The association of dysglycaemia metrics over the first 7 days of ICU stay with ICU mortality among patients with and without diabetes.

Background: Dysglycaemia metrics, defined as hyperglycaemia, increased glucose variability, hypoglycaemia and reduced time in the targeted blood glucose range (TIR), are linked to higher mortality. The relationship between dysglycaemia metrics and intensive care unit (ICU) mortality over time for patients with and without diabetes remains inconclusive, posing challenges for ICU medical staff in accurately identifying and distinguishing various risk factors and taking timely and appropriate responses.

Aim: To explore which dysglycaemia metrics within the first 7 days of ICU stay are associated with ICU mortality among patients with and without diabetes.

Study design: This retrospective cohort study included 712 patients without diabetes and 222 patients with diabetes. Clinical data were collected within the first 7 days of ICU stay. Binary logistic regression models were built to analyse which dysglycaemia metrics (hyperglycaemia, coefficient of variation [CV], hypoglycaemia and TIR) on the first day, over the first 3, 5 and 7 days of ICU stay were associated with ICU mortality.

Results: In patients with diabetes, hyperglycaemia on the first day (OR: 4.90, 95% CI: 1.51-15.90, p = .008) and TIR <70% during the first 7 days of ICU stay (OR: 16.31, 95% CI: 1.50-176.89, p = .022) were associated with increased ICU mortality. In patients without diabetes, CV >20% on the first day (OR: 1.46, 95% CI: 1.03-2.07, p = .035), and TIR <70% during the first 3 (OR: 2.01, 95% CI: 1.35-2.98, p < .001) and 5 days (OR: 1.66, 95% CI: 1.09-2.54, p = .019) were associated with increased ICU mortality (p < .05). The proportion of hypoglycaemia did not significantly correlate with ICU mortality in patients with or without diabetes (p > .05).

Conclusions: Specific dysglycaemia metrics are associated with ICU mortality between patients with and without diabetes. In patients with diabetes, hyperglycaemia on the first day and TIR <70% on the first 7 days with higher mortality. In patients without diabetes, CV >20% on the first day and TIR <70% in the first 3 and 5 days are associated with higher mortality. Monitoring these metrics may potentially help develop strategies to decrease ICU mortality through individualized glycaemic management.

Relevance to clinical practice: Close monitoring of dysglycaemia metrics, especially TIR, and personalized glucose management based on diabetic status may help identify high-risk ICU patients and improve targeted care strategies.