Our aim was to construct predictive models to estimate the risk of long-term end-organ complications and death in patients with type 2 diabetes (T2DM) and obesity who are considering bariatric surgery.


On a cohort of 288,692 patients with T2DM in the Cleveland Clinic Health System between 2004-2017, 2287 patients with obesity (BMI ≥30 kg/m2) who underwent bariatric surgery were matched 1:5 resulting in 11435 control patients based on the index date, age, gender, BMI, site, insulin use, and presence of diabetes complications with follow-up through December 2018. Multivariable time-to-event models were built and internally validated using 5-fold cross-validation to predict the 7-year risk for 5 outcomes of interest.


The prediction tools demonstrated the following discrimination ability based on the cross-validated time-dependent area under the curve (1=perfect discrimination, 0.5=coin flip) averaged over the years 5-7 for the 5 outcomes in the surgical and control groups respectively: all-cause mortality (0.77 and 0.78), coronary artery disease (0.66 and 0.69), cerebrovascular event (0.72 and 0.61), heart failure (0.79 and 0.78), and nephropathy (0.80 and 0.81). The Individualized Diabetes Complications (IDC) Risk Scores were integrated into user-friendly web and smartphone applications for clinical use. When a patient’s data is entered into the application, it calculates the 7-year morbidity and mortality rates with and without undergoing bariatric surgery.


The IDC Risk Scores can provide a strong evidence-based message for patients about their future health outcomes, based on their current status of obesity, T2DM, and metabolic disease with and without having bariatric surgery.