<![CDATA[BACKGROUNDAdmission blood glucose is routinely used to assess diabetic patients in the emergency department (ED), but it cannot distinguish chronic from acute stress-induced hyperglycemia. The stress hyperglycemia ratio (SHR), derived by dividing admission glucose by the HbA1c-estimated average glucose, adjusts for the patient's chronic glycemic baseline and may more accurately capture acute metabolic stress. This study aimed to determine the discriminative performance of SHR versus admission blood glucose in predicting mortality among type 2 diabetes patients. METHODSA retrospective cohort study was conducted involving 200 adults with type 2 diabetes and admission glucose ≥200 mg/dL. The primary outcome was mortality. Secondary outcomes were ICU/HCU admission, composite adverse outcome, and prolonged hospital length of stay. Multivariable logistic regression and ROC curve analyses were performed to evaluate the independent associations and compare the discriminative ability of SHR and admission blood glucose using AUC. RESULTSAmong 200 participants (mean age 57.6±10.2 years; mean SHR 1.5±0.59), 41(20.5%) died and 50(25.0%) experienced composite adverse outcomes. SHR was independently associated with mortality (AOR 3.25; 95%CI 1.44–7.32; p=0.004) and composite adverse outcome (AOR 3.21; 95%CI 1.45–7.10; p=0.004), while admission glucose was not significantly associated with any outcome. SHR outperformed admission glucose in discriminating mortality (AUC 0.734 vs 0.567) and composite adverse outcome (AUC 0.714 vs 0.542). Neither predictor significantly discriminated ICU/HCU admission or prolonged LOS. CONCLUSIONSHR outperforms admission blood glucose in predicting mortality and adverse outcomes in hyperglycemic ED patients with type 2 diabetes. Routine SHR calculation at ED admission may support early risk stratification in high-risk populations.]]>
