Article Tools
ORIGINAL REPORTS
Rapid Communications
Article Tools
National Prostate Cancer Screening Rates After the 2012 US Preventive Services Task Force Recommendation Discouraging Prostate-Specific Antigen–Based Screening
Abstract
In 2012, the US Preventive Services Task Force (USPSTF) discouraged prostate-specific antigen (PSA) –based prostate cancer screening. Previous USPSTF recommendations did not appreciably alter prostate cancer screening. Therefore, we designed a trend analysis to determine the population-based impact of the 2012 recommendation.
The nationally representative National Health Interview Survey was used to estimate the proportion of men age 40 years and older who saw a physician and were screened for prostate cancer in 2013. An externally validated 9-year mortality index was used to analyze screening rates based on remaining life expectancy. Screening rates from 2005, 2010, and 2013 were compared using logistic regression.
PSA-based screening did not significantly change from 2010 to 2013 among 40- to 49-year-old men (from 12.5% to 11.2%; P = .4). Screening rates significantly declined in men age 50 to 59 years (from 33.2% to 24.8%; P < .01), age 60 to 74 years (from 51.2% to 43.6%; P < .01), and age 75 years or older (from 43.9% to 37.1%; P = .03). A large percentage of men were screened for prostate cancer despite a high risk (> 52%) of 9-year mortality, including approximately one third of men older than age 75 years. Approximately 1.4 million men age 65 years or older with a high risk (> 52%) of 9-year mortality were screened in 2013.
Prostate cancer screening significantly declined among men older than age 50 years after the 2012 USPSTF guideline discouraging PSA-based screening. A significant proportion of men continue to be screened despite a high risk of 9-year mortality, including one third of men age 75 years and older.
The debate over prostate-specific antigen (PSA)– based prostate cancer screening intensified after two randomized screening trials reported conflicting results on prostate cancer–specific mortality.1,2 The tenor of the controversy heightened after the 2012 US Preventive Services Task Force (USPSTF) recommendation discouraging PSA screening for asymptomatic men.3 That recommendation differed from guidelines published by the American Cancer Society and the American Urological Association, both of which endorsed consideration of screening based on age and estimated remaining life expectancy.4,5
Although many organizations publish cancer screening recommendations, primary care physicians in the United States have cited the USPSTF as the most influential.3–10 Despite this, we demonstrated that a large proportion of men in the United States are screened for prostate cancer even without meeting criteria defined by the USPSTF and other organizations. For example, guidelines have encouraged physicians to only engage in shared decision making with men who have at least 10 years of remaining life expectancy.4–8 Yet, the prostate cancer screening rate among men age 70 years or older with a greater than 48% probability of 5-year mortality was 31% in 2005, representing approximately 777,000 men screened despite a high risk for overdiagnosis and overtreatment.11 In addition, screening rates among men age 65 years and older with a high (> 52%) likelihood of 9-year mortality did not differ before and after the 2008 USPSTF recommendation, despite that recommendation's emphasis on limiting screening to men with at least 10 years of remaining life expectancy.12 Similarly, the 2008 USPSTF guidelines discouraging screening in men older than age 75 years did not appreciably alter screening rates.13
Given the public health implications, we conducted a population-based trend analysis to determine the effects of the highly publicized 2012 USPSTF recommendation discouraging prostate cancer screening. Widespread national and social media coverage followed the initial release of the recommendation in October 2011.14–19 We used a large, nationally representative survey, the National Health Interview Survey (NHIS), to investigate three questions regarding prostate cancer screening. First, we estimated the proportion of men age 40 years and older screened in 2013. Second, we used an externally validated 9-year mortality index to analyze screening rates based on estimated remaining life expectancy.20 Finally, we compared screening rates in 2005, 2010, and 2013 to identify changes in national-level screening. All analyses were limited to men who saw a physician in the year before the survey.
This study was exempt from institutional review board review. The NHIS is a face-to-face, computer-assisted, cross-sectional survey performed on a rolling, annual basis in the United States. It is one of the major data collection programs from the Centers for Disease Control and Prevention and excludes active-duty military personnel as well as citizens residing in nursing homes, long-term care facilities, or correctional facilities. The survey oversamples minorities, and a single individual is unlikely to be sampled in multiple years. Employees from the US Census Bureau administer the survey, and results are released with sample weights based on population data. These weights may be used to extrapolate results to the US population.21
The NHIS typically releases cancer screening data every 5 years. The 2013 release, however, contained data regarding prostate cancer screening.21 We extracted data from the 2005, 2010, and 2013 NHIS databases for this study.
We included men age 40 years and older and excluded men with a history of prostate cancer. We excluded men who did not know what a PSA test was or who had not seen a physician in the year before the survey. The NHIS asked all men about the purpose of their most recent PSA test (ie, as “part of a routine examination,” “because of a problem,” “other reason,” or “don't know”). The dependent variable of interest was a PSA test as “part of a routine examination,” which was considered to be a test for screening purposes.
We incorporated sampling weights into all analyses using the strata and primary sampling units in the database. We determined race/ethnicity, highest education level, marital status, colonoscopy utilization, body mass index, tobacco use, personal cancer history, and self-reported health status. We investigated age-based prostate cancer screening by dividing the cohort into 5-year subgroups. We determined the proportion of each group who had a PSA test in the year before the 2005, 2010, and 2013 surveys.
Next, we used an externally validated mortality index designed specifically for the NHIS to calculate each man's predicted risk of death in the 9 years after the survey.20 The index has only been validated in men age 65 years and older; thus, we limited the analysis investigating screening rates by predicted 9-year mortality risk to this age range. We divided the men into the following two groups: healthier men with a ≤ 52% predicted risk of 9-year mortality and less healthy men with a more than 52% predicted 9-year mortality. We stratified these men into age groups (65 to 69, 70 to 74, and ≥ 75 years) and calculated screening rates based on predicted 9-year mortality.
