The influence of physician recommendation on prostate-specific antigen screening

Urologic Oncology: Seminars and Original Investigations 33 (2015) 424.e1–424.e7

Original article

The influence of physician recommendation on prostate-specific antigen screening Daniel Pucheril, M.D., M.B.A.a, Deepansh Dalela, M.D.a, Jesse Sammon, D.O.a, Akshay Sood, M.D.a, Maxine Sun, Ph.D.b, Quoc-Dien Trinh, M.D.b, Mani Menon, M.D.a, Firas Abdollah, M.D., F.E.B.U.a,* b

a VUI Center for Outcomes Research, Analytics and Evaluation, Henry Ford Health System, Detroit, MI Division of Urologic Oncology and Center for Surgery and Public Health, Dana-Farber Cancer Institute, Harvard Medical School, Boston, MA

Received 4 February 2015; received in revised form 14 June 2015; accepted 18 June 2015

Abstract Purpose: Prostate-specific antigen (PSA) screening is controversial, and little is known regarding a physician’s effect on a patient’s decision to undergo screening. This study’s objective was to evaluate the effect of a patient’s understanding of the risks and benefits of screening compared to the final recommendation of the provider on the patient’s decision to undergo PSA screening. Materials and methods: Using the 2012 Behavioral Risk Factor Surveillance System, men older than 55 years who did not have a history of prostate cancer/prostate “problem” and who reported a PSA test within the preceding year were considered to have undergone screening. The percentages of men informed and not informed of the risks and benefits of screening and the percentage men receiving recommendations for PSA screening from their provider were reported. Multivariable complex-sample logistic regression calculated the odds of undergoing screening. Results: In all, 75% of men were informed of screening benefits; however, 32% were informed of screening risks. After being informed of both, 56% of men opted for PSA screening if the provider recommended it, compared with only 21% when not recommended. Men receiving a recommendation to undergo PSA testing had higher odds of undergoing screening (odds ratio [OR] ¼ 4.98, 95% CI: 4.53–5.48) compared with those who were only informed about screening benefits (OR ¼ 2.40, 95% CI: 2.18–2.65) or risks (OR ¼ 0.92, 95% CI: 0.86–0.98). Significant limitations include recall and nonresponse bias. Conclusions: Patients’ decision to undergo or forgo PSA screening is heavily influenced by the recommendation of their physician; it is imperative that physicians are cognizant of their biases and facilitate a shared decision-making process. r 2015 Elsevier Inc. All rights reserved. Keywords: Prostate-specific antigen; Screening; Physician influence

1. Introduction Routine prostate-specific antigen (PSA) screening for prostate cancer continues to be a controversial topic, especially considering the conflicting recommendations given by the various professional and regulatory bodies. Although the most recent United States Preventive Services Task Force (USPSTF) guidelines unequivocally recommend against PSA screening for any age [1], the American Corresponding author. Tel.: þ1-313-916-2062. E-mail address: [email protected] (F. Abdollah). *

http://dx.doi.org/10.1016/j.urolonc.2015.06.013 1078-1439/r 2015 Elsevier Inc. All rights reserved.

Urological Association guidelines endorse shared decisionmaking for men between 55 and 69 years of age [2] and the American Cancer Society recommends shared decisionmaking for average-risk men older than 50 years and for the highest risk men, older than 40 years [3]. As little is known regarding a physician’s effect on a patient’s decision to undergo screening, we analyzed data from the Behavioral Risk Factor Surveillance System (BRFSS) 2012 to understand the relative effect of a patient’s understanding of the risks and benefits of PSA screening compared to the final recommendation of the health care provider (HCP) on the patient’s decision to undergo PSA screening. The BRFSS, a

424.e2

D. Pucheril et al. / Urologic Oncology: Seminars and Original Investigations 33 (2015) 424.e1–424.e7

joint initiative of the Centers for Disease Control and U.S. states/territories, is the world’s largest continuously conducted annual health survey.

