were the least likely to report PSA screening (25.0%) (Figure 2).

Whereas changes in sampling methodology preclude direct comparison with prior years of the BRFSS, our data show that the effect of previous guidelines recommending against the routine screening of elderly men has been minimal at best.² Prior work exploring PSA screening trends following earlier US Preventative Services Task Force recommendations against screening in men older than 75 years demonstrated no significant reduction, and yet men aged 75 to 79 years remain the third most likely age group to undergo screening in the year preceding the survey (weighted estimate, 45.7% [95% CI, 43.6%-47.9%]).³ Equally remarkable is the low rate of PSA screening (weighted estimate, 25.0% [95% CI, 23.8%-26.2%]) among men between 50 and 54 years old, for whom screening has previously been recommended by several professional organizations. These findings likely reflect both the considerable disagreement among experts and the conflicting recommendations. Taken together, these results suggest that national guidelines have had limited effect on clinical practice among health care providers.

The degree of heterogeneity in state-by-state PSA screening prevalence is another concerning and surprising study finding. Evidence from the colorectal and breast cancer screening literature suggests that state-by-state and regional variability is expected but not to such a pronounced extent. For example, regional variability in colorectal screening rates is only 7.5%.⁴ It is alarming that the prevalence of PSA screening can double from one state to the next. Whereas the causes of this heterogeneity may mirror those in mammography, including variability in the availability of large university hospitals, geographic density of providers, levels of insurance coverage, and income,⁵ the effect of physician preferences on the odds of screening is far more pronounced for PSA, relative to other screening tests.⁶

Limitations of our study include recall and nonresponse bias. Inaccuracies in self-reported PSA screening have been found to reflect underreporting, when compared with medical record extraction. Also, only individuals with telephones were sampled, exerting an uncertain bias.

Corresponding Author: Quoc-Dien Trinh, MD, Dana-Farber Cancer Institute, Harvard Medical School, 45 Francis St, ASB II-3, Boston, MA 02115 (trinh.qd@gmail.com).

Published Online: September 1, 2014. doi:10.1001/jamainternmed.2014.4117.

Author Contributions: Drs Sammon and Pucheril served as co–first authors, each with equal contribution to the manuscript. Drs Trinh and Sammon had full access to all of the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.

Study concept and design: Sammon, Pucheril, Kibel, Kantoff, Menon, Trinh.

Acquisition, analysis, or interpretation of data: Sammon, Pucheril, Diaz, Trinh.

Drafting of the manuscript: Sammon, Pucheril, Trinh.

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

Statistical analysis: Sammon, Diaz, Trinh.

Obtained funding: Menon.

Administrative, technical, or material support: Pucheril, Kantoff, Menon.

Study supervision: Kibel, Trinh.

Conflict of Interest Disclosures: Dr Kibel consults for Sanofi-Aventis, Dendreon, and Myriad Genetics. No other disclosures are reported.

Funding/Support: This work is supported by the Professor Walter Morris-Hale Distinguished Chair in Urologic Oncology at Brigham and Women’s Hospital.

Role of the Sponsor: The funder had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.

Additional Contributions: Akshay Sood, MD, Vattikuti Urology Institute Center for Outcomes Research, Analytics, and Evaluation, Henry Ford Health System; Simon P. Kim, MD, Department of Urology, Yale University; Jim C. Hu, MD, Department of Urology, David Geffen School of Medicine at UCLA; Paul L. Nguyen, MD, Department of Radiation Oncology, Dana-Farber Cancer Institute, Brigham and Women's Hospital, Harvard Medical School; and Maxine Sun, PhD, Cancer Prognostics and Health Outcomes Unit, University of Montreal Health Center, were instrumental in creating an earlier manuscript draft of this Research Letter. They were not compensated for their contributions.