We performed univariable analyses using logistic regressions to investigate the association between PSA testing and factors potentially associated with screening in the 2005, 2010, and 2013 surveys. We included men age 65 years and older who visited a physician in the year prior and analyzed self-reported race, highest education level, marital status, alcohol use, tobacco use, colonoscopy utilization, frequency of light or moderately strenuous exercise, body mass index, self-reported personal health status, ability to independently perform activities of daily living, walking ability, chronic anxiety or depression, predicted 9-year mortality risk, nation of origin, diabetes, chronic obstructive pulmonary disease, previously diagnosed cancer (other than prostate cancer), and number of nights hospitalized during the previous year. We constructed two multivariable logistic regression models using variables associated with prostate cancer screening in univariable analyses (P < .05). Model 1 included all variables significant in univariable analyses. Because many component variables in the mortality index independently predicted screening, we excluded predicted 9-year mortality to reduce the risk of colinearity. Model 2 included predicted 9-year mortality risk but omitted variables used to calculate the 9-year mortality risk variable. Both models included men age 65 and older in the 2005, 2010, and 2013 surveys and excluded men who had not seen a physician in the year prior. Finally, we investigated variables associated with screening changes from 2010 to 2013 by including an interaction-with-age term in logistic regressions (interaction analysis). We used Stata/SE version 13.0 (StataCorp, College Station, TX) and considered a P < .05 to be statistically significant.
|
| Characteristic | No. of Respondents (%) | ||
|---|---|---|---|
| 2005 Survey (n = 8,490) | 2010 Survey (n = 7,326) | 2013 Survey (n = 9,803) | |
| Race/ethnicity | |||
| Non-Hispanic white | 6,045 (77.1) | 4,548 (73.8) | 6,424 (72.0) |
| Non-Hispanic black | 1,022 (9.3) | 1,152 (9.7) | 1,353 (9.7) |
| Hispanic | 1,073 (9.1) | 1,110 (11.2) | 1,326 (11.9) |
| Asian | 236 (3.2) | 410 (4.1) | 505 (4.8) |
| Other | 114 (1.3) | 106 (1.2) | 195 (1.5) |
| Highest education level | |||
| Some high school | 1,558 (15.9) | 1,359 (15.2) | 1,702 (14.5) |
| High school diploma | 2,434 (29.9) | 1,997 (27.5) | 2,579 (25.8) |
| Some college | 2,072 (24.8) | 1,869 (25.9) | 2,587 (27.0) |
| College degree | 2,328 (29.4) | 2,061 (31.4) | 2,886 (32.8) |
| Marital status | |||
| Not married | 3,149 (23.9) | 2,894 (26.2) | 4,040 (27.4) |
| Married/living with partner | 5,290 (76.1) | 4,419 (73.8) | 5,738 (72.6) |
| Colonoscopy history | |||
| Never had colonoscopy | 4,930 (77.9) | 3,746 (69.2) | 5,137 (69.0) |
| Colonoscopy longer than 10 years ago | 162 (2.5) | 171 (3.1) | 299 (4.0) |
| Colonoscopy within last 10 years | 1,272 (19.6) | 1,451 (27.7) | 2,057 (27.0) |
| Body mass index | |||
| Underweight/normal | 2,344 (26.9) | 1,888 (24.7) | 2,418 (23.9) |
| Overweight | 3,766 (45.7) | 3,145 (44.4) | 4,224 (44.1) |
| Obese | 2,154 (27.5) | 2,152 (30.9) | 2,997 (32.1) |
| Tobacco use | |||
| Never smoker | 3,653 (44.5) | 3,373 (47.1) | 4,561 (48.0) |
| Former smoker | 2,884 (34.3) | 2,402 (33.6) | 3,293 (33.4) |
| Current smoker | 1,861 (21.2) | 1,491 (19.3) | 1,904 (18.6) |
| Personal cancer history | |||
| No personal history of cancer | 7,835 (92.7) | 6,718 (91.9) | 8,956 (91.9) |
| Personal history of cancer | 645 (7.3) | 598 (8.1) | 835 (8.1) |
| Self-reported personal health status | |||
| Excellent/very good | 4,429 (54.7) | 3,696 (54.2) | 4,970 (53.2) |
| Good | 2,512 (29.0) | 2,236 (29.1) | 2,894 (29.2) |
| Fair/poor | 1,542 (16.3) | 1,391 (16.7) | 1,936 (17.6) |
| 9-year predicted mortality risk* | |||
| Low risk | 1,531 (65.8) | 1,352 (65.0) | 2,047 (67.8) |
| High risk | 853 (34.2) | 787 (35.0) | 1,068 (32.2) |
*Only for men age 65 years and older.
Final survey response rates for adults in 2005, 2010, and 2013 were 69%, 61%, and 61%. The final adult survey response incorporated nonresponse at the household and family levels in addition to nonresponse at the individual adult level. The NHIS did not provide demographic data for nonrespondents compared with adults who completed the survey.21 After incorporating survey weights, these results represented approximately 60,813,000 men age 40 years and older in 2005, 66,065,000 in 2010, and 69,660,000 in 2013. In 2005, 2010, and 2013, the rates of all men age 40 years and older who answered prostate cancer screening questions were 88.7%, 90.0%, and 94.7%, respectively. Appendix Table A1 (online only) compares respondents and nonrespondents. The majority of men age 65 years and older (97.5%) had insurance coverage from Medicare and/or private insurance. The proportion of men who visited a physician in the year before the 2013 survey was lowest among those age 40 to 49 years (64.4%) and highest among men age 75 years and older (92.0%).
