BACKGROUND: Measurement of serum prostate-specific antigen (PSA) and digital rectal examination (DRE) are commonly used for prostate cancer screening. Clinicians have been advised to avoid measuring serum PSA after DRE because of the possibility of spurious elevation. However, studies in healthy volunteers have found no change. We sought to determine whether DRE performed by internists affects PSA levels in a primary care clinic population. METHODS: Men older than 49 years enrolled in a Veterans Affairs internal medicine clinic were eligible for the study. For all patients, prostate size and nodularity were assessed by DRE. Blood for determination of PSA levels was drawn immediately before DRE and 30 minutes after DRE. All patients completed a questionnaire regarding voiding. RESULTS: Two hundred two men aged 50 to 85 years (mean, 67 years) were enrolled and had complete data. The mean PSA level increased by 0.26 micrograms/L after DRE (P < .001). Six patients (2.9%; 95% confidence interval, 0.6% to 5.3%) had an increase in PSA level from less than 4 micrograms/L to more than 4 micrograms/L after DRE. There was a statistically significant but weak correlation between voiding score and PSA level (coefficient, .17; P = .02). Patients with an enlarged prostate gland had higher PSA levels than did patients with normal prostates or borderline results of examination. Nodularity did not predict an increased PSA level. CONCLUSIONS: We conclude, in an elderly veteran population, that DRE causes a statistically significant but clinically insignificant increase in serum PSA level. Additionally, we found that symptoms and physical examination results cannot be used to select a population for whom PSA screening would likely be useful.
BACKGROUND: Measurement of serum prostate-specific antigen (PSA) and digital rectal examination (DRE) are commonly used for prostate cancer screening. Clinicians have been advised to avoid measuring serum PSA after DRE because of the possibility of spurious elevation. However, studies in healthy volunteers have found no change. We sought to determine whether DRE performed by internists affects PSA levels in a primary care clinic population. METHODS:Men older than 49 years enrolled in a Veterans Affairs internal medicine clinic were eligible for the study. For all patients, prostate size and nodularity were assessed by DRE. Blood for determination of PSA levels was drawn immediately before DRE and 30 minutes after DRE. All patients completed a questionnaire regarding voiding. RESULTS: Two hundred two men aged 50 to 85 years (mean, 67 years) were enrolled and had complete data. The mean PSA level increased by 0.26 micrograms/L after DRE (P < .001). Six patients (2.9%; 95% confidence interval, 0.6% to 5.3%) had an increase in PSA level from less than 4 micrograms/L to more than 4 micrograms/L after DRE. There was a statistically significant but weak correlation between voiding score and PSA level (coefficient, .17; P = .02). Patients with an enlarged prostate gland had higher PSA levels than did patients with normal prostates or borderline results of examination. Nodularity did not predict an increased PSA level. CONCLUSIONS: We conclude, in an elderly veteran population, that DRE causes a statistically significant but clinically insignificant increase in serum PSA level. Additionally, we found that symptoms and physical examination results cannot be used to select a population for whom PSA screening would likely be useful.