2. Methods The BRFSS surveys are conducted by each state via telephone and designed to measure behavioral risk factors for the adult population living in households and are administered to a stratified random sample of the U.S. population aged 18 years and older living in the 50 states, the District of Columbia, Guam, and Puerto Rico [4]. The current BRFSS questionnaire consists of 3 parts: (1) the core component (consisting of fixed core, rotating core, and emerging core), (2) optional modules, and (3) stateadded questions. The fixed-core questions are a standard set of questions that are asked by all states annually, whereas rotating-core questions address different topics and are asked by all states in alternating years. Emerging-core questions focus on “late-breaking” topics and can be added to both fixed and rotating cores. States also have the option to participate in optional module questions or design and ask questions unique to the state [4]. Prostate Cancer Screening questions are included as part of the rotating module and are asked by all states in even numbered years starting in 2002. In 2011 the BRFSS for the first time incorporated cell phone data, thereby changing the weighting methodology and thus BRFSS data collected after 2011 are not comparable to BRFSS data from prior years. The 2012 BRFSS was used in this work to provide the most contemporary analysis of PSA screening patterns. Of note, the 2012 BRFSS included an optional Prostate Cancer Screening Decision Making module (optional module 14) in which only 3 states participated; the results of this module are not included in this work. Rather, this work is based on the analysis of the Prostate Cancer Screening module (core module 16; Appendix 1) [4]. The BRFSS provides nationally representative estimates via iterative proportional fitting weighting. This methodology allows the incorporation of cellular telephone survey data, reduces nonresponse bias, and reduces error within estimates. Patients are weighted by age, gender, race/ ethnicity, educational attainment, marital status, property ownership, and telephone ownership. The median combined response rate in the 2012 BRFSS survey was 45.2%, covering 50 states (including District of Columbia, Guam, and Puerto Rico) [5]. The population estimates provided in Table 1 reflect the results of a weighting calculation applied to the actual number of respondents. Baseline population characteristics were determined for the overall population of men aged Z55 years and also for the cohort of men aged Z55 self-reporting PSA screening. Men aged Z55 years who did not have a prior history of prostate cancer/ prostate “problem” and who reported a PSA test within the preceding 12 months were considered to have undergone

PSA screening (n ¼ 77,627). The percentage of men informed and not informed of the risks and benefits of PSA screening in addition to the percentage men receiving recommendations for PSA screening from their HCP was reported. Finally, multivariable complex-sample logistic regression calculated the odds ratio (OR) of undergoing screening in men informed of PSA screening benefits, men informed of PSA screening risks, and for men receiving a recommendation from their HCP to undergo screening. 3. Results Descriptive statistics and screening incidence according to age, race/ethnicity, education level, income category, presence of health insurance, access to HCP, marital status, and area of residence are reported in Table 1. Patients informed of PSA screening’s benefits and/or risks had a higher incidence of screening. In 75% of men were informed of PSA screening benefits, and of these 53% were likely to undergo screening, compared with 14% of noninformed patients, whereas 32% were informed of PSA screening’s risks, of whom 51% underwent screening vs. 39% of noninformed patients (both P o 0.001, Table 1). Likewise, men who received a PSA screening recommendation from their HCP had a significantly higher PSA screening incidence when compared with men who did not receive such a recommendation (56% vs. 12%; P o 0.001). A significantly higher number of men chose to undergo PSA screening when they were informed about its benefits than when they were not, irrespective of the HCP’s recommendation (Table 2). After being informed both about the benefits and risks of PSA screening, 56% of men opted for a PSA screening if the HCP recommended it, compared with only 21% when the HCP did not recommend it. Of the men who were informed neither about the benefits nor the risks of PSA screening, 47% of the men chose to undergo screening if the HCP recommended it, although only 5% opted for screening if it was not recommended by the HCP. On multivariate analysis, men who received a recommendation to undergo PSA screening had higher odds of undergoing the screening test (OR ¼ 4.98, 95% CI: 4.53–5.48) relative to those informed only of PSA screening’s benefits (OR ¼ 2.40, 95% CI: 2.18–2.65) or risks (OR ¼ 0.92, 95% CI: 0.86–0.98) (Table 3). 4. Discussion In recent years, there has been significant change to the prostate cancer screening guidelines. Specifically, the USPSTF and American Urological Association guidelines of 2011/12 and 2013, respectively, mark a dramatic shift in screening recommendations. It is largely unknown to what degree clinicians have adopted these guidelines; whether physicians clearly enumerate the risks and benefits of PSA screening; and finally, how a physician’s recommendation

