State Variation in PSA Screening After the 2011-2012 USPSTF Update
Not only is there immense state-by-state heterogeneity in PSA screening, but there is also marked variation within each state.
The US Preventive Services Task Force (USPSTF) is an authoritative source on disease prevention, and its screening recommendations often greatly impact the practice of medicine. The USPSTF continually considers new evidence in its guidelines, and thus its recommendations are liable to change to reflect the latest research. Since its first edition in 1989, there have been 5 updates to its prostate cancer screening guidelines (Table 1). Screening is performed through PSA, with or without digital rectal examination (DRE). PSA screening has been in wide use since 1992 when both the American Urological Association (AUA) and American Cancer Society recommended annual PSA and DRE screening for men ≥ 50 years old.1 Since then, the medical community has been vigorously debating and researching the benefits vs harms of prostate cancer (PCa) detection and early treatment. PCa screening has become divisive and controversial because even though it is the most commonly diagnosed non-skin cancer in men in the United States, with a lifetime risk for diagnosis of 15.9%, the chance of dying from it is only 2.8%.2
Opponents of screening argue that it leads to a significant “overdiagnosis” and “overtreatment,” as many men with PCa will remain asymptomatic throughout their lifetime, even without treatment. The negative impact of unnecessary screening is not only financial; it also has a negative psychological effect associated with false-positive results that often leads to additional testing and prostate biopsy. Besides, it increases the possibility of diagnosing and treating “indolent” tumors, which is associated with significant patient morbidity, such as urinary incontinence and sexual dysfunction.
Last year, the USPSTF released a draft of its 6th PCa screening recommendation that advised men aged 55–69 years to speak with their doctors about the benefits and harms of screening to reach an informed, personalized decision (grade C), and recommended against screening for men ≥ 70 years old (grade D).3 This is a big change from its 2011–2012 recommendation (a draft statement was released in October 2011 and was finalized in May 2012) that argued against PSA screening for men of all ages (grade D).2
Table 1. Evolution of USPSTF's Recommendation on Prostate Cancer Screening
D for PSA and TRUS; I for DRE*4
I for men < 75 years old; D for men ≥ 75 years old7
C for men 55-69 years old; D for men ≥ 70 years old3
*The USPSTF did not have a grading system in 1989 for its recommendations. Rather, these grades are what their 1989 guidelines would translate to now.
Effects on PCa Screening Practices
Given the recent 2017 USPSTF draft recommendation, it is both timely and important to understand how USPSTF changes affect real world screening practices. Although several other organizations publish PCa screening guidelines, primary care physicians are most greatly influenced by USPSTF recommendations.8 Furthermore, the majority of PSA screening tests are ordered by primary care physicians because they are first-line providers and far outnumber urologists.9 The 2008 USPSTF update that advised against PCa screening for men aged 75 year or older was associated with a decrease in PSA screening in men aged 50 years or older, but especially among those aged 75 years or older. However, studies differ as to whether this decline was significant,10,11,12 suggesting that the 2008 modification did not have an appreciable impact on PCa screening practices.
In contrast, the 2011–2012 USPSTF change that recommended against PCa screening in men of all ages, was associated with a significant decline in PSA screening in men aged 50 years or older across multiple studies.11,13,14 In a study we published in JAMA in 2015, we used data from the National Health Interview Survey (NHIS) and identified 20,757 men aged 50–74 years.14 We found that between 2010 and 2013, PSA screening dropped from 36% to 31% (p < 0.001), a relative percent change of 14%. A nonsignificant decline was also seen in men aged 75 years or older. A similar study by Jemal et al11 also found a reduction in PSA testing among men aged 50 years or older from 37.8% to 30.8%, a relative decrease of 18% between the same years. There was also a nonsignificant decline in PSA screening in the subset of men aged 75 years or older.
These results suggest that unlike the 2008 USPSTF recommendation change, the 2011–2012 update had a substantial effect on screening practices. But why? One likely reason is the greater media coverage on the latter. The USPSTF recommended against PSA screening when many prominent medical societies – including the AUA,15 the American Cancer Society,16 the American Society of Clinical Oncology,17 and the American College of Physicians18 – argued that PSA screening should be individualized and based on shared decision making between doctors and patients, which includes a discussion of risks and benefits, while also integrating patients' personal values and goals of care. This subsequently caused heated debate both in the medical community and the media.
