Abstract
This study of National Health Interview Survey data evaluates recent patterns of prostate-specific antigen testing for prostate cancer screening in the United States
Recommendations for prostate-specific antigen (PSA)-based prostate cancer screening have changed considerably in recent years. In 2008, the US Preventive Services Task Force (USPSTF) recommended against PSA-based prostate cancer screening among men 75 years or older, and in 2012, they recommended against PSA testing for men of all ages. Other organizations emphasize shared-decision making for men 50 years or older with a long life expectancy. As a result of shifting recommendations, PSA screening rates declined from 37.8% in 2010 to 30.8% in 2013 among men 50 years or older, resulting in substantial declines in prostate cancer incidence. However, it is not known if PSA testing has continued to decline. The objective of the present study was to examine PSA testing patterns using recently released 2015 National Health Interview Survey (NHIS) data.
Methods
Data on 19 690 male respondents 50 years or older were selected from the 2010, 2013, and 2015 NHIS, an in-person household survey of noninstitutionalized people with respective response rates of 60.8%, 61.2%, and 55.2%. The primary outcome was self-reported PSA testing in the past year for routine reasons. Men who reported a history of or were missing data on prostate cancer diagnosis (n = 1115) or PSA testing (n = 1574) or who had PSA testing for nonscreening reasons (n = 805) were excluded, resulting in an analytic population of 16 196 men. The χ2 test (α = .05) was used to examine differences in PSA screening rates (SRs). Adjusted screening rate ratios (SRRs) and corresponding 99% confidence intervals (CIs) were estimated using logistic regression models with predicted marginal probabilities and controlled for sociodemographic factors, insurance, and health care characteristics. All analyses were conducted using SAS-callable SUDAAN software, version 9.4, and accounted for the complex survey design.
Results
Of the 16 196 men included in our analyses, more than half were aged 50 to 64 years, and three-quarters had visited their primary care physician in the past year, proportions that were similar across survey years. Among men 50 years or older, unadjusted SRs of PSA testing for routine reasons in the past year decreased from 38.3% in 2010 to 31.5% in 2013 (P < .001) and remained stable at 32.1% through 2015 (P = .62) (Table 1). Adjusted analyses were similar where the PSA testing in the past year was significantly lower in 2013 than in 2010 (SRR, 0.83; 99% CI, 0.77-0.89) but not significantly different in 2015 and 2013 (SRR, 0.99; 95% CI, 0.91-1.08). The pattern of declining PSA testing between 2010 and 2013, but not between 2013 and 2015, was similar across age groups (50-64, 65-74, and ≥75 years) (Table 1 and Table 2).
Table 1. Unadjusted PSA Testing in the Past Year for Routine Reasons Among Male NHIS Respondents by Survey Year.
| Characteristic | Respondents, No. (%) [99% CI] (N = 16 196) | P Value | |||
|---|---|---|---|---|---|
| 2010 | 2013 | 2015 | 2010 vs 2013 | 2015 vs 2013 | |
| Overall | 4106 (38.3) [35.9-40.8] | 6065 (31.5) [29.6-33.5] | 6025 (32.1) [29.9-34.4] | <.001 | .62 |
| Age category, y | |||||
| 50-64 | 2535 (32.9) [29.9-36.0] | 3624 (25.2) [23.0-27.5] | 3384 (27.6) [25.0-30.4] | <.001 | .08 |
