We examine differences in screening, detection, staging and treatment of prostate cancer between urologists in the United States and Canada.
An anonymous questionnaire was developed and mailed to 700 randomly selected American and 350 Canadian urologists. The 7 domains of prostate cancer management comprised screening/case identification, radical prostatectomy indications, staging evaluations, neoadjuvant therapy, nerve sparing techniques, postoperative management and treatment of biochemical recurrence. The Dillman method of questionnaire administration was used. All data were stratified by country and analyzed using the generalized chi-square test.
Surveys were adequately completed by 45% and 79% of American and Canadian urologists, respectively. Practice experience and clinical volumes were not significantly different between the 2 cohorts. Overall, there were few differences in prostate cancer screening, staging, postoperative management, biochemical failure and use of neoadjuvant therapy. However, practicing American urologists tended to pursue more aggressive case finding practices, such as a higher age cutoff for prostate specific antigen testing (p = 0.0001) and more frequent use of transition zone biopsies (p = 0.0001). American urologists also displayed a tendency toward more liberal indication for extirpative surgery. They were more likely to perform radical prostatectomy in men older than 70 years, those with higher prostate specific antigen and those with node positive disease. Among both national cohorts there was considerable variation in management patterns for all domains of prostate cancer. Variation was most common among treatment of patients with adverse pathological conditions (positive margins, seminal vesicle involvement) and postoperative biochemical failure. Even when credible evidence exists (biopsy technique, preoperative staging) significant proportions of urologists in both countries continued to practice contrary to existing data.
American and Canadian practice patterns for prostate cancer differ significantly only in the domains of case identification and surgical indications. In addition, considerable intra-national variation in practice patterns exists. These data highlight the necessity to support randomized clinical trials in prostate cancer.