To the editor:
The conclusion by Collins and colleagues1 to their recent interesting paper on age-specific patterns of prostate-specific antigen (PSA) testing is a little misleading. In their computation of the primary care physician visits with PSA testing by age, the authors chose to include all persons aged 18 to 49 years in one stratum. They subsequently obtained a relatively low proportion of visits with PSA testing of 1.4% for persons younger than 50 years (actually 18-49 years) compared with 5.6% for persons aged 50 to 59 years and 7.0% for those aged 70 to 74 years. They then concluded that the PSA screening decisions by primary care physicians during the mid-1990s were not sensitive to age. However, we know that the risk of developing prostate cancer increases with age and that more than 75% of all prostate cancers are diagnosed in men older than 65 years.2 Prostate cancers and other prostate conditions are very rare in men younger than 40 years. In spite of this, Collins and coworkers included all persons aged 18 to 49 years in the first age stratum (50 years) in their analysis.
Inclusion of persons aged 18 to 40 years, who incidentally may have represented a substantial number of visits, would inflate the denominator while maintaining the numerator, with a resultant underestimation of the computed proportion. The least the authors could have done was to create another age stratum encompassing men aged 40 to 49 years, given that the age range for requesting PSA testing recommended by the American Cancer Society and the American Urological Association begins at age 40. One wonders what their results would look like if age was uniformly stratified-for example, using uniform national census age groupings. An advantage of using standardized age groupings is to permit comparison of study results from one setting to another.
Samuel N. Forjuoh, MB, ChB, DrPH
Glen R. Couchman, MD
Don B. Cauthen, MD
Scott & White Memorial Hospital
Temple, Texas
- Collins MM, Stafford RS, Barry M. Age-specific patterns of prostate-specific antigen testing among primary care physician visits. J Fam Pract 2000; 49:169-72.
- American College of Physicians. Screening for prostate cancer. Ann Intern Med 1997; 126:480-84.
- Forjuoh SN, Weiss HB, Coben JH, Garrison HG. The standardization of statewide hospitalized injury morbidity reports. Am J Public Health 1995; 85:732-33.
The preceding letter was referred to Drs McNaughton Collins, Stafford, and Barry, who responded as follows:
Our study examined the effect of patient age on patterns of prostate-specific antigen (PSA) testing and found relatively high rates of testing in both young men and the most elderly men. Drs Forjuoh, Couchman, and Cauthen requested information on the proportion of primary care physician visits with PSA testing among men aged 40 to 49 years. Analysis of our data indicated that PSA testing occurred in 3.6% of visits by men 40 to 49 years, compared with 5.6% by men aged 50 to 59 years and approximately 7.0% by men aged 60 to 69 years. Although the American Cancer Society and the American Urological Association have recommended that clinicians begin screening high-risk men (black men and men with a family history of prostate cancer) at age 40, the United States Preventive Services Task Force and the American College of Physicians clinical guideline have not recommended screening high-risk men any differently because of the lack of direct or indirect evidence quantifying the value of earlier or more aggressive screening. Although Forjuoh and colleagues correctly point out that age is a risk factor for prostate cancer, our finding that PSA screening occurred in more than 3% of visits by men 80 years or older strongly suggests that some prostate cancer screening decisions by primary care physicians are not as sensitive to age as national guidelines suggest they should be. Increasing overall mortality rates at older ages works against any potential benefit of PSA screening in these men.
Mary McNaughton Collins
Randall S. Stafford
Michael J. Barry
Massachusetts General Hospital
Boston