Michal R Janik, MD, Tomasz G Rogula, MD, PhD, Rami R Mustafa, Adel Alhaj Saleh, MD, Mujjahid Abbas, MD, Leena Khaitan, MD. University Hospital Cleveland Medical Center / Case Western Reserve University, Cleveland, US
Introduction: At present Laparoscopic Sleeve Gastrectomy (LSG) is fast becoming the most frequently performed bariatric procedure worldwide. Although patients expect 50-60% excess weight loss (EWL) on average, some do not achieve this goal. If there was a way to determine which patients will be successful with weight loss, then many patients may be able to avoid the risk associated with this procedure. The objective of this study is to develop a predictive model for a successful outcome after LSG.
Material and methods: A retrospective analysis of 344 consecutive patients after primary LSG was performed. Patients underwent LSG between January 2011-September 2015. The data set, including 29 patients’ variables, was identified to develop the predictive model. Successful outcome was defined as % EWL >50 at 1 year follow-up. Multiple logistic regression with stepwise variable selection was used to create the model. Independent variables with an association with the outcome in univariate analyses were entered into a model. With backward elimination procedure, clinically important risk factors were kept in the model. Afterward, a forward stepwise selection was also used to find a stable model.
Results: Through this time period, 63% of patients achieved successful outcome. The mean age of patients was 45.8 (±10.6) years and mean BMI was 47.5 (±8.8) kg/m2 at the time of surgery. Twenty-nine variables were analyzed and 4 were included to the final model. These variables are as follows: age, preoperative BMI, gender, and preoperative weight loss (Table1). The model demonstrated a good calibration (Hosmer-Lemeshow goodness-of-fit test, χ2 =3.356, P =0.910) and good discrimination (c-statistic 0.753).
Conclusion: The result revealed the association between successful outcome and following factors: lower age, lower preoperative BMI, male gender, and higher preoperative weight loss. Further studies are warranted to externally validate this risk model in a different population of LSG patients.
Variable | Value/Units | Odds Ratio | 95% Confidential Interval | Estimate | Standard Error |
---|---|---|---|---|---|
Age | Years | 0.938 | 0.902 – 0.974 | -0.0645 | 0.0195 |
Preoperative BMI | Kg/m2 | 0.922 | 0.878 – 0.967 | -0.0816 | 0.0247 |
Preoperative weight loss | ΔBMI | 1.234 | 0.979 – 1.557 | 0.2106 | 0.1184 |
Gender | Female vs. male | 0.177 | 0.053 – 0.585 | -0.8662 | 0.3053 |
Presented at the SAGES 2017 Annual Meeting in Houston, TX.
Abstract ID: 79227
Program Number: P505
Presentation Session: Poster (Non CME)
Presentation Type: Poster