There were no significant differences in screening rates by age group between 2005 and 2010. Rates significantly declined from 2010 to 2013 among all age groups except those age 40 to 49 years (Fig 1). Screening did not significantly differ from 2010 to 2013 among 40- to 49-year-old men (odds ratio [OR], 0.88; P = .4), with rates of 12.5% (95% CI, 10.2% to 14.7%) and 11.2% (95% CI, 9.2% to 13.2%) in 2010 and 2013, respectively. Notably, men age 60 to 74 years were most heavily screened, with a rate of 51.2% (95% CI, 48.1% to 54.2%) in 2010, declining to 43.6% (95% CI, 41.0% to 46.2%) in 2013 (OR, 0.74; P < .01). Screening significantly declined for men age 50 to 59 years, from 33.2% (95% CI, 30.1% to 36.3%) in 2010 to 24.8% (95% CI, 22.3% to 27.3%) in 2013 (OR, 0.66; P < .01), as well as for men age 75 years and older, from 43.9% (95% CI, 39.1% to 48.7%) in 2010 to 37.1% (95% CI, 33.2% to 41.0%) in 2013 (OR, 0.75; P = .03). After incorporation of sample weights, approximately 1.4 million men age 40 to 49 years, 3.6 million age 50 to 59 years, 6.6 million age 60 to 74 years, and 2.0 million age 75 years and older were screened for prostate cancer in 2013. We performed two sets of sensitivity analyses confirming the trends observed in 2005, 2010, and 2013 (Appendix Figs A1 and A2). The first assumed all nonrespondents did not undergo PSA screening in the year prior, and the second assumed all nonrespondents did (Appendix Figs A1 and A2).
Fig 1. Proportion of men, by 5-year age group, who saw a physician in the year prior and received a prostate-specific antigen (PSA) test for screening purposes.
Overall, in 2013, men age 65 years and older were screened at a lower rate if they were less healthy and had a higher likelihood (> 52%) of dying in the next 9 years (OR, 0.48; P < .01; Fig 2). This observation persisted in men age 65 to 69 years (OR, 0.38; P < .01), 70 to 74 years (OR, 0.44; P < .01), and 75 years and older (OR, 0.62; P = .01). Large percentages of men were screened in 2013 despite a high risk (> 52%) of predicted 9-year mortality, as follows: 27.2% (95% CI, 16.6% to 37.8%) of men age 65 to 69 years, 32.3% (95% CI, 22.2% to 42.4%) of men age 70 to 74 years, and 32.2% (95% CI, 26.9% to 37.6%) of men age 75 years and older. After incorporation of sample weights, approximately 1.4 million men age 65 years and older with a high risk (> 52%) of predicted 9-year mortality visited a physician and were screened for prostate cancer in 2013.
Fig 2. Prostate cancer screening rates, in 2013, by age group and predicted 9-year mortality among men age 65 years and older who had seen a physician in the year prior.
The majority of all PSA tests performed for men younger than age 80 years in 2013 were drawn from healthier men with a low predicted risk of 9-year mortality (≤ 52%; Fig 3). The majority of tests (73.3%; 95% CI, 63.7% to 82.9%) ordered for men age 80 years and older were in men unlikely to live another 9 years (> 52% mortality likelihood).
Fig 3. The proportion of all men who saw a physician in the year before the 2013 survey and received a prostate-specific antigen screening test who had ≤ 52% or more than 52% likelihood of 9-year mortality at time of testing.
A multivariable model was used to predict screening in men age 65 and older who visited a physician in the year before the 2005, 2010, and 2013 surveys. Model 1, which excluded predicted 9-year mortality, demonstrated that men who were college educated, were married, consumed alcohol, received a colonoscopy in the 10 years before the survey, regularly performed light or moderate activities, did not need assistance with activities of daily living, and rated their health more favorably were significantly more likely to be screened (all P < .05; Table 2). Model 2, which excluded component variables of the 9-year mortality index, demonstrated that men who were white, were more highly educated, were married, had a history of consuming alcohol, received a colonoscopy in the 10 years prior, regularly performed light or moderate activities, and had a lower predicted risk of 9-year mortality were significantly more likely to be screened for prostate cancer (all P < .05; Table 2).
|
| Variable | Model 1,* Without Mortality Risk | Model 2,† With Mortality Risk | ||||
|---|---|---|---|---|---|---|
| OR | 95% CI | P | OR | 95% CI | P | |
| Race | ||||||
| Non-Hispanic white | REF | REF | .07 | REF | REF | .01 |
| Non-Hispanic black | 0.82 | 0.61 to 1.11 | 0.75 | 0.56 to 1.01 | ||
| Hispanic | 0.88 | 0.60 to 1.29 | 0.86 | 0.59 to 1.25 | ||
| Asian | 0.75 | 0.42 to 1.32 | 0.71 | 0.41 to 1.22 | ||
| Other | 0.28 | 0.10 to 0.79 | 0.25 | 0.09 to 0.70 | ||
| Highest education level | ||||||
| Some high school | REF | REF | .01 | REF | REF | < .01 |
| High school diploma | 1.33 | 1.04 to 1.70 | 1.35 | 1.05 to 1.72 | ||
| Some college | 1.29 | 0.98 to 1.70 | 1.28 | 0.97 to 1.69 | ||
| College degree | 1.62 | 1.21 to 2.15 | 1.66 | 1.25 to 2.21 | ||
| Marital status | ||||||
| Not married | REF | REF | < .01 | REF | REF | < .01 |
| Married/living with partner | 1.55 | 1.30 to 1.85 | 1.55 | 1.31 to 1.85 | ||
| Alcohol use | ||||||
| Never drinker | REF | REF | < .01 | REF | REF | < .01 |
| Former drinker | 1.45 | 1.08 to 1.94 | 1.49 | 1.12 to 1.97 | ||
| Current drinker | 1.60 | 1.22 to 2.09 | 1.68 | 1.30 to 2.17 | ||
| Tobacco use | ||||||
| Never smoker | REF | REF | .5 | |||
| Former smoker | 1.08 | 0.90 to 1.31 | ||||
| Current smoker | 0.94 | 0.69 to 1.27 | ||||
| Colonoscopy history | ||||||
| Never had colonoscopy | REF | REF | < .01 | REF | REF | < .01 |
| Colonoscopy longer than 10 years ago | 1.02 | 0.70 to 1.48 | 1.10 | 0.76 to 1.59 | ||
| Colonoscopy within last 10 years | 1.78 | 1.47 to 2.15 | 1.80 | 1.49 to 2.17 | ||
| Activity level | ||||||
| Does not perform light/moderate activity | REF | REF | .02 | REF | REF | .01 |
| Performs light/moderate activities | 1.23 | 1.03 to 1.47 | 1.25 | 1.05 to 1.50 | ||
| Nationality | ||||||
| Born in the United States | REF | REF | .5 | REF | REF | .5 |
| Not born in the United States | 0.89 | 0.63 to 1.26 | 0.88 | 0.63 to 1.25 | ||
| BMI | ||||||
| Underweight/normal | REF | REF | .1 | |||
| Overweight | 1.20 | 0.98 to 1.46 | ||||
| Obese | 1.21 | 0.97 to 1.50 | ||||
| Self-reported personal health status | ||||||
| Excellent/very good | REF | REF | .02 | |||
| Good | 0.84 | 0.68 to 1.02 | ||||
| Fair/poor | 0.70 | 0.54 to 0.90 | ||||
| ADLs | ||||||
| Does not need ADL assistance | REF | REF | .03 | |||
| Needs ADL assistance | 0.82 | 0.68 to 0.98 | ||||
| Walking | ||||||
| Difficulty walking/cannot walk | REF | REF | .2 | |||
| No difficulty walking | 0.95 | 0.89 to 1.02 | ||||
| 9-year mortality risk | ||||||
| Low risk (< 52%) | REF | REF | < .01 | |||
| High risk (≥ 52%) | 0.60 | 0.49 to 0.72 | ||||
Abbreviations: ADL, activities of daily living; BMI, body mass index; OR, odds ratio; REF, reference.
