Glycemic Control in Severe Acute Brain Injury: Systematic Review and Meta-analysis with Trial Sequential Analysis

Hyperglycemia is common in intensive care unit (ICU) patients, including those with acute brain injury. Although many studies associate hyperglycemia with increased mortality, the optimal blood glucose target remains uncertain. This systematic review compares strict versus liberal glycemic control in ICU patients with acute brain injury, focusing on all-cause mortality, neurological outcome, and health-related quality of life.

Two authors independently searched CENTRAL, EMBASE, Medline, Web of Science, CINAHL, and trial registers for studies comparing glycemic control targets. Data were extracted in duplicate with discrepancies resolved by two screeners using Covidence. We performed meta-analysis, trial sequential analysis (TSA), and risk-of-bias assessment, and evaluated certainty of evidence using Grading of Recommendations Assessment, Development and Evaluation (GRADE).

From 29,221 records, we included 18 randomized clinical trials (RCTs) (N = 3196 patients) and 13 observational studies (N = 8354 patients). Fixed-effect meta-analysis suggested no effect on mortality from strict compared with liberal control (N = 3196 patients; RR 0.95; 95% CI 0.84-1.08; I² 0%; very low level of evidence), with TSA indicating a need for additional trials. Fixed-effect meta-analysis showed no clear effect on functional outcome (N = 1239 patients; RR 0.91; 95% CI 0.82-1.00; I² 0%; low level of evidence), and TSA indicated evidence against an effect. No studies reported health-related quality of life.

Low to very low certainty evidence suggests there is no difference between strict and liberal glycemic control regarding mortality or functional neurological outcome after severe acute brain injury. While additional large, low-bias RCTs could improve certainty, the benefit from strict glycemic control remains uncertain.