Continuous glucose monitoring in the ICU: clinical considerations and consensus
INTRODUCTION While tight glucose control has been widely adopted in the critical care setting, the optimal target glucose level following acute traumatic brain injury (TBI) remains debatable. This observational study was conducted to delineate the relationship between glucose levels and clinical outcomes during acute phase (first 5 days) of TBI. METHODS We retrospectively identified 429 TBI patients admitted to the intensive care unit (ICU) from January 2005 to December 2006. Of those, 380 patients were retained for final analysis. Collected data included demographics, admission Glasgow Coma Scale (GCS), and APACHE II, glucose on admission and during the first 5 days of admission, and insulin use. Clinical outcomes included mortality, ICU, and hospital length of stay. RESULTS The overall hospital mortality was 13.2% (n = 50). Demographics were similar between survivor and nonsurvivor groups; however, nonsurvivors were older and had worse disease severity on admission. Nonsurvivors also had significantly higher glucose levels at admission and during the first 24 h of admission (P < 0.001). Based on the receiver operating characteristic (ROC) curve, admission and day-1 peak glucose were better predictors for mortality compared to hospital days 2-5 glucose levels, with day-1 peak glucose being the best predictor of mortality (AUC = 0.820). A Kaplan-Meier survival analysis also showed that patients with glucose <160 mg/dl during the first day of ICU admission had a significantly better survival rate compared to those with glucose > or =160 mg/dl (P < 0.001). Two glucose bands, <60 and > or =160 mg/dl, were identified to be associated with increased mortality irrespective of injury severity (OR = 1.130; 95% CI 1.034-1.235; P = 0.007; OR = 1.034; 95% CI 1.021-1.047, P < 0.001; respectively). CONCLUSIONS Findings from our study suggest a glucose level > or =160 mg/dl within the first 24 h of admission following TBI is associated with poor outcomes irrespective of severity of injury, and this presents a timeframe for which active therapeutic interventions may improve clinical outcomes. Prospective efficacy trials are needed to corroborate these findings.