The Learning Curve for Multi-parametric MRI/US Fusion Guided Prostate Biopsy: A Single Center Experience
Multi-parametric MRI (mpMRI) with ultrasound fusion targeted biopsy has been increasingly utilized as a diagnostic procedure for patients with suspected prostate cancer. Several aspects of fusion biopsy require learning, including lesion targeting and the operational knowledge of the device. As targeted biopsy gains further adoption in prostate cancer diagnostics, understanding the learning curve of the procedure will be helpful for institutions considering implementation into their practice.
We retrospectively reviewed 173 mpMRI/US fusion targeted biopsies performed utilizing the Artemis (Eigen) fusion biopsy device. Each biopsy was performed by one of five urologists with no prior experience performing fusion targeted biopsy. An average of 5 biopsy cores were obtained from each region of interest (ROI), followed by a 12-core systematic biopsy using a software generated template. Operative records were used to document the primary end point of length of procedure (LOP). Analysis of variance and chi-square tests were used to compare continuous and categorical variables respectively. Multiple linear regression was utilized to assess independent predictors of LOP.
Overall, LOP decreased with increasing operator experience. Average LOP for cases 1-10 was 30.36 minutes, (standard deviation 9.3). There was a significant decrease in average LOP with increasing biopsy experience; for cases 10-20, 21-30, 31-40 LOP was 25.1, 21.6, and 18.6 minutes, respectively (p
Results of our study demonstrate an improvement in LOP with increasing user experience, independent of number of ROIs. In addition, the number of ROIs were shown to independently influence LOP. Although use of this new technology is associated with a steep learning curve, our study demonstrates a substantial improvement within the first twenty procedures, suggestive of basic proficiency, and continued improvement with the second twenty procedures. Additional longitudinal data may further elucidate variables associated with physician learning curve.