| 65-74 | 975 (51.1) [45.9-56.2] | 1526 (45.9) [41.4-50.4] | 1661 (42.1) [37.7-46.6] | .046 | .10 |
| >75 | 596 (44.8) [38.7-51.1] | 915 (37.8) [32.4-43.5] | 980 (35.1) [30.0-40.6] | .04 | .36 |
| Race/ethnicity | |||||
| Hispanic | 518 (21.5) [16.6-27.5] | 667 (19.7) [14.6-26.2] | 650 (21.3) [16.2-27.6] | .57 | .58 |
| Non-Hispanic white | 2719 (41.5) [38.7-44.4] | 4210 (34.0) [31.8-36.3] | 4299 (35.4) [32.8-38.1] | <.001 | .29 |
| Non-Hispanic black | 630 (32.0) [26.2-38.3] | 856 (28.9) [23.9-34.4] | 731 (27.4) [22.6-32.8] | .33 | .61 |
| Other | 239 (30.4) [20.3-42.8] | 332 (22.0) [14.8-31.5] | 345 (16.3) [10.4-24.5] | .12 | .13 |
| Educational attainmenta | |||||
| <HS diploma | 784 (24.5) [20.0-29.6] | 1073 (21.6) [17.1-26.8] | 905 (18.1) [13.9-23.3] | .26 | .18 |
| HS or GED | 1121 (31.4) [27.1-36.1] | 1634 (26.2) [22.5-30.2] | 1637 (28.2) [24.3-32.5] | .02 | .36 |
| Some college | 1013 (39.1) [34.5-43.9] | 1630 (32.3) [28.5-36.3] | 1663 (31.8) [27.6-36.3] | .003 | .84 |
| College graduate | 1171 (50.4) [46.0-54.9] | 1707 (39.7) [35.9-43.6] | 1792 (41.4) [37.0-45.9] | <.001 | .42 |
| Insurance category | |||||
| Private | 1623 (37.9) [34.3-41.7] | 2206 (30.1) [27.1-33.3] | 2277 (31.3) [28.0-34.9] | <.001 | .50 |
| Medicaid/state plan | 146 (19.5) [11.2-32.0] | 261 (12.4) [7.0-21.3] | 314 (14.2) [8.5-22.7] | .15 | .64 |
| Medicare | 1455 (45.4) [41.3-49.5] | 2284 (39.1) [35.7-42.6] | 2362 (38.2) [34.6-41.9] | .002 | .60 |
| Military | 190 (48.1) [36.9-59.5] | 223 (33.5) [23.7-44.9] | 214 (33.1) [21.7-46.8] | .02 | .95 |
| Uninsured | 459 (12.8) [8.6-18.7] | 676 (10.1) [6.7-15.0] | 394 (8.8) [4.5-16.4] | .26 | .62 |
| Other | 233 (47.8) [38.7-57.0] | 415 (48.4) [39.4-57.4] | 464 (39.1) [31.3-47.5] | .91 | .047 |
| Has a usual source of primary careb | |||||
| Yes | 3630 (41.9) [39.3-44.5] | 5379 (34.4) [32.4-36.6] | 5423 (34.5) [32.1-37.0] | <.001 | .98 |
| No | 474 (8.3) [5.1-13.2] | 686 (4.8) [2.8-8.1] | 601 (7.4) [4.0-13.3] | .06 | .18 |
| Visited a PCP in the past 12 monthsc | |||||
| Yes | 3101 (45.5) [42.7-48.3] | 4514 (37.8) [35.5-40.1] | 4600 (37.8) [35.3-40.5] | <.001 | .94 |
| No | 1005 (13.6) [10.4-17.7] | 1545 (11.9) [9.0-15.6] | 1421 (12.0) [9.2-15.6] | .37 | .94 |
| US born | |||||
| Yes | 3408 (40.1) [37.5-42.8] | 5145 (32.6) [30.6-34.8] | 5119 (33.8) [31.4-36.2] | <.001 | .34 |
| No | 698 (26.5) [21.3-32.5] | 916 (25.5) [20.9-30.7] | 903 (23.2) [18.4-28.7] | .72 | .39 |
| Geographic region | |||||
| Northeast | 703 (38.4) [32.5-44.7] | 1061 (33.3) [29.1-37.6] | 1054 (32.6) [27.3-38.4] | .07 | .79 |
| Midwest | 941 (40.5) [35.5-45.6] | 1233 (31.5) [27.4-36.0] | 1273 (31.8) [27.4-36.6] | <.001 | .91 |
| South | 1522 (38.7) [34.7-42.7] | 2255 (34.2) [30.8-37.7] | 1997 (36.1) [32.4-39.9] | .02 | .28 |
| West | 940 (35.1) [30.3-40.1] | 1516 (25.6) [22.3-29.2] | 1701 (25.5) [21.3-30.1] | <.001 | .94 |
Abbreviations: GED, graduate equivalent degree; HS, high school; NHIS, National Health Interview Survey; PCP, primary care physician; PSA, prostate-specific antigen.
Data on educational attainment were missing for 66 respondents.
Data on having a usual source of care were missing for 2 respondents.
Data on visiting a physician in the past year were missing for 10 respondents.
Table 2. Adjusted PSA Testing in the Past Year for Routine Reasons Among Male NHIS Respondents by Survey Year.
