Defining the learning curve for robotic-assisted endoscopic harvesting of the left internal mammary artery

Robotic-assisted techniques are continuing to cement their role in coronary surgery, particularly in facilitating the endoscopic harvesting of the left internal mammary artery (LIMA), regardless of how the subsequent bypass grafting is performed. As more surgeons attempt to become trained in robotic-assisted procedures, we sought to better define the learning curve associated with robotic-assisted endoscopic LIMA harvest.

Between January 2011 and July 2012, a total of 77 patients underwent robotic-assisted minimally invasive direct coronary artery bypass surgery at our institution. The LIMA was harvested endoscopically in all patients, using standard robotic instruments, followed by direct grafting to anterior wall myocardial vessels via a small thoracotomy. Intraoperative times for various components of the procedure were collated and analyzed.

The mean ± SD time taken to insert and position the ports for the robotic instruments was 3.9 ± 1.4 minutes. The mean ± SD LIMA harvest time was 31.8 ± 10.1 minutes, and the mean ± SD total robotic time was 44.2 ± 12.9 minutes. All time variables consistently continued to decrease as the experience of the operating surgeon increased, with the greatest magnitude of improvement being evident within the first 20 cases. The logarithmic learning curves for LIMA harvest time and total robot time during our entire experience were both calculated as 90%, correlating to an expected 10% improvement in performance for each doubling of cases completed.

Coronary surgeons can rapidly become proficient in robotic-assisted endoscopic LIMA harvest, with significant improvement in operative times evident within the first 20 cases completed. These data may be useful in designing appropriate training programs for newer surgeons seeking to gain experience in robotic-assisted coronary surgery.