*Model 1 omits the predicted 9-year mortality variable to reduce colinearity.
†Model 2 omits the component variables used to calculate the 9-year mortality variable.
From 2010 to 2013, there were no significant predictors of declines in prostate cancer screening among men age 50 years and older who visited a physician in the year prior. There was a trend toward significance for education level and nation of birth (United States v other) as predictors of declining screening rates from 2010 to 2013 (Table 3).
|
| Variable | Odds Ratio* | 95% CI† | P |
|---|---|---|---|
| Race | |||
| Non-Hispanic white | 0.72 | 0.63 to 0.81 | .3 |
| Non-Hispanic black | 0.89 | 0.66 to 1.20 | |
| Hispanic | 0.84 | 0.57 to 1.24 | |
| Asian | 0.54 | 0.30 to 1.0 | |
| Other | 0.44 | 0.17 to 1.14 | |
| Highest education level | |||
| Some high school | 0.87 | 0.64 to 1.18 | .1 |
| High school diploma | 0.78 | 0.62 to 0.98 | |
| Some college | 0.75 | 0.60 to 0.93 | |
| College degree | 0.58 | 0.48 to 0.70 | |
| Marital status | |||
| Not married | 0.74 | 0.65 to 0.85 | .4 |
| Married/Living with partner | 0.67 | 0.56 to 0.80 | |
| Alcohol use | |||
| Never drinker | 0.69 | 0.51 to 0.95 | .4 |
| Former drinker | 0.64 | 0.51 to 0.80 | |
| Current drinker | 0.75 | 0.66 to 0.86 | |
| Tobacco use | |||
| Never smoker | 0.77 | 0.66 to 0.89 | .4 |
| Former smoker | 0.72 | 0.61 to 0.86 | |
| Current smoker | 0.60 | 0.44 to 0.81 | |
| Colonoscopy history | |||
| Never had colonoscopy | 0.80 | 0.65 to 0.98 | .4 |
| Colonoscopy longer than 10 years ago | 0.55 | 0.31 to 0.97 | |
| Colonoscopy within last 10 years | 0.71 | 0.58 to 0.87 | |
| Activity level | |||
| Does not perform light/moderate activity | 0.74 | 0.62 to 0.87 | .6 |
| Performs light/moderate activities | 0.69 | 0.60 to 0.80 | |
| Nationality | |||
| Born in the United States | 0.70 | 0.63 to 0.79 | .1 |
| Not born in the United States | 0.91 | 0.66 to 1.25 | |
| COPD | |||
| No history of COPD | 0.72 | 0.65 to 0.81 | .6 |
| History of COPD | 0.66 | 0.45 to 0.97 | |
| Diabetes | |||
| No history of diabetes | 0.70 | 0.62 to 0.80 | .4 |
| History of diabetes | 0.79 | 0.63 to 1.0 | |
| Personal history of cancer | |||
| No personal history of cancer | 0.71 | 0.64 to 0.80 | .6 |
| Personal history of cancer | 0.81 | 0.55 to 1.19 | |
| Hospitalizations in past year | |||
| No hospitalization | 0.74 | 0.66 to 0.84 | .4 |
| One night in hospital in past year | 0.58 | 0.41 to 0.82 | |
| Multiple nights in hospital in past year | 0.64 | 0.37 to 1.13 | |
| BMI | |||
| Underweight/normal | 0.72 | 0.57 to 0.91 | .6 |
| Overweight | 0.77 | 0.65 to 0.90 | |
| Obese | 0.67 | 0.55 to 0.81 | |
| Self-reported personal health status | |||
| Excellent/very good | 0.68 | 0.58 to 0.79 | .4 |
| Good | 0.79 | 0.65 to 0.96 | |
| Fair/poor | 0.75 | 0.60 to 0.95 | |
| ADLs | |||
| Does not need ADL assistance | 0.72 | 0.65 to 0.81 | .7 |
| Needs ADL assistance | 0.66 | 0.41 to 1.07 | |
| Walking | |||
| Difficulty walking/cannot walk | 0.70 | 0.61 to 0.80 | .5 |
| No difficulty walking | 0.77 | 0.63 to 0.95 | |
| Anxiety | |||
| No history of anxiety or depression | 0.73 | 0.65 to 0.81 | .5 |
| Anxiety and/or depression | 0.55 | 0.26 to 1.17 | |
| 9-year mortality risk‡ | |||
| Low risk (< 52%) | 0.85 | 0.70 to 1.04 | .3 |
| High risk (≥ 52%) | 0.70 | 0.52 to 0.93 |
Abbreviations: ADL, activities of daily living; BMI, body mass index; COPD, chronic obstructive pulmonary disease; OR, odds ratio.
