Association Between Time in Targeted Blood Glucose Range With All-Cause Mortality in Critically Ill Patients With...

Heart failure (HF) patients in ICU experience high mortality rates, often complicated by glycemic dysregulation. Time in targeted blood glucose range (TTBR) provides a comprehensive assessment of glycemic control, but its association with outcomes in critically ill HF patients remains unclear.

This retrospective cohort study analyzed 5396 critically ill HF patients from the MIMIC-IV database. TTBR, defined as the percentage of time that glucose values remain between 70-139 mg/dL, was calculated using linear interpolation. Cox proportional hazards regression models assessed the association between TTBR and mortality. Restricted cubic splines examined dose-response relationships. Receiver operating characteristic curve analysis evaluated the predictive ability of TTBR for mortality. Subgroup analyses were performed across demographic and clinical categories.

A total of 5396 patients were included in the analysis with a mean age of 71.77. Mortality rates decreased progressively across TTBR quartiles (Q1: 20.83%, Q2: 15.56%, Q3: 11.06%, Q4: 8.9%; P < .001). In model 3, TTBR as a continuous variable was significantly associated with reduced mortality (hazard ratio [HR] 0.98, 95% CI 0.98-0.99, P < .001). Compared to Q1, patients in Q3 (HR 0.64, 95% CI 0.51-0.79, P = .004) and Q4 (HR 0.58, 95% CI 0.47-0.72, P = .003) had significantly lower mortality risk. The model fully adjusted for demographics, comorbidities, and acute clinical variables achieved an area under the curve of 0.77 (95% CI: 0.73-0.81), demonstrating good discriminative ability. The protective effect of higher TTBR was consistent across most subgroups.

Higher TTBR is independently associated with lower mortality in critically ill HF patients, supporting its potential role as a prognostic indicator of glycemic control in this setting.