D. Pucheril et al. / Urologic Oncology: Seminars and Original Investigations 33 (2015) 424.e1–424.e7

424.e3

Table 1 Population characteristics for overall men population 55 years or older, weighted by age, race/ethnicity, education attainment, marital status, property ownership, and telephone ownership; BRFSS, 2012 Characteristic

Weighted 95% CI population distribution

Population estimate (million)

95% CI

Percent undergoing 95% CI PSA screening

Overall Age group 55–59 60–64 65–69 70–74 75–79 80–99

100%

27.988

27.802–28.174

42.9

42.3–43.6

6.642–6.981 6.584–6.916 4.905–5.175 3.561–3.795 2.582–2.788 2.913–3.129

36.2 44.5 49.4 48.9 45.6 34.2

34.8–37.5 43.2–45.9 47.9–50.8 47.2–50.5 43.7–47.6 32.5–36.0

24.834–25.188 2.828–3.125

43.5 38.5

42.8–44.1 35.9–41.0

o0.001 24.3 24.1 18.0 13.1 9.6 10.8

23.8–24.9 23.6–24.7 17.5–18.5 12.7–13.6 9.2–10.0 10.4–11.2

6.812 6.750 5.040 3.678 2.685 3.021

89.4 10.6

88.9–89.9 10.1–11.1

25.011 2.977

12.1 28.7 28.7 30.3

11.6–12.6 28.2–29.3 28.1–29.3 29.8–30.9

3.393 8.038 8.026 8.493

3.246–3.541 7.869–8.207 7.841–8.210 8.364–8.622

28.5 40.3 44.0 50.2

26.5–30.5 39.1–41.5 42.8–45.3 49.2–51.1

0.1

0.1–0.2

0.036

0.026–0.046

36.5

24.2–50.8

Income category o15,000 15,000–o25,000 25,000–o35,000 35,000–o50,000 Z50,000 Unknown

7.2 14.1 10.8 15.1 43.1 9.7

6.8–7.5 13.7–14.6 10.4–11.2 14.7–15.6 42.5–43.8 9.3–10.1

2.001 3.951 3.014 4.239 12.072 2.709

1.903–2.099 3.822–4.079 2.900–3.129 4.110–4.368 11.883–12.261 2.605–2.813

22.4 33.8 39.5 44.5 50.0 41.1

20.5–24.5 32.2–35.4 37.6–41.4 42.9–46.2 49.1–51.0 39.2–43.1

Health insurance Yes No Unknown

92.8 7.1 0.2

92.4–93.1 6.7–7.4 0.1–0.2

25.965 1.977 0.045

25.770–26.160 1.879–2.076 0.033–0.057

44.8 19.2 31.2

44.1–45.4 17.2–21.5 20.5–44.4

Access to regular health care provider Yes No Unknown

90.8 8.9 0.3

90.4–91.2 8.6–9.3 0.2–0.3

25.414 2.503 0.071

25.215–25.612 2.401–2.605 0.049–0.092

45.5 17.1 20.0

44.9–46.2 15.6–18.8 11.9–31.7

Marital status Married Never married Divorced, widowed, or separated

68.3 8.8 22.8

67.7–68.9 8.4–9.2 22.3–23.4

19.120 2.473 6.394

18.908–19.333 2.359–2.587 6.249–6.540

46.9 33.3 34.8

46.1–47.7 31.0–35.6 33.6–36.0

Metropolitan status Urban/suburban Rural Unknown

64.2 19.3 16.4

63.8–64.6 19.0–19.7 16.2–16.7

17.979 5.408 4.600

17.804–18.155 5.309–5.507 4.526–4.673

44.7 42.4 36.5

43.9–45.6 41.3–43.5 34.8–38.2

Race/ethnicity White, non-Hispanic Black, non-Hispanic Education level Did not graduate high school Graduated high school Attended college or technical school Graduated from college or technical university Unknown

