Disease-Specific Optimal Blood Glucose Control Strategies in Critically Ill Patients: Retrospective Cohort Study and Network Meta-analysis.
Xinyao He, Jiahao Meng, Yiqun Zhu et al.
✦ AI-curated · Sources linked
In 30 seconds
This study combined a retrospective MIMIC-IV cohort analysis with a frequentist network meta-analysis of randomized trials to examine glucose-control targets in critically ill patients with sepsis, cardiac disease, or neurological disorders. In the cohort, intermediate strict control (110-150 mg/dL) was associated with the lowest 28-day mortality in sepsis and cardiac patients, while strict control (80-110 mg/dL) was associated with lower mortality in neurological patients (HR 0.76, p = 0.013). The network meta-analysis broadly supported disease-specific differences, but found no statistically significant pairwise differences and ranked liberal control highest for cardiac disease.
Key findings
- In sepsis and cardiac patients, strict and liberal control were associated with higher 28-day mortality than intermediate strict control (110-150 mg/dL).
- Among neurological patients, strict control was associated with lower mortality (HR 0.76, p = 0.013).
- The network meta-analysis ranked intermediate strict control highest for sepsis (P-score 74.9%) and strict control highest for neurological disorders (P-score 71.2%).
- For cardiac disease, liberal control ranked highest in the network meta-analysis (P-score 83.5%), but no pairwise differences were statistically significant.
Why it matters
The findings suggest that glucose targets associated with better outcomes may differ across critical illness categories, challenging the assumption that one target suits all ICU patients. However, the differing cardiac results between the cohort and network meta-analysis, alongside the absence of significant pairwise differences in the latter, mean the evidence does not establish a definitive target for every condition.
What to keep in mind
The cohort analysis was retrospective, and the network meta-analysis found no statistically significant pairwise differences; the optimal target for cardiac disease therefore remains uncertain.