Background: To test the diagnostic utility of the triage serum glucose in differentiating major from minor injuries.
Methods: Prospective database at Kings County Hospital, a Level I trauma center, from August 2005 to August 2008.
Inclusion criteria: trauma patients 13+ years.
Exclusion criteria: diabetes or obvious life-threatening injuries requiring immediate surgery, isolated head trauma, or transferred or dead on arrival. We recorded age, sex, injury mechanism, base deficit (BD), lactate (LAC), and serum glucose and calculated Injury Severity Scores. Major injury: a change in hematocrit >10, blood transfused within 24 hours, or Injury Severity Score >15. Data were reported as mean differences (95% confidence interval [CI]). Groups were compared by Student's t test; receiver operator characteristic curves were compared by Wilcoxon test (two-tailed, [alpha] = 0.05).
Results: One thousand six hundred forty-nine patients with an average age of 35.5 years (13-95 years), 79.5% male, and 50% blunt trauma were studied. Patients with major (n = 278) compared with minor injury (n = 1371) had significantly (p < 0.0001) lower BD and higher LACs (p < 0.0001). Major injury patients had significantly (p < 0.0001) higher serum glucose levels (8.33 mMol/L, 95% CI: 7.94-8.69 mMol/L) compared with patients with minor injuries (6.49 mMol/L, 95% CI: 6.39-6.66 mMol/L). Areas under the curve for glucose (0.73, 95% CI: 0.70-0.76) are similar to BD (0.72, 95% CI: 0.68-0.76) and LAC (0.71, 95% CI: 0.67-0.75).
Conclusions: Serum glucose was as discriminating as BD or LAC in differentiating minor from major injury. An initial glucose >/=11.1 mMol/L had a low sensitivity (15%) but a high specificity (94%) for major injury.