P value

o0.001

o0.001

o0.001

o0.001

o0.001

o0.001

o0.001

o0.001

Informed of benefits of PSA screening Yes No

75.2 24.8

74.7–75.8 24.2–25.3

21.052 6.935

20.845–21.260 6.777–7.094

52.6 13.5

51.9–53.4 12.7–14.3

Informed of risks of PSA screening Yes No

32.0 68.0

31.4–32.6 67.4–68.6

8.945 19.042

8.763–9.127 18.836–19.248

50.9 39.2

49.7–52.0 38.4–39.9

Recommendation for PSA screening Yes No

70.6 29.4

70.0–71.1 28.9–30.0

19.749 8.239

19.543–19.955 8.064–8.413

55.7 12.2

55.0–56.5 11.5–13.0

o0.001

o0.001

424.e4

D. Pucheril et al. / Urologic Oncology: Seminars and Original Investigations 33 (2015) 424.e1–424.e7

for or against PSA testing ultimately affects a patient’s decision to be screened. To further elucidate these questions, we analyzed the self-reported responses of men captured in the 2012 BRFSS regarding information they received about PSA screening, their physician’s recommendation, and their ultimate decision to undergo screening. This study has a number of noteworthy findings. First, although the verdict is still uncertain on whether HCPs have significantly altered their practice regarding PSA screening, patients remain heavily influenced by their physicians. In our analysis, 46% of patients with access to a regular HCP received screening compared with only 17% of patients without consistent access to a HCP (Table 1). The recent implementation of the Affordable Care Act in the United States may improve access to primary care and thereby increase the prevalence of PSA screening. In fact, Aizer et al. found that the improved coverage fostered by the Affordable Care Act may translate into better oncologic outcomes among most young adults with cancer [6]. Further, our findings highlight and corroborate prior analysis of American cohorts demonstrating the positive effect of higher income bracket, educational attainment, and marital status on incidence of PSA screening [7–10]. Second, 71% (95% CI: 70.0–71.1) of patients stated that their physician recommended PSA screening (Table 1). Interestingly, among all examined covariates, receiving a screening recommendation resulted in the highest incidence of PSA screening (namely 56%, Table 1), which on multivariable analysis translated into an almost 5 folds higher probability of getting screened compared with patients not receiving a recommendation for screening (Table 3). Based on our literature search, this is the first report of the profound and direct effect on patient adoption of PSA screening based on the physician’s recommendation in favor of PSA screening. This finding may be expected as recommendations from a primary care provider have been demonstrated to be one of the strongest predictors of adherence to other cancer screening guidelines [11–13]. Intriguingly, there is evidence to support that the effect of

physician preferences on the odds of PSA screening is more pronounced relative to other screening tests. Specifically, Jaramillo et al. recently showed that 27% of the variance in PSA screening among Medicare beneficiaries in Texas was dependent on the primary care provider. By comparison, variance due to physician influence in mammography and colorectal cancer screening was only 10% and 9%, respectively [14]. Yet, unlike colorectal, cervical, and breast cancer, PSA screening is not supported by current guidelines [1,2] and may represent inappropriate use of health care resources, particularly in the older population. Another striking finding is the disproportion in the frequency in which men report being informed of the benefits (75%, 95% CI: 74.4–75.8) vs. the risks (32%, 95% CI: 31.4–32.6) of PSA screening (Table 1). This indicates that physicians have a bias toward the benefits of PSA screening, and therefore may fail to adequately counsel patients about the risks associated with screening and thus violate a basic tenet of shared decision-making. In fact, a Swedish survey of 305 physicians regarding their attitude toward PSA testing found that most physicians reported a positive attitude toward PSA testing, and that the physician’s opinion about the PSA test as a sufficient screening tool was the only variable affecting the physician’s decision of ordering PSA testing regardless of patient age [15]. A recent review by Hayes and Barry [16] reiterates that the available evidence regarding PSA screening favors clinical discussion of the pros and cons of PSA screening with average-risk men aged 55 to 69 years and only men who express a definitive preference for screening should have PSA testing. In light of these findings, it is critical that HCPs understand their inherent bias in favor of PSA screening and consciously make an effort to present a full and unbiased recommendation regarding whether PSA screening is right for the individual patient. Indeed, improved shared decision-making regarding PSA screening has been a focus of discussion within the family medicine literature [17,18], as has the development of various decision aids that can be used to assist physicians and

Table 2 Overall population of men 55 years or older, undergoing PSA screening; BRFSS, 2012

PSA recommended Yes by health provider

No

% Men who were informed about benefits of PSA screening (95% CI) % Men who were not informed about the benefits of PSA screening (95% CI) % Men who were informed about benefits of PSA screening (95% CI) % Men who were not informed about the benefits of PSA screening (95% CI)

