1362-P: Mortality Risk after Hypoglycaemia in Acute Settings in People with Diabetes: Development of a Prognostic Model

Introduction and Objective: In-hospital hypoglycemia increases the risk of mortality, but there is a lack of risk-stratification tools to identify those at risk of hypoglycemia-related adverse events. We aimed to develop a prognostic model to estimate mortality risk within 24 hours of resolution of in-hospital hypoglycemia in people living with diabetes (PLwD). Methods: All in-hospital hypoglycemic episodes (glucose <3.9mmol/l) in PLwD between January 2023 and December 2024 were extracted from a multi-centre (11 hospitals) NHS trust dataset in England. A derivation model was developed using a multiple logistic regression, with collinearity, performance and calibration being assessed. Discrimination was assessed using the Area Under the Receiver Operating Characteristic Curve (AUROC), and the magnitude of individual predictors was assessed using odds ratios (OR). Internal validation was performed using bootstrapping. Results: 2160 hypoglycaemic episodes were included in the dataset. 7.0% were associated with mortality within 24 hours of resolution. Type 2 diabetes (OR 2.1 compared to type 1 diabetes), Charlson Comorbidity Index score (OR 1.1 per unit score increase), IHSG level 2 or 3 (OR 4.4 Level 2 and OR 10.1 Level 3 compared to Level 1) and precipitating factors (OR 2.0 for fasting/missing a meal and OR 2.1 for multiple factors vs unclear cause) were included in the final model. The model demonstrated moderate discrimination (AUROC: 0.74), good calibration, and high stability in internal validation (AUROC 0.73 after optimisation). Conclusion: We successfully developed and internally validated a prognostic model to predict mortality within 24 hours of hypoglycaemia in PLwD in acute settings. This tool could support early risk stratification and clinical decision-making in acute settings. Disclosure S. Sherratt-Mayhew: None. A. Ling Jie Yee: None. C. Page: None. A. Buchipudi: None. K. Persad: None. A. Khatoon: None. M. Idrissi: None. S. Abdi Osman: None. A. Manta: None. A. Iqbal: Research Support; Current; Abbott Diabetes, Dexcom, Inc. Advisory Panel; Current; Sanofi. Speaker's Bureau; Current; Sanofi, Daiichi Sankyo, Eli Lilly and Company. P. Choudhary: Advisory Panel; Current; Medtronic. Speaker's Bureau; Current; Medtronic. Research Support; Current; Medtronic. Consultant; Current; Medtronic. Speaker's Bureau; Current; Abbott. Advisory Panel; Current; Abbott. Research Support; Current; Abbott. Speaker's Bureau; Current; Dexcom, Inc. Advisory Panel; Current; Dexcom, Inc. Research Support; Current; Dexcom, Inc. Consultant; Current; Dexcom, Inc. Speaker's Bureau; Current; Insulet Corporation. Advisory Panel; Current; Insulet Corporation. Research Support; Current; Insulet Corporation. Consultant; Current; Insulet Corporation. Speaker's Bureau; Current; Sanofi. Consultant; Current; Sanofi. Speaker's Bureau; Current; Lilly. Advisory Panel; Current; Lilly. Consultant; Current; Lilly. Advisory Panel; Current; Ypsomed AG. Advisory Panel; Ended; Embecta. Speaker's Bureau; Current; Roche Diabetes Care. Research Support; Current; Roche Diabetes Care. Consultant; Current; Roche Diabetes Care. Advisory Panel; Current; Vertex Pharmaceuticals Incorporated. Consultant; Current; Glooko, Inc., Cambridge Mechatronics Ltd, vTv Therapeutics. M.A. Karamat: None. R.P. Raghavan: Speaker's Bureau; Current; Eli Lilly and Company. Other - Chaired Healthcare educational meetings sponsored by Abbott; Current; Abbott Diabetes. P. Kempegowda: None. Funding This work was not directly funded. Grants from NIHR and ABCD supported author PK.