From the national to state level
How does the national reduction in PSA screening after the 2011–2012 update matter for each state? How do we apply this knowledge? Unfortunately, national data often are not representative and sometimes vastly different from regional realities. Therefore, individual state experiences cannot be inferred from purely national analyses. Rather, deeper insights into interstate differences must be investigated to arrive at actionable facts and figures. State-specific data are the first step for state health agencies to address any shortcomings and to set tangible goals. There is limited information in the literature regarding state-by-state differences in PSA screening trends, and therefore it was the focus of our study that was recently published in Urology.19 To the best of our knowledge, this report was the first of its kind to examine the impact of the USPSTF 2011–2012 recommendation on PSA screening on a state-level.
We used data from the Behavioral Risk Factor Surveillance System (BRFSS), the world's largest health-related telephone survey, which collects data from more than 400,000 adults per year in the United States. We selected data from 2012 and 201420 to represent early and late post-USPSTF 2011–2012 recommendation populations, respectively. We identified 222,475 men aged 50 years or older who had self-reported PSA tests within the last year and who did not have the test because of “a prostate problem” or because they “were told [they] had prostate cancer.” Covariates included age group, race or ethnicity, education, income, health insurance status, access to regular health care, marital status, and smoking status. We selected Oklahoma as the reference state because it represented the median PSA screening rate in the pooled 2012 and 2014 BRFSS sample data. We then calculated the relative percent change in PSA screening using Oklahoma as a baseline.
The adjusted nationwide prevalence of PSA screening dropped from 39.4% in 2012 to 36.0% in 2014 (relative change of −8.5%; 95% CI −10.5% to −6.4%; p <0.001). On a state-by-state level, striking differences were detected. Specifically, PSA screening significantly decreased in 34 states and the District of Columbia, and significantly increased in Hawaii and Texas. No significant changes in PSA screening were seen in the remaining 14 states. The maximum relative decrease in PSA screening was observed in Vermont (−26.6%), and the maximum relative increase was observed in Hawaii (+10.2%) – a staggering intranational difference of 36.8% (Figures 1 and 2).
Figure 1. Waterfall plot depicting adjusted and weighted state-by-state percent changes with 95% confidence intervals in self-reported prostate-specific antigen screening prevalence in 222,475 men, within the Behavioral Risk Factor Surveillance System between 2012 and 2014.
Figure 2. Heat map depicting adjusted and weighted state-by-state percent changes in self-reported prostate-specific antigen screening prevalence in 222,475 men, within the Behavioral Risk Factor Surveillance System between 2012 and 2014.
How can there be a 36.8% difference in change in relative PSA screening rates between two states in the same country? How can we determine how much of this difference is even attributable to the USPSTF recommendation? There is likely a multitude of factors that influence screening practices, including a complex gestalt of external factors or “background noise.” This background noise is applicable to all types of disease screenings and is unlikely to affect different screening programs in different ways. To quantify the effect of this background noise, we correlated changes in PSA screening with those of breast cancer screening (mammography) and colorectal cancer screening (fecal occult blood test and sigmoidoscopy or colonoscopy). We selected these screenings because they target similar age groups to that of PSA screening, but their USPSTF guidelines did not change between 2012 and 2014. Again, we used BRFSS data and the same methodology to analyze changes in their screening patterns.
We found that on a state-level there were weakly positive correlations between changes in PSA screening and changes in breast cancer screening (Pearson r = 0.40; p = 0.004), and between changes in PSA screening and changes in colorectal cancer screening (Pearson r = 0.43; p = 0.002). Using the coefficient of determination (R2), 84.0% and 81.5% of the observed changes in PSA screening were not reflected by parallel changes in breast cancer and colorectal cancer screenings, respectively. In other words, only 16%–18.5% of changes in PSA screening were due to background noise.
What's causing these disparities?