% Men who were informed about risks of PSA screening (95% CI)

% Men who were not informed about the risks of PSA screening (95% CI)

55.6 (54.4–56.9)

57.0 (56.0–58.0)

31.9 (24.2–40.7)

47.3 (44.4–50.1)

20.6 (18.1–23.4)

29.8 (27.4–32.3)

15.3 (9.7–23.4)

5.4 (4.9–6.1)

Figures in each cell represent the percentage of men opting for PSA screening.

o0.001

o0.001

D. Pucheril et al. / Urologic Oncology: Seminars and Original Investigations 33 (2015) 424.e1–424.e7 Table 3 Multivariable analysis of likelihood to undergo PSA screening in men older than 55 years given information about benefits, risks, or recommendation of PSA screening, controlling for potential confounders including age, race, educational status, income category, insurance status, access to health care provider, marital status, metropolitan status, and information about benefits/risks/recommendation for PSA screening. BRFSS 2012 Screening information/ recommendation

Odds ratio (95% confidence interval) for undergoing screening

Informed of benefits of PSA screening Informed of risks of PSA screening Recommended to undergo PSA screening

2.404 (2.180–2.652) 0.917 (0.860–0.977) 4.979 (4.525–5.478)

patients to facilitate this process [19–21]. Specifically, Feng et al. found that a brief web-based interactive educational intervention for family medicine or internal medicine physicians improved shared decision-making behaviors regarding PSA screening in the intervention group compared with a control group. Further work should be done to determine if these types of interventions change long-term practice patterns [22]. Finally, our analysis of the 2012 BRFSS reveals noncompliance with the recently published, and aforementioned, guidelines [1,2]. Although 43% of men Z 55 years of age reported PSA screening in the year preceding the 2012 BRFSS, only 36% (95% CI: 34.8–37.5) of men aged 55 to 59 years received screening, yet 49% (95% CI: 47.2–50.5) and 46% (95% CI: 43.7–47.6) men aged 70 to 74 and 75 to 79 years reported screening, respectively (Table 1). These findings are consistent with previous studies examining various subpopulations within the United States. For example, studies using data from the National Health Interview Survey between 2000 and 2005 [9] and between 2005 and 2010 [23] showed similar distributions of PSA screening prevalence. Although differences in survey and sampling methodology limit direct comparisons, these data would suggest that the prevalence of PSA screening has remained stable between 2000 and 2012. These data are of concern, as they indicate no substantive decrease in screening frequency of men older than 75 years despite the USPSTF recommending against screening for men older than 75 years for the past 5 years [24]. More recently, however, Cohn et al. [25] analyzed an institutional database of patients seen by family practice or internal medicine physicians and found that there were significant decreases in the incidence of PSA screening in men aged 40 to 49 and 70 to 79 years in preUSPSFT (2007–2011) and post-USPSTF (2012) cohorts demonstrating compliance with the guidelines. The conflicting outcomes regarding physician compliance may be secondary to the smaller institutional cohort analyzed in the Cohn paper compared with the national cohorts presented in the National Health Interview Survey studies.

424.e5

Nonetheless, Cohn’s results are promising as they indicate that improved compliance is possible, especially at an institutional level. Furthermore, compliance with the guidelines may vary across specialty. For example, Kim et al. [26], after surveying a random sample of 1,366 radiation oncologists and urologists, found that urologists were more likely to screen men aged 40 to 49 and 50 to 59 years, but less likely to screen men older than 75 years compared with radiation oncologists. Our findings have several clinical implications. Primarily, physicians must understand the power of their recommendations. Our findings suggest that a physician’s recommendation for or against PSA screening is the most influential factor in determining whether a patient would undergo prostate cancer screening. Secondly, physicians must stay current with constantly changing guidelines and present their patients with the risks and benefits of PSA screening to allow the patient to make an educated and informed decision. Our findings clearly demonstrate that patients place a significant amount of trust and confidence in the recommendation of their individual physician, and thus physicians have a great responsibility to provide sound guidance to their patients. Significant limitations of our study include recall and nonresponse biases, and the failure of the BRFSS to record the screening preferences of the respondents. The tendency to recall/report screening is associated with certain study variables such as education, socioeconomic status, and characteristics of the intervention, thus introducing a bias toward underestimation of screening prevalence and incorrect recollection of physician recommendations. For example, patients are more likely to correctly recall a definitive diagnosis (e.g., prostate cancer) or invasive treatment/diagnostic test (e.g., prostatectomy or prostate biopsy) than a less invasive blood test (PSA screening) [27]. Additionally, when compared with medical record extraction, errors in self-reporting of PSA screening have been found to reflect underreporting [28]. Conversely, the BRFSS samples only individuals with telephones, and the relationship between telephone ownership and predictors of screening (including higher income and education) is likely correlated, therefore introducing bias toward overreporting of screening prevalence. Therefore, we emphasize that the findings of this study can be interpreted only within the confines of these notable limitations.