*OR < 1 indicates a decline and OR > 1 indicates an increase in prostate-specific antigen screening rate.
†P value for interaction with year.
‡Only men age 65 years and older were assessed for 9-year mortality risk because the index is only validated for men in this age group.
Prostate cancer screening significantly declined among men age 50 years and older in the United States after the 2012 USPSTF recommendation discouraging PSA-based screening. No specific patient-related factors predicted declining rates of screening from 2010 to 2013. Despite these declines, in 2013 approximately one third of men age 65 years and older with a high risk (52%) of predicted 9-year mortality were screened—approximately 1.4 million men.
Our study is the first to demonstrate national declines in self-reported PSA testing after the 2012 USPSTF recommendation discouraging PSA-based prostate cancer screening. The largest change occurred among men age 50 to 59 years, in whom absolute and relative screening rates decreased by 8% and 25% from 2010 to 2013. Screening decreased among all age groups, similar to other studies examining regional health networks.22,23 This decrease, regardless of age and even in those with lengthy estimated remaining life expectancy, suggests some physicians are adhering to the USPSTF guidelines compared with age-based and life expectancy–based guidelines.3,4,24 Notably, a similar decline was not observed after the 2008 recommendation discouraging prostate cancer screening among men age 75 and older.13,25 Persistently elevated screening rates among men with limited remaining life expectancies are troubling and merit further interventions. These may include increasing awareness of initiatives such as Choosing Wisely recommendations from the American Society of Clinical Oncology and recommendations from the American Geriatrics Society, both of which discourage screening men with limited remaining life expectancies, as well as supporting physician-led quality initiatives such as the Michigan Urological Surgery Improvement Collaborative, which has demonstrated success in other areas of urologic oncology.26–30 Reducing or eliminating reimbursement for screening and subsequent interventions performed for men with limited remaining life expectancies may be considered if physician-led efforts fall short.
There are many potential effects of decreased prostate cancer screening after the 2012 USPSTF recommendation. Decreased screening will result in fewer prostate biopsies and fewer men diagnosed with prostate cancer.31–34 Models constructed by Gulati et al35 predicted that a continuation of screening patterns from the year 2000 would result in approximately 710,000 to 1,120,000 overdiagnosed prostate cancers and 280,000 prostate cancer deaths from 2013 to 2025. The predicted number of overdiagnosed cancers would decrease by two thirds in a program limiting testing to men younger than age of 70 years, but prostate cancer deaths would increase by 5% to 8%. Predicted prostate cancer deaths would increase by 13% to 20% in a program completely discontinuing screening. These models did not account for contemporary guideline nonadherence in the form of persistent screening of men of advanced age and/or limited remaining life expectancy, an important distinction because approximately 80% of men age 75 and older diagnosed with prostate cancer pursue active treatment.12,13,35,36
Patients at low risk for prostate cancer may be monitored with extended screening intervals, an approach reinforced by data from the European Randomized Study of Screening for Prostate Cancer, the newest American Urological Association recommendation, and other previously published studies.2,24,37,38 Modeling studies from the Rotterdam component of the European Randomized Study of Screening for Prostate Cancer demonstrated that annual screening of men age 55 to 67 years results in an overdiagnosis rate of 50%.39 Heijnsdijk et al40 demonstrated that greatest cost-effectiveness was achieved with a screening program offered to patients between age 55 and 60 years with 1- to 2-year intervals. Gulati et al37 showed that men with PSA values less than the median age group–matched levels may be screened up to every 5 years without dramatically impacting patient-centered outcomes. The approach of Gulati et al37 would reduce overdiagnoses by approximately 27% and false-positive results by 50%, yet would still save 83% of lives compared with an annual screening interval approach. A prospective study by Aus et al38 similarly demonstrated that men age 50 to 66 years with low baseline serum PSA levels could be safely screened every 3 years with minimal effects on prostate cancer–specific mortality.
The contemporary objective of screening is the detection of high-volume or high-grade cancers that are likely to become clinically evident. A physician caring for a patient with an elevated PSA level should interpret the value relative to age-specific controls and not reflexively proceed to biopsy, because up to 55% of elevated PSA levels will normalize within the course of 1 year. The NHIS does not provide data regarding the percentage of men with elevated PSA levels who proceeded to have a prostate biopsy. The probability of high-risk cancer should be estimated using existing validated risk calculators.41–45 Multiple novel biomarkers have also recently become commercially available to improve on PSA alone in predicting the likelihood of a man harboring a high-grade cancer. Tests such as the prostate CA 3, kallikrein panel, and Prostate Health Index have superior operating characteristics compared with PSA alone.46–48 The adoption and integration of these tests, along with downstream effects, require further study.
Limitations of the NHIS database include its interview-based reporting component. A meta-analysis investigating agreement between self-reported cancer screening and medical records, however, indicated a tendency toward under-reporting of prostate cancer screening.49 Similarly, informative details about ordering provider (primary care physician or specialist) or practice structure (private, academic, and so on) were not available. Data regarding survey nonrespondents were not provided by the NHIS, which limited the investigation of potential selection biases. The overall survey response rate for prostate cancer screening questions, however, was high (91.2%), and sensitivity analyses (Appendix Figs A1 and A2, Appendix Table A1) demonstrated preserved prostate cancer screening trends after the 2012 USPSTF guideline. We limited this analysis to men who visited a physician in the year before the survey, which only captures screening behavior among individuals who interfaced with the health care system. This contrasts with an analysis performed in 2011 in which we analyzed screening among all men regardless of their interaction with the health care system.11 The USPSTF recommendation was finalized in May 2012, which allowed for a relatively short amount of time for dissemination of the guidelines into practice. Nevertheless, the guideline was covered widely in every major media outlet, was likely known to an overwhelming majority of primary care physicians and urologists, and has clearly impacted clinical practice.