In addition to common “background noise,” there are undoubtedly many measurable and unmeasurable determinants of screening that, when taken together, explain this profound state-by-state variation in adoption of USPSTF guidelines. One theory proposed in an insightful editorial comment21 about our study was that states most affected by PCa would be less likely to reduce PSA screening following the 2011–2012 update. Interestingly, most of the states (and the District of Columbia) with the highest mortalities from PCa, such as the District of Columbia, Idaho, North Dakota, and Alaska, were associated with significant declines in PSA screening. In contrast, states with the lowest rates of mortality, like Hawaii and Texas, surprisingly had the highest increases in screening.22
So, what are other possible causes? Race must always be considered as it historically has been and currently still is the basis of many health care disparities, which pervade all medical specialties. In a recent study we published in Urologic Oncology,23 we demonstrated that on a state-level there were significant racial differences in recommended PSA screening and non-recommended PSA screening (those that are done to men with a life expectancy of less than 10 years) across the United States. Again, we looked at BRFSS data in 2012 and 2014, but this time for men aged 40–99 years. The prevalence of PSA screening was 30.7% in non-Hispanic whites (NHWs) vs 28.1% in blacks (p < 0.001). On a state-by-state analysis, NHWs were screened more frequently in 43 states, and blacks were screened more frequently in 8 states. Additionally, non-recommended screening was performed more frequently to NHWs in 19 states, and to blacks in 24 states. Therefore, even though blacks have a greater risk of developing PCa and suffer higher rates of morbidity and mortality from the disease, a larger proportion of NHWs are getting screened, whereas more black men are getting improperly screened.
In a different study we published in The Journal of Urology in 2016,24 we learned that socioeconomic predictors of screening behavior do not apply equally between blacks and NHWs. Analyzing BRFSS data from 2012, we found that all parameters of higher socioeconomic status (education, income, rural vs urban/suburban residence, health insurance, and regular access to a health care provider) were predictive of undergoing PSA screening in NHWs, yet only higher education was significantly associated with screening in black men in multivariate analysis. Additionally, the effect of a physician's recommendation to undergo PSA screening was more pronounced in NHWs (OR 12.98; 95% CI 12.09–13.93) than in blacks (OR 8.43; 95% CI 6.57–10.82).
The influence of a physician's recommendation as well as a discussion of the risks and benefits may be highly influential on a patient's decision to undergo PSA screening. In another study we conducted using 2012 BRFSS data,25 we found that 75% of men were informed of screening benefits, but only 32% were informed of screening risks. After being informed of both, 56% opted for PSA screening if the provider recommended it compared with only 21% when the physician did not recommend it (p < 0.001). Moreover, physicians were likely to give different recommendations depending on whether they were a primary care physician (PCP) or a urologist. In a study by Zavaski et al26 that examined data from the National Ambulatory Medical Care Survey in 2010 and 2012, PSA screening decreased from 36.5% to 16.4% (-20.1%) among primary care visits (adjusted odds ratio, 0.44; 95% CI, 0.24–0.80), whereas it only decreased from 38.7% to 34.5% (4.2%) among urology visits (adjusted odds ratio, 0.60; 95% CI, 0.19–1.84). However, the difference between physician-specific testing practices was not statistically significant (p = 0.63), likely due to a small sample size of 1164 doctor's visits.
In a separate but related study by Colli and Amling27 that used data from BRFSS, the US Census Bureau, and the member directory from the AUA, there were statistically significant inverse correlations between PCa mortality and 3 factors: urologist population density (r = -0.44, p = 0.002), PSA screening rates (r = -0.41, p = 0.003), and living in an urban area (r = -0.33, p = 0.02). The most likely explanation that relates these factors are that urologists (like many other physicians) tend to live in urban areas and are more likely to screen for PCa than primary care physicians, and therefore the mortality from PCa is lower in urban areas because of greater access to medical care.
Consistent with this theory, the 2 states with the highest decrease in PSA screening after the 2011–2012 update, Vermont and Maine, also had the greatest percent rural populations per the US 2010 census, respectively.28 However, in contrast to this theory, Vermont and Maine had the 4th and 10th highest urologist-to-population ratios.29 Hawaii and Texas were the only 2 states that had significant increases in PSA testing, and they had the 7th and 17th greatest percent urban populations in 2010,28 but also had the 16th and 48th greatest urologist-to-population ratios.29 These findings suggest that there is not only wide interstate variation in PSA testing, but also broad intrastate heterogeneity as well, possibly due to pockets of urban populations with higher densities of urologists. There also appears to be a great amount of locational shift year-to-year among urologists. For example, thus far the data provided here on urologist density have been taken from the 2014 edition of AUA's The State of Urology Workforce and Practice in the United States. In 2014, Vermont and Maine had the 4th and 10th highest urologist densities, but in 2016 they had the 18th and 36th highest urologist-to-population ratios.30
Lastly, another explanation for state-level variations is state politics. In yet another recent study, we found that a state's dominant political ideology influences health care policy implementation.31 We revealed a significant correlation between the change in PSA screening rates between 2012 and 2014, and a state's level of political conservativeness. In the most politically conservative states according to Gallup US Daily,32 PSA screening was unchanged (OR = 0.92, 95% CI 0.84–1.00), whereas the least conservative states experienced a significant decline in PSA screening (OR = 0.72, 95% CI 0.64–0.81). Up to 22% of the observed variation in PSA screening trends were attributable to a state's dominant political leaning (R2 = 0.22).