5. Conclusion In conclusion, patients are heavily swayed toward undergoing PSA screening based on the recommendation of their physician. Therefore, it is imperative that physicians are cognizant of their personal biases and take part in an unbiased shared decision-making process that is best suited to each patient’s individual attributes. Fortunately, much work is being done to develop tools to better facilitate this process between physicians and their patients.

424.e6

D. Pucheril et al. / Urologic Oncology: Seminars and Original Investigations 33 (2015) 424.e1–424.e7

Appendix A For prostate cancer screening questions, see Fig. A1.

QUESTION

QUESTION PROMPT AND ANSWER CHOICES

NUMBER 16.1

16.2

16.3

A Prostate-Specific Antigen test, also called a PSA test, is a blood test used to check men for prostate cancer. Has a doctor, nurse or other health professional EVER talked with you about the advantages of the PSA test? 1. Yes 2. No 7. Don’t Know/Not Sure 9. Refused Has a doctor, nurse, or other health professional EVER talked with you about the disadvantages of the PSA test? 1. Yes 2. No 7. Don’t Know/Not Sure 9. Refused Has a doctor, nurse, or other health professional EVER recommended that you have a PSA test?

16.4

1. Yes 2. No 7. Don’t Know/Not Sure 9. Refused Have you EVER HAD a PSA test?

16.5

1. Yes 2. No 7. Don’t Know/Not Sure 9. Refused How long has it been since you had your last PSA test?

16.6

1. Within the past year (anytime less than 12 months ago) 2. Within the past 2 years (1 year but less than 2 years) 3. Within the past 3 years (2 years but less than 3 years) 4. Within the past 5 years (3 years but less than 5 years) 7. Don’t know/Not sure 9. Refused What was the MAIN reason you had this PSA test – was it? 1. Part of a routine exam 2. Because of a prostate problem 3. Because of a family history of prostate cancer 4. Because you were told you had prostate cancer 5. Some other reason 7. Don’t know/Not sure 9. Refused

*

Available at http://www.cdc.gov/brfss/questionnaires/pdf-ques/2012_brfss.pdf

Fig. A1. Prostate cancer screening questions (module 16) excerpted from 2012 BRFSS*.

References [1] Moyer VA. Screening for prostate cancer: U.S. Preventive Services Task Force recommendation statement. Ann Intern Med 2012;157(2):120–34. [2] Carter HB, Albertsen PC, Barry MJ, et al. Early detection of prostate cancer: AUA guideline. J Urol 2013;190(2):419–26.

[3] Society AC. American Cancer Society recommendations for prostate cancer early detection, Available at: http://www.cancer.org/cancer/ prostatecancer/moreinformation/prostatecancerearlydetection/prostatecancer-early-detection-acs-recommendations; 2015. [4] Available at: http://www.cdc.gov/brfss/. http://www.cdc.gov/brfss/ [accessed 11.12.13].

