Learning Curves for Robot-Assisted and Laparoscopic Partial Nephrectomy

Authors:
Publisher website
Google Scholar
Abstract
Objectives: To evaluate the learning curve of robot-assisted partial nephrectomy (RAPN) and laparoscopic partial nephrectomy (LPN) between two surgeons at a single institution. Methods: A prospectively maintained, Institutional Review Board (IRB)-approved kidney surgery database was reviewed retrospectively and the first 116 consecutive LPNs performed by one surgeon (Hyung Kim) and 116 consecutive RPNs performed by a second surgeon (Thomas Schwaab) were identified. The learning curve was evaluated by examining the operative times, warm ischemia times (WITs), estimated blood loss, the postoperative estimated glomerular filtration rate (eGFR), and intra- and postoperative complications in the quartiles of 29 patients. The LPNs performed by Hyung Kim were done following completion of a minimally invasive fellowship. Thomas Schwaab had minimal experience with LPN and no fellowship training before starting RAPN. Results: The RAPN and LPN groups had similar patient and tumor characteristics. The RAPN group had a higher preoperative eGFR (74.1±22.04 vs. 80.95±21.25 mL/minutes, p =0.015) and a worse Eastern Cooperative Oncology Group (ECOG) performance status (ECOG 1+ in 12% vs. 2.6%, p p =0.203) and postoperative ( p =0.193) complications were similar. In the RAPN group, operating room (OR) time (161±51 vs. 203±55 minutes, p vs. 21.8±9.1 minutes, p vs. 1.67±1.1 days, p versus the LPN (9.6% vs. 10%). The learning curves differed for log tumor size, log WIT, and postoperative complications. Conclusions: The variables of the learning curve for RAPN can be obtained earlier than the same variables for LPN. RAPN had a shorter OR time and WITs. The shorter WITs, earlier in the series, led to consistently lower fluctuations in GFR and preservation of the renal function. The learning curves for each procedure need to be re-evaluated at longer intervals to ensure their accuracy.

This publication has 20 references indexed in Scilit: