Resident participation in laparoscopic roux-en-Y gastric bypass: a comparison of outcomes from the ACS NSQIP database

Laura Doyon, MD1, Alejandro Moreno-Koehler, MPH2, Dmitry Nepomnayshy, MD, FACS1. 1Lahey Hospital and Medical Center, 2Tufts University School of Medicine

Background: As clinical outcome data becomes more closely tied with hospital reimbursement, understanding the balance between providing quality care and training surgical residents becomes more important. We sought to determine if resident participation in this type of complex laparoscopic bariatric case affected 30 day morbidity and mortality of patients using the ACS NSQIP database.

Methods: We queried the ACS NSQIP database from 2005-2012 using CPT codes for laparoscopic gastric bypass. Cases were divided between those involving a resident and those performed by an attending alone. Preoperative, operative, and postoperative data were analyzed. Univariate and multivariate analyses of intraoperative and postoperative outcomes were assessed, including operative time, length of stay, mortality, morbidity, and readmission. Resident groups were divided into Junior (R1-3), Senior (R4+5), and Fellow (R6+), for subanalyis to address whether different level of residents affected outcomes. Results were examined for clinical relevance as well as statistical significance.

Results: Preoperative characteristics were similar between the two groups. Of the 43,477 laparoscopic gastric bypass cases available for analysis, 22,189 had resident involvement (Resident = R), and 21,288 did not (No resident = NR). There was no clinical difference in the following outcomes: length of stay(R = 2.59 days v. NR= 2.47 days, p<0.0001), return to the operating room (R= 2.8% v. NR= 2.7%, p=0.586), pulmonary embolism (R=0.2% v. NR=0.2%, p=0.676), organ space infection [R = 0.7% v. NR = 0.6%, p=0.439, OR 1.02 (95% CI 0.99-1.01)], or mortality [OR 1.15 (95% CI 0.70-1.88)]. On multivariate analysis the resident group had increased risk of complications, although the overall incidence was still low: superficial site infection [R=2.1% v. NR =1.5%, p<0.0001, OR = 1.47 (95% CI 1.27-1.70)], renal insufficiency [R= 0.2% v. NR= 0.1%, p= 0.002, OR 2.26 (95% CI 1.34 – 3.83)], urinary tract infection [R = 1.1% v. NR = 0.9%, OR 1.26 (95% CI 1.03-1.54], sepsis [R= 0.8% v. NR = 0.6%, OR 1.29 (95% CI 1.02 -1.62)]. Readmission data was available from 2011 to 2012 (n = 3686 cases) and revealed on multivariate analysis that the resident group had an increased risk of readmission [R = 8.6% v. NR = 6.7%, p = 0.008, OR 1.55 (95% CI 1.17-2.06)]. There was no increased risk of organ space infection. Operative time was increased in the resident group by 29.3 minutes (p<0.0001). Resident group subanalysis revealed that operative time increased with resident seniority: Junior 141.6 minutes, Senior 145.6 minutes, Fellow 164.9 minutes. Other preoperative and postoperative outcomes were similar.

Conclusion: Resident participation in laparoscopic roux-en-Y gastric bypass was associated with increased incidence of several adverse outcomes which were statistically significant, although still infrequent in both groups. The greatest absolute difference appeared to be in longer operative time, especially with increased seniority of the trainee. This can likely be attributed to increased participation in the procedure. Although small differences in the rates of some complications are statistically significant given the large dataset, there seems to be little clinically significant adverse effect of resident involvement in operative care of gastric bypass patients.

« Return to SAGES 2015 abstract archive