BRIEF REPORT: Physicians and Their Personal Prostate Cancer-screening Practices with Prostate-specific Antigen

A National Survey

Authors


  • The authors have no conflicts of interest to report.This study was presented at the European Society of Medical Decision Making. Erasmus MC Universitaire Medisch Centrum Rotterdam, Rotterdam, the Netherlands on June 7, 2004 and at the Society of General Internal Medicine, Chicago, IL on May 14, 2004.

Address correspondence and requests for reprints to Dr. Chan: Biomedical Ethics, Division of General Internal Medicine, Department of Medicine, The University of Texas-Houston Medical School, 6431 Fannin, 1.122 MSB, Houston, TX 77030 (e-mail: Evelyn.C.Chan@uth.tmc.edu).

Abstract

BACKGROUND: There is inconclusive evidence that prostate cancer screening with prostate-specific antigen (PSA) reduces mortality. Although PSA testing is widespread, it is unknown how many physicians have taken the PSA test themselves.

OBJECTIVE: To determine the prevalence of PSA testing among physicians.

DESIGN: Cross-sectional survey.

SUBJECTS: A nationwide stratified random sample of urologists (response rate 61%, n=247), Internists (response rate 51%, n=273), and family physicians (response rate 64%, n=249) were surveyed by mail in 2000. After excluding female respondents and men who either reported a positive history of prostate cancer or did not respond to that query, there were 146 urologists, 96 Internists, and 118 family physicians.

MEASUREMENTS: Whether physicians had undergone prostate cancer screening with PSA.

RESULTS: Eighty-seven percent (155/178) of male physicians aged, 50 and older and 21% (31/150) of white male physicians under age 50 reported having had a PSA test. More urologists than nonurologists in both age groups reported having had a screening PSA test.

CONCLUSION: Most physicians aged 50 and older report undergoing PSA testing. This may reflect a belief in its efficacy and contribute to its widespread use.

Screening for prostate cancer with prostate-specific antigen (PSA) is controversial because there is inconclusive evidence that regular testing reduces mortality.1–3 Until clinical trials4,5 determine whether there is a net benefit, professional organizations recommend that physicians inform men of the potential risks and benefits of testing through informed or shared decision making.6–10 These risks include potential harms, such as false-positive results and unnecessary anxiety, biopsies, and potential treatment complications for some cancers whose natural course may not lead to morbidity and mortality.10

Despite its uncertain efficacy, PSA testing has become widespread.11–16 It has become even more prevalent than colorectal cancer screening, an intervention with proven efficacy and unanimous support among professional organizations.14,16

Although physicians have contributed to the widespread uptake of PSA testing among men,11–16 it is unknown just how many U.S. physicians have taken the PSA test themselves. One study in Victoria, Australia, found evidence to support high levels of PSA testing (45%) among physicians there.17 In this study, we determined the prevalence of personal PSA testing among a national sample of urologists, family physicians and Internists.

METHODS

We obtained institutional board review approval from the University of Texas-Houston Health Sciences Center to draw a nationwide random sample of physicians. The sample was stratified by specialty from the Official ABMS Directory of Board-Certified Medical Specialists 2000, 32nd edn, Vols. 1–4. We determined physician eligibility and mailed surveys and postcard reminders as described previously.18 Eligible physicians practiced medicine at least 20 hours/week, cared for men aged 40 and older, and were not in residency training. In 2000, we surveyed 273 eligible internists, 249 family physicians, and 247 urologists. We received completed surveys from 139 internists, 160 family physicians, and 151 urologists (overall response rate: 59%). For our analysis, we dropped 84 women, 2 men who reported a personal history of prostate cancer, and 4 men who did not respond to that question, leaving 96 Internists, 118 family physicians, and 146 urologists. Because family physicians and internists had similar demographic characteristics and responses, we combined them as “nonurologists” and compared them with urologists.

We asked physicians: “Have you ever undergone prostate cancer screening with prostate specific antigen?” We also ascertained respondents' age, years with a medical degree, race/ethnicity, and marital status.

For men at average risk for prostate cancer, professional organizations recommend offering PSA testing beginning at age 506–10; therefore, the primary focus of our analysis was on male physicians aged 50 and older since they would be potential candidates for PSA testing. As a secondary focus, we examined white male physicians under the age of 50. Men under age 50 are not candidates for PSA testing unless they are at a higher than average risk for prostate cancer (e.g., African Americans). Therefore, urologists and nonurologists under the age of 50 who are white would not be expected to have undergone PSA screening themselves. We used χ2 tests to compare the frequency distributions.

RESULTS

The demographic characteristics of survey respondents are shown in Table 1. Of male physicians aged 50 and older, 87% (95% confidence interval: 81.2% to 91.6%) (155/178) reported having had a screening PSA test. Among physicians aged 50 and older, more urologists (95%) than nonurologists (78%) reported a PSA test (Table 2). Among white male physicians under the age of 50, 21% (95% confidence interval: 14.5% to 28.0%) (31/150) reported a PSA screening test, and more urologists (40%) than nonurologists (13%) reported being screened (Table 2).

