Back to 2025 Abstracts
THE DIABETES REMISSION INDEX (DRI): AN INNOVATIVE PREDICTIVE MODEL FOR DIABETES REMISSION AFTER METABOLIC PROCEDURES
Wissam Ghusn
*1,3, Pearl Ma
2, Robert A. Vierkant
3, Manpreet Mundi
3, Matyas Fehervari
4, Kayla Ikemiya
2, Karl Hage
5, Andres Acosta
3, Michael Camilleri
3, Barham Abu Dayyeh
3, Kelvin Higa
2, Omar M. Ghanem
31Internal Medicine, Boston University, Boston, MA; 2University of California San Francisco Department of Medicine Fresno, Fresno, CA; 3Mayo Foundation for Medical Education and Research, Rochester, MN; 4Imperial College London, London, United Kingdom; 5Cleveland Clinic, Cleveland, OH
Introduction: Type 2 diabetes (T2D) represents a significant public health burden, with metabolic and bariatric surgeries (MBS), such as Roux-en-Y gastric bypass (RYGB) and sleeve gastrectomy (SG), offering effective treatment options. While these procedures are associated with high rates of diabetes remission, outcomes remain variable. This study aimed to develop two novel prognostic models, the Diabetes Remission Index (DRI) and the Weight Loss-Adjusted Diabetes Remission Index (W-DRI), to predict T2D remission after MBS.
Methods: This multicenter, retrospective cohort study evaluated patients with T2D and obesity who underwent RYGB or SG between 2008 and 2018. Inclusion criteria required a diagnosis of T2D and complete follow-up data, while patients with conditions affecting weight or glucose metabolism were excluded. The DRI model was developed using preoperative variables, while the W-DRI incorporated postoperative weight loss. Predictive accuracy was assessed using the area under the receiver operating characteristic curve (AUC), calibration plots, and stratified analysis.
Results: The study included 503 patients from Institution 1 (I-1) and 409 from Institution 2 (I-2) (
Table 1). The DRI model incorporated preoperative variables, including T2D duration, HbA1c, insulin use, number of diabetes medications, and presence of vascular complications. The W-DRI further integrated postoperative weight loss as a variable. In I-1, 44.7% of patients achieved T2D remission, with a DRI AUC of 0.80 (95% CI: 0.77–0.83;
Figure 1A). External validation in I-2 demonstrated remission in 52.6% of patients and a comparable AUC of 0.78 (95% CI: 0.75–0.81;
Figure 1B). Incorporating weight loss into the W-DRI improved predictive performance, with AUCs of 0.82 (95% CI: 0.79–0.85;
Figure 1C) in I-1 and 0.79 (95% CI: 0.76–0.82;
Figure 1D) in I-2. Patients achieving ?25% total body weight loss had significantly higher remission rates. Calibration analysis showed strong alignment between predicted and observed remission rates. An interactive DRI and W-DRI calculator is now available on the Mayo Clinic webpage (
https://newsnetwork.mayoclinic.org/dri-calculator/).
Conclusion: This study developed and validated the DRI and W-DRI models, which demonstrated robust predictive accuracy for T2D remission after MBS. These models provide valuable tools for tailoring patient care and optimizing treatment outcomes. Future studies should expand these models to diverse populations and explore additional predictive variables.
Table 1. Demographic and Clinical Characteristics of the two cohorts
Figure 1. DRI model ROC curves and areas under the curve (AUCs) for the five cross-validation validation sets for I-1 (Panel A) and for I-2 (Panel B), and W-DRI model ROC curves and AUCs for the five cross-validation validation sets for I-1 (Panel C) and for I-2 (Panel D).
Back to 2025 Abstracts