| Characteristic | SRR (99% CI) |
|---|---|
| Model 1a | |
| Year | |
| 2013 vs 2010 | 0.83 (0.77-0.89) |
| 2015 vs 2013 | 0.99 (0.91-1.08) |
| Age category, y | |
| 50-64 | 1 [Reference] |
| 65-74 | 1.49 (1.31-1.69) |
| >75 | 1.29 (1.12-1.48) |
| Race/ethnicity | |
| Non-Hispanic white | 1 [Reference] |
| Hispanic | 0.82 (0.68-0.98) |
| Non-Hispanic black | 0.94 (0.85-1.04) |
| Other | 0.62 (0.48-0.78) |
| Educational attainment | |
| <HS diploma | 0.58 (0.50-0.67) |
| HS or GED | 0.70 (0.63-0.77) |
| Some college | 0.82 (0.75-0.89) |
| College graduate | 1 [Reference] |
| Insurance category | |
| Private | 1 [Reference] |
| Medicaid/state plan | 0.67 (0.52-0.88) |
| Medicare | 0.92 (0.80-1.04) |
| Military | 1.09 (0.92-1.29) |
| Uninsured | 0.68 (0.56-0.84) |
| Other | 1.02 (0.87-1.20) |
| Has a usual source of primary care | |
| Yes | 1 [Reference] |
| No | 0.45 (0.33-0.61) |
| Visited a primary care physician in the past year | |
| Yes | 1 [Reference] |
| No | 0.45 (0.39-0.53) |
| Place of birth | |
| United States | 1 [Reference] |
| Outside the United States | 1.02 (0.91-1.14) |
| US Geographic region | |
| Northeast | 1 [Reference] |
| Midwest | 1.02 (0.91-1.14) |
| South | 1.11 (1.00-1.23) |
| West | 0.90 (0.80-1.02) |
| Model 2b: Age and Year Interaction | |
| 2013 vs 2010 among men | |
| 50-64 y | 0.80 (0.71-0.90) |
| 65-74 y | 0.89 (0.71-1.02) |
| >75 y | 0.82 (0.68-0.99) |
| 2015 vs 2013 among men | |
| 50-64 y | 1.06 (0.94-1.19) |
| 65-74 y | 0.92 (0.81-1.04) |
| >75 y | 0.95 (0.79-1.15) |
Abbreviations: GED, graduate equivalent degree; HS, high school; NHIS, National Health Interview Survey; PSA, prostate-specific antigen; SRR, screening rate ratio.
Model 1 adjusts for race/ethnicity, educational attainment, having a usual source of care, visiting a primary care physician in the past year, insurance type, and US geographic region. Only men with complete data on these factors were included in the model. There were 16 112 men included in the adjusted analysis (2010, n = 4087; 2013, n = 6036; and 2015, n = 5989).
Model 2 includes an interaction term for survey year and age and adjusts for race/ethnicity, educational attainment, having a usual source of care, visiting a primary care physician in the past year, insurance type, and US geographic region. Only men with complete data on these factors were included in the model. There were 16 112 men included in the adjusted analysis (2010, n = 4087; 2013, n = 6036; 2015, n = 5989).
Discussion
In this nationally representative study, we found that previously reported declines in PSA testing have not continued in recent years, and approximately a third of men 50 years or older still receive routine PSA testing. Physicians interested in deadopting PSA testing may have done so, closely following the USPSTF recommendation and the media attention that came with it. In addition, other public health organizations still support PSA testing, albeit with shared decision making, and physicians may have chosen to continue to offer PSA testing based on their beliefs about screening and interpretation of clinical trial results. We relied on self-reported PSA testing, which is subject to recall bias, and some men may not have been informed of testing, which is a limitation of the study. However, this study provides data on contemporary nationwide PSA testing patterns.
Conclusions
In conclusion, previous declines in routine PSA testing between 2010 and 2013 did not continue in the most recent time period (between 2013 and 2015). A recent study reported a modest short-term (from 2011-2013) increase in the incidence of metastatic prostate cancer among men 75 years or older. However, continued evaluation on how testing patterns influence prostate cancer outcomes over the long term are needed.
References
- 1.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(2):120-134. [DOI] [PubMed] [Google Scholar]
- 2.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(2):70-98. [DOI] [PubMed] [Google Scholar]
- 3.Jemal A, Fedewa SA, Ma J, et al. . Prostate cancer incidence and PSA testing patterns in relation to USPSTF screening recommendations. JAMA. 2015;314(19):2054-2061. [DOI] [PubMed] [Google Scholar]
- 4.US Department of Health and Human Services National Health Interview Survey: Survey Description. Hyattsville, MD: National Center for Health Statistics; 2016. [Google Scholar]
- 5.Rauscher GH, Johnson TP, Cho YI, Walk JA. Accuracy of self-reported cancer-screening histories: a meta-analysis. Cancer Epidemiol Biomarkers Prev. 2008;17(4):748-757. [DOI] [PubMed] [Google Scholar]
- 6.Hu JC, Nguyen P, Mao J, et al. . Increase in prostate cancer distant metastases at diagnosis in the United States [published online December 29, 2016]. JAMA Oncol. doi: 10.1001/jamaoncol.2016.5465 [DOI] [PMC free article] [PubMed] [Google Scholar]