Table 1. Demographic Characteristics of Male Physician Respondents*
 Age 50 and OlderUnder Age 50
Urologists (n=98)Nonurologists (n=80)P ValueUrologists (n=48)Nonurologists (n=134)P Value
  • *

    Two sample t-tests were used to compare the mean values of continuous variables between physicians in each specialty. χ2 tests for independence were used between categorical variables and physician specialty.

  • Missing data ranged from 0% of respondents (age) to 5.6% (years since having medical degree).

  • Missing data ranged from 1.1% of respondents (age, race, marital status) to 4.4% (years since having medical degree).

Mean age, y±SD58.9±6.158.3±6.4.55544.4±4.242.3±5.2.013
Mean years with medical degree±SD32.5±7.130.5±8.6.11019.0±7.415.6±6.1.003
Ethnic origin (%)  .826  .430
 White84.788.6 89.681.1 
 Black2.01.3 2.11.5 
 Hispanic1.01.3 0.06.1 
 Asian8.27.6 8.310.6 
 Other4.11.3 0.00.8 
Marital status (%)  .345  .436
 Married89.893.8 89.685.1 
 Not currently married10.26.3 10.414.9 
Table 2. Experience with Prostate-Specific Antigen (PSA) Screening Among Male Physicians
 Urologists (%)Nonurologists (%)P Value*
(95% Confidence Interval), n=98(95% Confidence Interval), n=80
Aged 50 and older
 Had PSA screening93/98 (95%)62/80 (78%).0006
(88.5% to 98.3%)(66.8% to 86.1%) 
 Urologists (%)Nonurologists (%)P Value*
(95% Confidence Interval), n=43(95% Confidence Interval), n=107
  • *

    P value based on χ2 test.

Under age 50 (whites only)
 Had PSA screening17/43 (40%)14/107 (13%).0003
(25.0% to 55.6%)(7.3% to 21.0%) 

DISCUSSION

The 2000 U.S. National Health Interview Survey (NHIS) found that 57% of men aged 50 and older with no history of prostate cancer reported ever having had a PSA test.16 Although NHIS data are weighted to provide estimates representative of men in the general population16 and our data are not (and therefore may have been skewed by incomplete survey participation), our results suggest that PSA screening may be more widespread among physicians than in the general population.

Despite professional guidelines against offering the PSA test to men under the age of 50 in the absence of risk factors for prostate cancer, we found a high percentage of physicians under the age of 50 who had undergone screening. The 2001 Behavioral Risk Factor Surveillance System (BRFSS) survey14 found that 34% of the men in the general population aged 40 to 49 years reported having had a PSA test. The prevalence of PSA screening that we found for white urologists under the age of 50 is higher than would be expected if they were testing based on a family history of prostate cancer, which is found in only 10% of men in the general population.19

Our findings suggest that physicians may be requesting and, at the least, are not refusing PSA screening when offered, which in turn suggests that physicians believe that the potential harms of PSA testing do not outweigh the potential benefits. It is possible that they may be adhering to perceived standards of care,14,15 believe that PSA screening is beneficial,15 or appreciate the psychological reassurance of a negative PSA screening test result.20

Compared with nonurologists, more urologists reported undergoing PSA screening, regardless of age. We previously found that urologists are less likely to consider that the uncertainties associated with PSA testing are important for men to know.18 Although attitudes do not necessarily predict behavior, urologists may hold a more favorable attitude toward screening with PSA.

Our results also have implications for the implementation of informed decision making. Although professional organizations recommend informed decision making for PSA testing,6–10 having had the PSA test, or harboring implicit beliefs that the risk/benefit ratio of PSA testing is more favorable than evidence indicates may make physicians biased in their discussions of PSA testing with patients.

Early detection of cancer may save lives; however, it also has the potential to disrupt lives and cause unnecessary harm. Even without the completion of randomized trials to assess efficacy, physicians are already undergoing PSA testing. Efforts to implement informed decision making for screening with PSA may need to take into account presumptions not only among patients, but also among physicians about the perceived benefits of screening with PSA.

Acknowledgments

This study was funded by National Cancer Institute grant K08-CA78615, awarded to Dr. Chan as a clinical scientist award. Additional support was provided by Centers for Disease Control/Association of Schools of Public Health, grant #S1171-19/20, and NIH grant M01-RR02558, the Clinical Research Center at the University of Texas-Houston.

Contributions of Authors: Dr. Chan and Dr. Barry participated in the conceptualization of the study, and with Dr. Vernon and Dr. Ahn designed the study and analyzed the data. All authors had access to all data in the study and held final responsibility for the decision to submit this work for publication. All authors approved of the final version.

Conflict of Interest Statement for Authors: The authors declare that they had no financial or personal conflicts of interest that could have inappropriately biased this study. Dr. Chan had full access to all data in the study and had final responsibility for the decision to submit this manuscript for publication.

Ethics Approval: The authors obtained institutional board review approval from the Committee for the Protection of Human Subjects at the University of Texas-Houston Health Sciences Center, Houston, TX, U.S.A., to conduct the study.

Ancillary