Predictors of Outcomes Following Roux En Y Gastric Bypass Surgery at the Ottawa Hospital

Chris G Smith, Dr, Balpreet G Brar, Dr, Joseph Mamazza, Dr, Husein Moloo, Dr, Haggar Fatima, Ms, Jean Denis Yelle, Dr, Robert Dent, Dr, Isabelle Raiche, Dr. The Ottawa Hospital


Introduction: Despite a low complication rate, the consequences of complications from bariatric surgery can be devastating. The purpose of this study was to identify predictors of outcomes in patients undergoing laparoscopic roux en y gastric bypass.

Methods and Procedures: Prospective analysis of patients undergoing roux en y gastric bypass at an academic center from November 2007 to July 2011. Exclusion criteria were: patients who underwent revisional surgery, those who had surgery at another center, and those who had incomplete follow up. Primary outcome measure was incidence of complications including: anastomotic leak, abscess, pulmonary or cardiac complications, internal hernia, stricture, renal failure and wound infection. Complications were pooled to create a single binary variable. Multiple logistic regression was used to identify predictors of post operative complications as well as length of stay. Linear regression was used to explore factors predictive of weight loss. Surgeon factors such as OR time and Patient factors including: age, BMI, comorbidities, distance from bariatric center, percent excess weight loss, and number of visits with the nurse practitioner and behaviourist preoperatively were explored.

Results: Data on 377 patients was analyzed with a complication rate of 18.7%. Mean duration of follow up was 7.2 months. The mean age, BMI, and OR time were 46.1+/-10.3 years, 47.9+/-6.8, and 154.6+/-38.2 minutes respectively. Mean excess weight loss was 82.2%. Univariate analysis identified BMI (odds ratio [OR] 1.058, 95% CI 1.019-1.098) and history of hypertension (OR 2.086, 95% CI 1.231-3.536) as being significantly associated with the incidence of complications. Multivariate analysis with forward selection showed that BMI alone (OR 1.043, 95% CI 1.000-1.088) was significantly associated with an increased risk of developing complications. Separate logistic models were used to analyze each complication type as a response variable. This identified hypertension (OR 2.027, 95 % CI 1.034-3.973) as being significantly linked to the incidence of wound infection. Number of visits with the behaviourist preoperatively (OR 2.375, 95% CI 1.289-4.375) was associated with a higher number of complications occurring greater than 90 days after surgery. When looking at length of stay as the response variable, only type 2 diabetes mellitus was shown to be significantly related (p = 0.0064). Linear regression analysis showed that increased BMI was negatively associated with excess weight loss (parameter estimate -1.43, p = 0.0003).

Conclusion: In our experience, BMI, hypertension, and number of visits to the behaviourist preoperatively were shown to be associated with a higher risk of complications post laparoscopic roux en y gastric bypass for morbid obesity. History of diabetes was predictive of increased length of stay while increased BMI was associated with decreased excess weight loss. These factors may be important when counselling patients considering bariatric surgery.

Session Number: Poster – Poster Presentations
Program Number: P488
View Poster

« Return to SAGES 2012 abstract archive