Persistent hyperglycemia is predictive of outcome in critically ill trauma patients

Abstract

Objectives: Our objectives were to determine whether persistent hyperglycemia was predictive of outcome in critically ill trauma patients.

Methods: Prospective data were collected daily on 942 consecutive trauma patients admitted to the ICU over a 2-year period. Patients were stratified by serum glucose level from day 1 to day 7 (low = 0-139 mg/dL, medium = 140-219 mg/dL, and high >220 mg/dL) age, gender, and ISS. Patients were further stratified by pattern of glucose control (all low, all moderate, all high, improving, worsening, highly variable (HV). Outcome was measured by ventilator days, infection, hospital (HLOS) and ICU (ILOS) length of stay and mortality. Multiple linear regression models were used to determine level of significance.

Results: 71% were victims of blunt trauma. The majority (74%) were male with a mean ISS of 21.3 +/- 15. 41% of patients acquired an infection. Patients with medium, high, worsening, and highly variable hyperglycemia were found to have increased ILOS, HLOS, ventilator days, infection rate and mortality by univariate analysis (p < 0.01). When controlling for age, ISS, and glucose pattern, patients with high, worsening and HV hyperglycemia were most predictive of increased ventilator days, ILOS, HLOS, infection and mortality. (p < 0.01).

Conclusion: Trauma patients with persistent hyperglycemia have a significantly greater degree of morbidity and mortality. A prospective randomized controlled study instituting aggressive hyperglycemic control is warranted.