The learning curve of transrectal ultrasound-guided prostate biopsies: implications for training programs.

ABSTRACT

OBJECTIVE: To assess the quality of specimens obtained from prostate biopsies performed by urology residents and evaluate the number of procedures required to perform high-quality transrectal ultrasound (TRUS)-guided prostate biopsies. MATERIALS AND METHODS: Between 2006 and 2009, 770 patients underwent TRUS-guided prostate biopsies in our academic center. During the 6 semesters of this period, 24 residents (4 per semester) performed 1 session of 5.6±1.5 procedures each month for a total of 33.6±9 procedures during the study. The first session was performed with a senior urologist. Prostate cancer detection rate and standards of quality (average length of prostatic core biopsy specimens and absence of prostatic tissue) were retrospectively studied between the beginning and the end of each semester. RESULTS: A total of 12,760 biopsy cores were performed for 770 procedures. Mean patient age (64.5±6.1 years), and median prostate-specific antigen (8.7±3.7 ng/mL) were comparable between the study periods. The average length of biopsy cores significantly improved (+10%) from the first (12±2.7 mm) to the last month (13.2±2.1 mm) with a plateau after 12 procedures. Overall, cancer detection rate was 47% and was stable during the semester (41.3% the first month vs 44.1% the last month; P=.39). On univariate and multivariate analysis the mean length of biopsy specimens was associated with the number of procedures (P<.001) and the number of cores performed (P<.001). CONCLUSION: Twelve procedures are necessary to perform high-quality TRUS-guided prostate biopsies without compromising prostate cancer detection. In current training programs, we strongly recommend that residents have direct supervision for a minimum of 12 cases before they are allowed to perform TRUS-guided biopsies with indirect supervision.