|
Back to 2016 Annual Meeting
The Learning Curve for Minimally Invasive Esophagectomy: Impact of Experience on Postoperative Complications and Operative Times
Patrick D. Lorimer*1, Danielle M. Boselli2, Joshua S. Hill1, Jonathan C. Salo1
1Surgery, Levine Cancer Institute, Carolinas Medical Center, Charlotte, NC; 2Biostatistics, Levine Cancer Institute, Carolinas Medical Center, Charlotte, NC
Introduction
The impact of surgical volume on patient outcomes is well described. Additionally, longer operative times have been associated with increased complications. As experience with procedures is gained, operative times tend to decrease. In this era of value-based care, a thorough understanding of mechanisms contributing to complications is necessary. This study represents the learning curve of a single surgical team at a quaternary referral center.
Methods
Data from a prospectively collected institutional database of esophageal resections for cancer from 2007-2015 was reviewed. Patients were divided into groups by chronological order of surgery (group 1 = 1-50, 2 = 51-100, 3 = 101-160). Operating room times were analyzed using ANOVA to determine differences between groups. These differences were used to identify a learning curve. Groups 1 and 2 were combined and a grafted polynomial linear regression model was estimated to describe the institutional trend in operating room times prior to and following operative case 100. Odds of postoperative complications in cases 101-160 versus 1-100 were estimated utilizing logistic regression modeling.
Results
160 patients met inclusion criteria; median age 64, 80% male and 90% White. Twenty-six (16%) patients were clinically Stage 1/is, 21 (13%) Stage 2, 112 (70%) Stage 3, and 1 (1%) Stage 4. Odds of pneumonia and ARDS were lower in cases >100 versus ≤100. (OR: 0.14 (95%CI 0.04-0.47), p<0.05) and (OR: 0.12 (95%CI 0.02-0.98), p<0.05), respectively). Odds of other postoperative complications, including pulmonary embolism, prolonged ventilation, MI, anastomotic leak, esophageal stricture, stroke, and renal failure were also decreased in this group.
Mean operating times decreased by chronologic group; group 1 had an average operating room time of 557 minutes (Standard Deviation [SD]: 80.9), 511 minutes (SD: 112) in group 2, and 432 minutes (SD: 88) in group 3 (Figure 1). Operating times between all three groups were significantly different (p<0.001), as well as all pairwise comparisons between the groups (1 v. 2 p=0.021, 2 v. 3 p=0.001, 1 v. 3 p<0.001). The decrease in operating room time between surgical cases 1 and 100 was 1.6 minutes per case (p<0.001). For operative cases ≥ 100, the estimated slope shows a gradual decrease in operating room time of 0.2 minutes per case (p<0.05). When adjusting for AJCC staging and histology the difference in slope remains statistically significant (p<0.05).
Conclusions
Postoperative complications and operating room time decreased over the course of our experience. The most dramatic improvements came within the first 100 cases. After this point, operating room time continued to decrease, but not in a significant fashion. This is also the inflection point after which complications began to decrease significantly for our surgical team.
 
Back to 2016 Annual Meeting
|
