Can a trained non‐physician provider perform transrectal ultrasound‐guided prostatic biopsies as effectively as an experienced urologist?

Abstract

WHAT'S KNOWN ON THE SUBJECT? AND WHAT DOES THE STUDY ADD?: PSA testing has resulted in a large number of patients being referred to urologists for investigation of potential prostate cancer. Despite limited evidence, non-physician providers now perform a number of routine urological procedures such as transrectal ultrasound-guided prostatic biopsies (TRUSP) in a bid to help relieve this increasing workload. In the largest series to date, we provide evidence that an adequately trained non-physician provider is able to perform TRUSP as effectively as an experienced urologist after an initial learning curve. To evaluate differences in cancer detection rates between a trained non-physician provider (NPP) and an experienced urologist performing transrectal ultrasound-guided prostatic biopsies (TRUSP) at a single UK institution. We retrospectively analysed a prospectively accrued database of patients (n = 440) referred for investigation of an abnormal digital rectal examination and/or a raised age-specific prostate-specific antigen (PSA) value undergoing first-time outpatient prostatic biopsies who were sequentially allocated to either an NPP or a physician-led TRUSP clinic. Differences in overall and risk-stratified prostate cancer detection rates were evaluated according to TRUSP operator. Continuous variables were analysed using Mann-Whitney U test whereas categorical variables were analysed using Pearson's chi-squared test. A multivariate binary logistic regression model was fitted for predictors of a positive biopsy. In all, 57.3% (126/220) of patients who underwent physician-led TRUSP were diagnosed with prostate cancer compared with 52.7% (116/220) in the NPP-led clinic (P = 0.338). Sub-group analysis revealed a lower cancer detection rate in men presenting with a low PSA level (<9.9 ng/mL) during the first 50 independent TRUSP procedures performed by the NPP (P = 0.014). This initial difference was lost with increasing case volume, suggesting the presence of a learning curve. Multivariate logistic regression analysis revealed age (odds ratio (OR) 1.054, 95% confidence interval (95% CI) 1.025-1.084, P ≤ 0.001), presenting PSA level (OR 1.05, 95% CI 1.02-1.081, P = 0.001), prostatic volume (OR 0.969, 95% CI 0.958-0.981, P ≤ 0.001) and clinical stage (OR 1.538, 95% CI 1.046-2.261, P = 0.029) to be predictors of a positive prostatic biopsy outcome. The choice of TRUSP operator was not predictive of a positive prostatic biopsy (OR 0.729, 95% CI 0.464-1.146, P = 0.171). An adequately trained NPP is able to perform TRUSP as effectively as an experienced urologist after an initial learning curve of 50 cases.