The evaluation of risk prediction models in predicting outcomes after bariatric surgery: a prospective observational...

As the prevalence of obesity is increasing, the number of patients requiring surgical intervention for obesity-related illness is also rising. The aim of this pilot study was to explore predictors of short-term morbidity and longer-term poor weight loss after bariatric surgery.

This was a single-centre prospective observational cohort pilot study in patients undergoing bariatric surgery. We assessed the accuracy (discrimination and calibration) of two previously validated risk prediction models (the Physiological and Operative Severity Score for the enumeration of Morbidity and Mortality, POSSUM score, and the Obesity Surgical Mortality Risk Score, OS-MS) for postoperative outcome (postoperative morbidity defined using the Post Operative Morbidity Survey). We then tested the relationship between postoperative morbidity and longer-term weight loss outcome adjusting for known patient risk factors.

Complete data were collected on 197 patients who underwent surgery for obesity or obesity-related illnesses between March 2010 and September 2013. Results showed POSSUM and OS-MRS were less accurate at predicting Post Operative Morbidity Survey (POMS)-defined morbidity on day 3 than defining prolonged length of stay due to poor mobility and/or POMS-defined morbidity. Having fewer than 28 days alive and out of hospital within 30 days of surgery was predictive of poor weight loss at 1 year, independent of POSSUM-defined risk (odds ratio 2.6; 95% confidence interval 1.28-5.24).

POSSUM may be used to predict patients who will have prolonged postoperative LOS after bariatric surgery due to morbidity or poor mobility. However, independent of POSSUM score, having less than 28 days alive and out of hospital predicted poor weight loss outcome at 1 year. This adds to the literature that postoperative complications are independently associated with poor longer-term surgical outcomes.