In conclusion, this is the first analysis demonstrating a nationwide decline in national prostate cancer screening rates among men age 50 years and older after the 2012 USPSTF guideline discouraging PSA testing for the early detection of prostate cancer. Encouragingly, an increased probability of 9-year mortality was associated with decreased screening rates from 2010 to 2013. Despite these trends, roughly a third of men older than age 65 years with a high probability (> 52%) of death within the next 9 years were screened for prostate cancer, exposing these approximately 1.4 million men to a high risk of overdiagnosis and overtreatment.
Processed as a Rapid Communication manuscript.
Presented at the 51st Annual Meeting of the American Society of Clinical Oncology, Chicago, IL, May 29-June 2, 2015.
Terms in blue are defined in the glossary, found at the end of this article and online at www.jco.org.
Authors' disclosures of potential conflicts of interest are found in the article online at www.jco.org. Author contributions are found at the end of this article.
Disclosures provided by the authors are available with this article at www.jco.org.
Conception and design: All authors
Collection and assembly of data: Michael W. Drazer, Dezheng Huo
Data analysis and interpretation: All authors
Manuscript writing: All authors
Final approval of manuscript: All authors
| 1. | GL Andriole, ED Crawford, RL Grubb 3rd , etal: Mortality results from a randomized prostate-cancer screening trial N Engl J Med 360: 1310– 1319,2009 Crossref, Medline, Google Scholar |
| 2. | FH Schröder, J Hugosson, MJ Roobol , etal: Screening and prostate-cancer mortality in a randomized European study N Engl J Med 360: 1320– 1328,2009 Crossref, Medline, Google Scholar |
| 3. | VA Moyer : Screening for prostate cancer: U.S. Preventive Services Task Force recommendation statement Ann Intern Med 157: 120– 134,2012 US Preventive Services Task Force Crossref, Medline, Google Scholar |
| 4. | AM Wolf, RC Wender, RB Etzioni , etal: American Cancer Society guideline for the early detection of prostate cancer: Update 2010 CA Cancer J Clin 60: 70– 98,2010 Crossref, Medline, Google Scholar |
| 5. | KL Greene, PC Albertsen, RJ Babaian , etal: Prostate specific antigen best practice statement: 2009 update J Urol 182: 2232– 2241,2009 Crossref, Medline, Google Scholar |
| 6. | A Qaseem, MJ Barry, TD Denberg , etal: Screening for prostate cancer: A guidance statement from the Clinical Guidelines Committee of the American College of Physicians Ann Intern Med 158: 761– 769,2013 Crossref, Medline, Google Scholar |
| 7. | Clinical Preventive Service Recommendation: Prostate Cancer American Academy of Family Physicians http://www.aafp.org/patient-care/clinical-recommendations/all/prostate-cancer.html Google Scholar |
| 8. | LS Lim, K Sherin : Screening for prostate cancer in U.S. men ACPM position statement on preventive practice Am J Prev Med 34: 164– 170,2008 ACPM Prevention Practice Committee Crossref, Medline, Google Scholar |
| 9. | MH Kawachi, RR Bahnson, M Barry , etal: NCCN clinical practice guidelines in oncology: Prostate cancer early detection J Natl Compr Canc Netw 8: 240– 262,2010 Crossref, Medline, Google Scholar |
| 10. | GE Tasian, MR Cooperberg, JE Cowan , etal: Prostate specific antigen screening for prostate cancer: Knowledge of, attitudes towards, and utilization among primary care physicians Urol Oncol 30: 155– 160,2012 Crossref, Medline, Google Scholar |
| 11. | MW Drazer, D Huo, MA Schonberg , etal: Population-based patterns and predictors of prostate-specific antigen screening among older men in the United States J Clin Oncol 29: 1736– 1743,2011 Link, Google Scholar |
| 12. | MW Drazer, SM Prasad, D Huo , etal: National trends in prostate cancer screening among older American men with limited 9-year life expectancies: Evidence of an increased need for shared decision making Cancer 120: 1491– 1498,2014 Crossref, Medline, Google Scholar |
| 13. | SM Prasad, MW Drazer, D Huo , etal: 2008 US Preventive Services Task Force recommendations and prostate cancer screening rates JAMA 307: 1692– 1694,2012 Crossref, Medline, Google Scholar |
| 14. | G Harris : U.S. panel says no to prostate screening for healthy men The New York Times 2011 http://www.nytimes.com/2011/10/07/health/07prostate.html?_r=0 Google Scholar |
| 15. | New England Journal of Medicine weighs in with views on PSA testing KHN Morning Briefing 2011 Kaiser Health News, Kaiser Family Foundation http://kaiserhealthnews.org/morning-breakout/thur-opinion-nejm-on-psa-testing/ Google Scholar |
| 16. | C Bankhead : Prostate test recommendation draws mixed reviews MedPage Today 2011 http://www.medpagetoday.com/HematologyOncology_/ProstateCancer/29015 Google Scholar |
| 17. | A Selyukh : Health panel takes heat on cancer screening advice Health, Reuters 2011 http://www.reuters.com/article/2011/12/18/us-health-taskforce-idUSTRE7BH0C620111218 Google Scholar |
| 18. | J Kane : PSA testing controversy reignites “over-screening” debate PBS Newshour 2011 http://www.pbs.org/newshour/rundown/psa-testing-controversy-reignites-over-screening-debate/ Google Scholar |
| 19. | V Prabhu, T Lee, S Loeb , etal: Twitter response to the United States Preventive Services Task Force recommendations against screening with prostate specific antigen BJU Int [epub ahead of print on March 25, 2014] Google Scholar |