The adjusted nationwide prevalence of PSA screening dropped by 8.5% after the 2011–2012 USPSTF PCa screening update. However, this figure masks substantial differences on a state-level, which ranged from a 26.6% relative decrease to a 10.2% relative increase – a difference of 36.8%. There are numerous interacting determinants of screening that form a complex and dynamic effect on the prevalence of PSA screening in each state. Not only is there immense state-by-state heterogeneity, but there is also marked variation within each state. Among others, factors that account for these startling discrepancies in implementation of USPSTF 2011–2012 guidelines include race, socioeconomic factors, living in an urban vs rural area, discussion of the risks vs benefits of screening, physician's recommendation about screening, type of physician (PCP vs urologist), and state politics.
Moving forward there is still a great deal of progress to be made in identifying the underlying mechanisms responsible for these differences. Further research is needed to address these inter- and intra- state disparities and to provide greater equality in care throughout the United States. Given this latest 2017 USPSTF draft recommendation, we are interested to see its effects on PSA screening practices on the national, state, and county levels.
- Potosky AL, Miller BA, Albertsen PC, Kramer BS. The role of increasing detection in the rising incidence of prostate cancer. JAMA. 1995;273: 548-552.
- Moyer VA. U.S. Preventive Services Task Force. Screening for prostate cancer: U.S. Preventive Services Task Force recommendation statement. Ann Intern Med. 2012;157:120-134.
- Bibbins-Domingo K, Grossman DC, Curry SJ. The US Preventive Services Task Force 2017 draft recommendation statement on screening for prostate cancer: An invitation to review and comment. JAMA. 2017;317:1949–1950. doi:10.1001/jama.2017.4413
- U.S. Preventive Services Task Force. Guide to clinical preventive services: An assessment of the effectiveness of 169 interventions. Baltimore, MD: Williams & Wilkins, 1989.
- Woolf SH. U.S. Preventive Services Task Force. Guide to clinical preventive services: Report of the U.S. Preventive Services Task Force. 2nd edition. Baltimore (MD): Williams & Wilkins; 1996.
- U.S. Preventive Services Task Force. Screening for prostate cancer: recommendations and rationale. Am Fam Physician. 2003 Feb 15;67:787-792.
- U.S. Preventive Services Task Force. Screening for prostate cancer: U.S. Preventive Services Task Force recommendation statement. Ann Intern Med. 2008 Aug 5;149:185-191.
- Tasian GE, Cooperberg MR, Potter MB, et al. PSA screening: determinants of primary-care physician practice patterns. Prostate Cancer Prostatic Dis. 2012;15:189-194. doi: 10.1038/pcan.2011.59. Epub 2011 Nov 29.
- Cohn JA, Wang CE, Lakeman JC, et al. Primary care physician PSA screening practices before and after the final U.S. Preventive Services Task Force recommendation. Urol Oncol. 2014;32:41.e23-e30. doi: 10.1016/j.urolonc.2013.04.013. Epub 2013 Aug 2.
- Scosyrev E, Wu G, Golijanin D, Messing E. Prostate-specific antigen testing in older men in the USA: data from the behavioral risk factor surveillance system. BJU Int. 2012;110:1485-1490.
- Jemal A, Fedewa SA, Ma J, et al. Prostate cancer incidence and PSA testing patterns in relation to USPSTF screening recommendations. JAMA. 2015;314:2054-2061.
- Howard DH, Tangka FK, Guy GP, et al. Prostate cancer screening in men ages 75 and older fell by 8 percentage points after task force recommendation. Health affairs (Project Hope). 2013;32:596-602. doi:10.1377/hlthaff.2012.0555.
- Drazer MW, Huo D, Eggener SE. National prostate cancer screening rates after the 2012 US Preventive Services Task Force recommendation discouraging prostate-specific antigen-based screening. J Clin Oncol. 2015;33:2416-2423.
