Early hyperglycemia predicts multiple organ failure and mortality but not infection.

BACKGROUND

Previous studies attempting to characterize the association between early hyperglycemia (EH) and subsequent outcome have been performed without utilization of a strict glycemic control protocol. We sought to characterize the clinical outcomes associated with EH in a cohort of severely injured trauma patients, when a strict glycemic control protocol was used.

METHODS

Data were obtained from a multicenter prospective cohort study evaluating clinical outcomes in blunt injured adults with hemorrhagic shock. Known diabetics and patients with isolated traumatic brain injury were excluded from the analysis. A strict glycemic protocol (target glucose, 80-110 mg/dL) was employed. Cox proportional hazard regression was used to evaluate the effects of EH on multiple organ failure (MOF), nosocomial infection (NI), and mortality, after adjusting for the effects of early death on subsequent infection rates.

RESULTS

Overall mortality, MOF, and NI rates for the entire cohort were 19.6%, 37.5%, and 42.2%, respectively, with a mean Injury Severity Score of 31.6 +/- 14. Cox proportional hazard regression confirmed that EH was independently associated with almost a twofold higher mortality rate and a 30% higher incidence of MOF, but was not an independent risk factor for NI, after controlling for all important confounders. There continued to be no independent association between EH and NI, even when stratified by infection type (pneumonia, catheter-related blood stream infection, or urinary tract infection).

CONCLUSION

These results suggest that EH is a marker of severe physiologic insult after injury, and that strict glycemic control may reduce or prevent the infectious complications previously shown to be associated with hyperglycemia early after injury.