| 20. | MA Schonberg, RB Davis, EP McCarthy , etal: External validation of an index to predict up to 9-year mortality of community-dwelling adults aged 65 and older J Am Geriatr Soc 59: 1444– 1451,2011 Crossref, Medline, Google Scholar |
| 21. | National Health Interview Survey. Public-use data file and documentation National Center for Health Statistics http://www.cdc.gov/nchs/nhis/quest_data_related_1997_forward.htm Google Scholar |
| 22. | JA Cohn, CE Wang, JC Lakeman , etal: Primary care physician PSA screening practices before and after the final U.S. Preventive Services Task Force recommendation Urol Oncol 32: 41.e23– 41.e30,2014 Crossref, Medline, Google Scholar |
| 23. | A Aslani, BJ Minnillo, B Johnson , etal: The impact of recent screening recommendations on prostate cancer screening in a large health care system J Urol 191: 1737– 1742,2014 Crossref, Medline, Google Scholar |
| 24. | HB Carter, PC Albertsen, MJ Barry , etal: Early detection of prostate cancer: AUA Guideline J Urol 190: 419– 426,2013 Crossref, Medline, Google Scholar |
| 25. | Screening for prostate cancer: U.S. Preventive Services Task Force recommendation statement Ann Intern Med 149: 185– 191,2008 US Preventive Services Task Force Crossref, Medline, Google Scholar |
| 26. | Choosing Wisely, American Society of Clinical Oncology: 10 things physicians and patients should question American Society of Clinical Oncology http://www.choosingwisely.org Google Scholar |
| 27. | Choosing Wisely, American Geriatrics Society: 10 things physicians and patients should question American Geriatrics Society http://www.choosingwisely.org Google Scholar |
| 28. | CB Riedinger, PR Womble, SM Linsell , etal: Variation in prostate cancer detection rates in a statewide quality improvement collaborative J Urol 192: 373– 378,2014 Crossref, Medline, Google Scholar |
| 29. | PR Womble, JE Montie, Z Ye , etal: Contemporary use of initial active surveillance among men in Michigan with low-risk prostate cancer Eur Urol 67: 44– 50,2015 Crossref, Medline, Google Scholar |
| 30. | CP Filson, B Boer, J Curry , etal: Improvement in clinical TNM staging documentation within a prostate cancer quality improvement collaborative Urology 83: 781– 786,2014 Crossref, Medline, Google Scholar |
| 31. | J Wade, DJ Rosario, RC Macefield , etal: Psychological impact of prostate biopsy: Physical symptoms, anxiety, and depression J Clin Oncol 31: 4235– 4241,2013 Link, Google Scholar |
| 32. | S Loeb, HB Carter, SI Berndt , etal: Complications after prostate biopsy: Data from SEER-Medicare J Urol 186: 1830– 1834,2011 Crossref, Medline, Google Scholar |
| 33. | RK Nam, R Saskin, Y Lee , etal: Increasing hospital admission rates for urological complications after transrectal ultrasound guided prostate biopsy J Urol 183: 963– 968,2010 Crossref, Medline, Google Scholar |
| 34. | S Loeb, A Vellekoop, HU Ahmed , etal: Systematic review of complications of prostate biopsy Eur Urol 64: 876– 892,2013 Crossref, Medline, Google Scholar |
| 35. | R Gulati, A Tsodikov, R Etzioni , etal: Expected population impacts of discontinued prostate-specific antigen screening Cancer 120: 3519– 3526,2014 Crossref, Medline, Google Scholar |
| 36. | MR Cooperberg, JM Broering, PR Carroll : Time trends and local variation in primary treatment of localized prostate cancer J Clin Oncol 28: 1117– 1123,2010 Link, Google Scholar |
| 37. | R Gulati, JL Gore, R Etzioni : Comparative effectiveness of alternative prostate-specific antigen–based prostate cancer screening strategies: Model estimates of potential benefits and harms Ann Intern Med 158: 145– 153,2013 Crossref, Medline, Google Scholar |
| 38. | G Aus, JE Damber, A Khatami , etal: Individualized screening interval for prostate cancer based on prostate-specific antigen level: Results of a prospective, randomized, population-based study Arch Intern Med 165: 1857– 1861,2005 Crossref, Medline, Google Scholar |
| 39. | G Draisma, R Boer, SJ Otto , etal: Lead times and overdetection due to prostate-specific antigen screening: Estimates from the European Randomized Study of Screening for Prostate Cancer J Natl Cancer Inst 95: 868– 878,2003 Crossref, Medline, Google Scholar |
| 40. | EA Heijnsdijk, TM de Carvalho, A Auvinen , etal: Cost-effectiveness of prostate cancer screening: A simulation study based on ERSPC data J Natl Cancer Inst 107: 366,2015 Crossref, Medline, Google Scholar |
| 41. | JE Oesterling, SJ Jacobsen, CG Chute , etal: Serum prostate-specific antigen in a community-based population of healthy men: Establishment of age-specific reference ranges JAMA 270: 860– 864,1993 Crossref, Medline, Google Scholar |
| 42. | JA Eastham, E Riedel, PT Scardino , etal: Variation of serum prostate-specific antigen levels: An evaluation of year-to-year fluctuations JAMA 289: 2695– 2700,2003 Crossref, Medline, Google Scholar |
| 43. | Individualized risk assessment of prostate cancer UT Health Science Center http://deb.uthscsa.edu/URORiskCalc/Pages/uroriskcalc.jsp Google Scholar |
| 44. | IM Thompson, DP Ankerst, C Chi , etal: Assessing prostate cancer risk: Results from the Prostate Cancer Prevention Trial J Natl Cancer Inst 98: 529– 534,2006 Crossref, Medline, Google Scholar |