D. Pucheril et al. / Urologic Oncology: Seminars and Original Investigations 33 (2015) 424.e1–424.e7 [5] Centers for Disease Control and Prevention (CDC). Behavioral risk factor surveillance system survey data, Available at: http://www.cdc.gov/brfss/ annual_data/2012/pdf/Overview_2012.pdf; 2012 [accessed 01.01.15]. [6] Aizer AA, Falit B, Mendu ML, et al. Cancer-specific outcomes among young adults without health insurance. J Clin Oncol 2014; 32(19):2025–30. [7] Woods VD, Montgomery SB, Herring RP, Gardner RW, Stokols D. Social ecological predictors of prostate-specific antigen blood test and digital rectal examination in black American men. J Natl Med Assoc 2006;98:492–504. [8] Hosain GM, Sanderson M, Du XL, Chan W, Strom SS. Racial/ethnic differences in predictors of PSA screening in a tri-ethnic population. Cent Eur J Public Health 2011;19:30–4. [9] Drazer MW, Huo D, Schonberg MA, Razmaria A, Eggener SE. Population-based patterns and predictors of prostate-specific antigen screening among older men in the United States. J Clin Oncol 2011; 29:1736–43. [10] Sammon JD, Pucheril D, Diaz M, et al. Contemporary nationwide patterns of self-reported prostate-specific antigen screening. JAMA Intern Med 2014;174:1839–41. [11] Wee CC, McCarthy EP, Phillips RS. Factors associated with colon cancer screening: the role of patient factors and physician counseling. Prev Med 2005;41:23–9. [12] Klabunde CN, Vernon SW, Nadel MR, Breen N, Seeff LC, Brown ML. Barriers to colorectal cancer screening: a comparison of reports from primary care physicians and average-risk adults. Med Care 2005;43:939–44. [13] Klabunde CN, Schenck AP, Davis WW. Barriers to colorectal cancer screening among Medicare consumers. Am J Prev Med 2006;30:313–9. [14] Jaramillo E, Tan A, Yang L, Kuo YF, Goodwin JS. Variation among primary care physicians in prostate-specific antigen screening of older men. J Am Med Assoc 2013;310(15):1622–4. [15] Hayat Roshanai A, Nordin K, Berglund G. Factors influencing primary care physicians’ decision to order prostate-specific antigen (PSA) test for men without prostate cancer. Acta Oncol 2013;52:1602–8. [16] Hayes JH, Barry MJ. Screening for prostate cancer with the prostatespecific antigen test: a review of current evidence. J Am Med Assoc 2014;311:1143–9.

424.e7

[17] Han PK, Kobrin S, Breen N, et al. National evidence on the use of shared decision making in prostate-specific antigen screening. Ann Fam Med 2013;11:306–14. [18] Li J, Berkowitz Z, Richards TB, Richardson LC. Shared decision making in prostate-specific antigen testing with men older than 70 years. J Am Board Fam Med 2013;26:401–8. [19] Evans R, Joseph-Williams N, Edwards A, et al. Supporting informed decision making for prostate specific antigen (PSA) testing on the web: an online randomized controlled trial. J Med Internet Res 2010;12:e27. [20] Taylor KL, Williams RM, Davis K, et al. Decision making in prostate cancer screening using decision aids vs usual care: a randomized clinical trial. JAMA Intern Med 2013;173:1704–12. [21] Williams RM, Davis KM, Luta G, et al. Fostering informed decisions: a randomized controlled trial assessing the impact of a decision aid among men registered to undergo mass screening for prostate cancer. Patient Educ Couns 2013;91:329–36. [22] Feng B, Srinivasan M, Hoffman JR, et al. Physician communication regarding prostate cancer screening: analysis of unannounced standardized patient visits. Ann Fam Med 2013;11:315–23. [23] Prasad SM, Drazer MW, Huo D, Hu JC, Eggener SE. 2008 US Preventive Services Task Force recommendations and prostate cancer screening rates. J Am Med Assoc 2012;307:1692–4. [24] U.S. Preventive Services Task Force: Screening for prostate cancer: U.S. Preventive Services Task Force recommendation statement. Ann Intern Med 2008;149:185–91. [25] 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;32:41e23–30. [26] Kim SP, Karnes RJ, Nguyen PL, et al. A national survey of radiation oncologists and urologists on recommendations of prostate-specific antigen screening for prostate cancer. BJU Int 2014;113:E106–11. [27] Sayre EC, Bunting PS, Kopec JA. Reliability of self-report versus chart-based prostate cancer, PSA, DRE and urinary symptoms. Can J Urol 2009;16(1):4463–71. [28] Hall HI, Van Den Eeden SK, Tolsma DD, et al. Testing for prostate and colorectal cancer: comparison of self-report and medical record audit. Prev Med 2004;39:27–35.