- Sammon JD, Abdollah F, Choueiri TK, et al. Prostate-specific antigen screening after 2012 US Preventive Services Task Force recommendations. JAMA. 2015;314:2077-2079.
- Carter HB, Albertsen PC, Barry MJ, et al. Early detection of prostate cancer: AUA guideline. J Urol. 2013;190:419-426.
- Wolf AM, Wender RC, Etzioni RB, et al. American Cancer Society Prostate Cancer Advisory Committee. American Cancer Society guideline for the early detection of prostate cancer: update 2010. CA Cancer J Clin. 2010;60:70-98. doi: 10.3322/caac.20066. Epub 2010 Mar 3.
- Basch E, Oliver TK, Vickers A, et al. Screening for prostate cancer with prostate-specific antigen testing: American Society of Clinical Oncology provisional clinical opinion. J Clin Oncol. 2012;30:3020-3025
- Qaseem A, Barry MJ, Denberg TD, et al. Screening for prostate cancer: a guidance statement from the Clinical Guidelines Committee of the American College of Physicians. Ann Intern Med. 2013;158:761-769.
- Vetterlein MW, Dalela D, Sammon JD, et al. State-by-state variation in prostate-specific antigen screening trends following the 2011 United States Preventive Services Task Force Panel update. Urology. 2017 Oct 19. pii: S0090-4295(17)31093-2. doi: 10.1016/j.urology.2017.08.055. [Epub ahead of print]
- Centers for Disease Control and Prevention (CDC). Behavioral Risk Factor Surveillance System Survey Data. Atlanta, Georgia: U.S. Department of Health and Human Services, Centers for Disease Control and Prevention, 2012 and 2014.
- Krane LS. Editorial comment. doi: https://doi.org/10.1016/j.urology.2017.08.057
- U.S. Cancer Statistics Working Group. United States Cancer Statistics: 1999-2014 Incidence and Mortality Web-based Report. Atlanta (GA): Department of Health and Human Services, Centers for Disease Control and Prevention, and National Cancer Institute; 2017. Available at: http://www.cdc.gov/uscs. Accessed on Feb 10, 2018.
- Jindal T, Kachroo N, Sammon J, et al. Racial differences in prostate-specific antigen-based prostate cancer screening: state-by-state and region-by-region analyses. Urol Oncol. 2017;35:460.e9-460.e20.
- Sammon JD, Dalela D, Abdollah F, et al. Determinants of prostate specific antigen screening among black men in the United States in the contemporary era. J Urol. 2016;195:913-918.
- Pucheril D, Dalela D, Sammon J, et al. The influence of physician recommendation on prostate-specific antigen screening. Urol Oncol. 2015;33:e421-e427.
- Zavaski ME, Meyer CP, Sammon JD, et al. Differences in prostate-specific antigen testing among urologists and primary care physicians following the 2012 USPSTF recommendations. JAMA Intern Med. 2016;176:546-547.
- Colli JL, Amling CL. Prostate cancer mortality rates compared to urologist population densities and prostate-specific antigen screening levels on a state-by-state basis in the United States of America. Prostate Cancer Prostatic Dis. 2008;11(3):247-51. doi: 10.1038/pcan.2008.7. Epub 2008 Feb 12.
- U.S. Census Bureau. Lists of Population, Land Area, and Percent Urban and Rural in 2010 and Changes from 2000 to 2010: Percent urban and rural in 2010 by state. Retrieved on Feb 10, 2018 from https://www.census.gov/geo/reference/ua/urban-rural-2010.html.
- American Urological Association, The State of Urology Workforce and Practice in the United States 2014. Linthicum, Maryland, U.S.A, May 15, 2015. Retrieved on Feb 10, 2018 from https://www.AUAnet.org/common/pdf/research/census/State-Urology-Workforce-Practice-US.pdf.
- American Urological Association, The State of Urology Workforce and Practice in the United States 2016. Linthicum, Maryland, U.S.A, April 11, 2017. Retrieved on Feb 10, 2018 from https://www.AUAnet.org/common/pdf/research/census/State-Urology-Workforce-Practice-US.pdf.
- Sammon J, Serrel E, Vetterlein M, et al. PD58-12 PSA screening at the intersection of politics and policy. J Urol. 2017;197:e1132. http://dx.doi.org/10.1016/j.juro.2017.02.2627.
- Gallup U.S. Daily. Retrieved on April 4, 2016 from http://www.gallup.com.