| 45. | DJ Parekh, DP Ankerst, BA Higgins , etal: External validation of the Prostate Cancer Prevention Trial risk calculator in a screened population Urology 68: 1152– 1155,2006 Crossref, Medline, Google Scholar |
| 46. | MM Vedder, EW de Bekker-Grob, HG Lilja , etal: The added value of percentage of free to total prostate-specific antigen, PCA3, and a kallikrein panel to the ERSPC risk calculator for prostate cancer in prescreened men Eur Urol 66: 1109– 1115,2014 Crossref, Medline, Google Scholar |
| 47. | T Nordström, A Vickers, M Assel , etal: Comparison between the four-kallikrein panel and Prostate Health Index for predicting prostate cancer Eur Urol [epub ahead of print on August 20, 2014] Google Scholar |
| 48. | JT Wei, Z Feng, AW Partin , etal: Can urinary PCA3 supplement PSA in the early detection of prostate cancer? J Clin Oncol 32: 4066– 4072,2014 Link, Google Scholar |
| 49. | GH Rauscher, TP Johnson, YI Cho , etal: Accuracy of self-reported cancer-screening histories: A meta-analysis Cancer Epidemiol Biomarkers Prev 17: 748– 757,2008 Crossref, Medline, Google Scholar |
| biomarker: | a functional biochemical or molecular indicator of a biologic or disease process that has predictive, diagnostic, and/or prognostic utility. |
| life expectancy: | the average number of years that an individual would live if he or she were to experience throughout life the age-specific mortality rates prevailing in a given year. |
| prostate-specific antigen (PSA): | a protein produced by cells of the prostate gland. The blood level of prostate-specific antigen (PSA) is used as a tumor marker for men who may be suspected of having prostate cancer. Most physicians consider 0 to 4.0 ng/mL to be the normal range. Levels of 4 to 10 and 10 to 20 ng/mL are considered slightly and moderately elevated, respectively. PSA levels have to be complemented with other tests to make a firm diagnosis of prostate cancer. |
The following represents disclosure information provided by authors of this manuscript. All relationships are considered compensated. Relationships are self-held unless noted. I = Immediate Family Member, Inst = My Institution. Relationships may not relate to the subject matter of this manuscript. For more information about ASCO's conflict of interest policy, please refer to www.asco.org/rwc or jco.ascopubs.org/site/ifc.
No relationship to disclose
No relationship to disclose
Consulting or Advisory Role: Myriad Genetics, Medivation, Janssen Pharmaceuticals, Genomic Health, OPKO Diagnostics, MDxHealth
Speakers' Bureau: Myriad Genetics, Janssen Pharmaceuticals
Research Funding: Myriad Genetics (Inst)
|
| Characteristic | No. (%) | P | |
|---|---|---|---|
| Respondents (n = 22,490) | Nonrespondents (n = 2,168) | ||
| Race/ethnicity | < .01 | ||
| Non-Hispanic white | 14,917 (74.0) | 1,420 (74.3) | |
| Non-Hispanic black | 2,986 (9.4) | 350 (10.5) | |
| Hispanic | 3,196 (11.2) | 257 (9.2) | |
| Asian | 1,017 (4.1) | 116 (4.6) | |
| Other | 374 (1.3) | 25 (1.4) | |
| Highest education level | .1 | ||
| Some high school | 3,994 (15.0) | 424 (16.9) | |
| High school diploma | 6,126 (27.4) | 634 (30.3) | |
| Some college | 5,798 (26.1) | 510 (25.4) | |
| College degree | 6,454 (27.5) | 539 (27.5) | |
| Marital status | .2 | ||
| Not married | 8,750 (25.6) | 991 (31.3) | |
| Married/living with partner | 13,690 (74.4) | 1,138 (68.7) | |
| Colonoscopy history | .03 | ||
| Never had colonoscopy | 13,216 (73.2) | 399 (65.9) | |
| Colonoscopy longer than 10 years ago | 580 (3.2) | 19 (2.1) | |
| Colonoscopy within last 10 years | 4,215 (23.6) | 197 (32.0) | |
| Body mass index | .6 | ||
| Underweight/normal | 5,799 (24.5) | 554 (28.8) | |
| Overweight | 9,936 (44.9) | 803 (42.5) | |
| Obese | 6,558 (30.6) | 496 (28.8) | |
| Tobacco use | .5 | ||
| Never smoker | 10,304 (46.8) | 898 (46.1) | |
| Former smoker | 7,420 (33.1) | 686 (34.2) | |
| Current smoker | 4,734 (20.1) | 428 (19.7) | |
| Personal cancer history | .6 | ||
| No personal history of cancer | 21,401 (95.2) | 2,017 (93.7) | |
| Personal history of cancer | 1,077 (4.8) | 131 (6.3) | |
| Self-reported personal health status | < .01 | ||
| Excellent/very good | 11,742 (55.0) | 1,002 (49.9) | |
| Good | 6,651 (28.9) | 649 (28.1) | |
| Fair/poor | 4,087 (16.1) | 514 (22.0) | |
| Age, years | < .01 | ||
| 40-49 | 7,048 (34.0) | 602 (30.0) | |
| 50-59 | 6,616 (30.4) | 619 (30.1) | |
| 60-74 | 6,405 (26.6) | 604 (26.5) | |
| 75+ | 2,421 (9.1) | 343 (13.4) | |
| Survey year | < .01 | ||
| 2005 | 7,278 (30.4) | 926 (38.1) | |
| 2010 | 6,301 (32.8) | 742 (41.3) | |
| 2013 | 8,911 (36.8) | 500 (20.6) | |
| Visited physician in year prior | < .01 | ||
| Visited a physician | 16,520 (75.0) | 1,401 (77.5) | |
| Did not visit a physician | 5,954 (25.1) | 412 (22.5) | |
| 9-year predicted mortality risk* | < .01† | ||
| Low risk | 4,185 (69.4) | 401 (58.4) | |
| High risk | 1,965 (30.6) | 311 (41.6) | |
*Only for men age 65 and older.
†Univariable analysis only.
Fig A1. Screening rates by year. Sensitivity analysis assuming all nonrespondents did not have a prostate cancer screen in the year before the survey.
Fig A2. Screening rates by year. Sensitivity analysis assuming all nonrespondents had a prostate cancer screen in the year before the survey.
