Correlating actual one-year weight loss with predicted weight loss by the MBSAQIP: bariatric surgical risk/benefit calculator.

Background: The American Society for Metabolic and Bariatric Surgery has released a Bariatric Surgical Risk/Benefit Calculator, an online tool with which patients and providers can input patient preoperative information and predict their 1-year weight loss. We seek to validate our institutional data with the national database and investigated patient factors that influence lack of treatment effect after bariatric surgery.

Materials And Methods: A retrospective review of all prospectively collected data of bariatric surgeries performed at Yale New Haven Hospital from 2017 to 2018 was conducted. By entering data into the MBSAQIP Calculator, the 1-year predicted Body Mass Index was calculated and compared to the actual weight loss. Statistical analysis was performed using an unpaired t-test with Welch's correction (Prism 8, GraphPad).

Results: The average difference between the actual and predicted weight loss at 1-year for 327 patients was 3.6 BMI points. When the actual weight loss was compared to predicted BMI at 1 year, a high correlation was found (R = 0.6, P = 0.003). We examined the outliers with a comparison of weight loss for those patients who's BMI fell within 5 points of the predicted versus those whose BMI recorded above 5. It was discovered those patients who had higher than 5 BMI points than predicted, had higher preoperative BMI (46.1 vs 43.6, P = 0.008).

Conclusions: The MBSAQIP calculator is a useful tool to guide surgeons with decision-making and informed consent. Our institution's 1-year weight loss data correlated closely with that predicted. From the outliers, we found that patients who did not meet the predicted weight loss had significantly higher preoperative BMI. This may alter preoperative discussions with class 3 or over obese patients regarding expected weight loss and warrant investigations with the national database to develop modifications of the calculator.