Date Published: July 7, 2017
Publisher: Public Library of Science
Author(s): Fumiaki Yano, Nobuo Omura, Kazuto Tsuboi, Masato Hoshino, Seryung Yamamoto, Shunsuke Akimoto, Takahiro Masuda, Hideyuki Kashiwagi, Katsuhiko Yanaga, Satoshi Inoue.
Although laparoscopic Heller myotomy and Dor fundoplication (LHD) is widely performed to address achalasia, little is known about the learning curve for this technique. We assessed the learning curve for performing LHD.
Of the 514 cases with LHD performed between August 1994 and March 2016, the surgical outcomes of 463 cases were evaluated after excluding 50 cases with reduced port surgery and one case with the simultaneous performance of laparoscopic distal partial gastrectomy. A receiver operating characteristic (ROC) curve analysis was used to identify the cut-off value for the number of surgical experiences necessary to become proficient with LHD, which was defined as the completion of the learning curve.
We defined the completion of the learning curve when the following 3 conditions were satisfied. 1) The operation time was less than 165 minutes. 2) There was no blood loss. 3) There was no intraoperative complication. In order to establish the appropriate number of surgical experiences required to complete the learning curve, the cut-off value was evaluated by using a ROC curve (AUC 0.717, p < 0.001). Finally, we identified the cut-off value as 16 surgical cases (sensitivity 0.706, specificity 0.646). Learning curve seems to complete after performing 16 cases.
Laparoscopic surgery is less invasive than open abdominal surgery, and has come into broad use because it is excellent from the viewpoint of cosmetic results . Additionally, in deep tissue regions such as the esophageal hiatus, mediastinum, and pelvis, laparoscopy can provide superior visibility compared with the naked eyes . However, one disadvantage is that tactile sensation is minimized thorough instruments. In spite of reports of serious complications and patient death [3, 4], laparoscopic surgery continue to increase each year. The Endoscopic Surgical Skill Qualification System was established in Japan to promote safer and more appropriate laparoscopic operations . Becoming certified as a qualified surgeon is a major goal for today’s gastrointestinal surgeons in Japan.
All statistical analyses were performed using SPSS version 22.214.171.124 (Armonk, NY, USA). Medians with interquartile ranges (IQR) are expressed for continuous variables. Chi-square test was used to compare categorical variables. Mann-Whitney’s U test were used to compare continuous variables. A ROC curve analysis was used to identify the cut-off value of the surgical experience number which was defined as the completion of the learning curve. Relationships between operation time, amount of blood loss, intraoperative complications, postoperative reflux esophagitis, and satisfaction scores, as reflected in questionnaire responses, were assessed using a Spearman’s rank correlation test. The level of significance was set at a p value <0.05. There have been reports pertaining to learning curves in various types of laparoscopic surgery, such as laparoscopic distal pancreatectomy and laparoscopic partial nephrectomy. Ricci et al.  examined 32 cases involving laparoscopic distal pancreatectomy performed by a single surgeon (pancreatic surgeon), with the primary endpoint being operation time and the secondary endpoint being the conversion to laparotomy, rate of complication, mortality rate, postoperative hospital stay, and number of unscheduled splenectomies. The ROC curve is used as a statistical method to calculate the cut-off value for the learning curve, as was conducted in the current study. The results indicated that conversion to laparotomy, incidence of complications, mortality rate, and postoperative hospital stay did not affect the surgeons’ ability to overcome the learning curve, and unscheduled splenectomies were performed relatively earlier than the completion of the learning curve. Based on these observations, the cut-off value was concluded to be approximately 17 surgical cases. However, Osaka et al.  evaluated 63 laparoscopic partial nephrectomies performed by one surgeon to address clinical T1a renal masses. To assess the results of surgery, predictive factors for achieving the designated performance targets were evaluated using three parameters: ischemia time of 25 minutes or less, negative margin, and no intraoperative complications. These 3 targets were met in 39 of 63 cases. Tumor size and the distance from the urinary tract were correlated with achievement of the 3 targets. Moreover, if separated by a learning curve of 30 cases, they concluded that the size of the tumor and the learning curve of the operator were predictors for achieving the 3 targets because the difference upon achievement of the targets was significant. Through this study, we estimated the learning curve of laparoscopic Heller myotomy and Dor fundoplication for achalasia can be plateaued by performing 16 cases. Source: http://doi.org/10.1371/journal.pone.0180515