(301) Glycemic Variability Within 24 Hours of ICU Admission as a Predictor of ICU Mortality: Cohort Study

Introduction: Glucose fluctuations in critically ill patients are associated with oxidative stress, endothelial dysfunction, and activation of pro-inflammatory signalling pathways. Glycemic variability (GV) measures these fluctuations, and some observational evidence suggests its usefulness as a predictor of ICU mortality. GV may help identify critically ill patients for whom a specific targeted glycemic control could be beneficial.

Methods: We adapted the STROBE guidelines to develop our conference abstract. We conducted an ambispective cohort study to assess GV within the first 24 hours of ICU admission as a predictor of ICU mortality. The study occurred from March 2022 to January 2024 in the polyvalent ICU at Hospital Civil de Guadalajara “Fray Antonio Alcalde.” Using our Electronic Health Record (EHR) system, researchers collected and verified data on patient age, gender, comorbidities, severity scores, nosocomial infections, length of mechanical ventilation, hospital stay, and ICU mortality. We calculated GV by measuring the standard deviation of capillary glucose records and applied this to prospective data. Regression analysis explored the relationship between GV and ICU mortality. Our local ethics committee approved the study.

Results: We included 266 patients (182 retrospective, 84 prospective) with mean ages of 36 and 32 years, respectively. We defined the GV cut-off at a 30% difference. In the retrospective phase, ICU mortality was 15.4% (28 patients), with GV >30% in 34% (62 patients); mortality was 3.3% (4 patients) for GV < 30% and 38.7% (24 patients) for GV >30%, yielding an OR of 18.3 (95% CI 5.9–56.1, p=0.0001). In the prospective phase, ICU mortality was 14% (12 patients), with GV >30% in 64% (55 patients); mortality was 3.2% (1 patient) for GV < 30% and 20% (11 patients) for GV >30%, yielding an OR of 7.51 (95% CI 0.91–62, p=0.049).

Conclusions: Our study found that GV within the first 24 hours of ICU admission predicts ICU mortality, with a GV >30% linked to significantly higher mortality rates in both retrospective and prospective analyses, confirming that GV is a predictor for mortality in critically ill patients, and may be an interesting target for glycemic control in critically ill patients.