Prospective study of the clinical predictors of a positive abdominal computed tomography in blunt trauma patients

J Trauma. 2004 Aug;57(2):296-300. doi: 10.1097/01.ta.0000130612.60661.c3.


Background: CT scans are often used in the evaluation of blunt trauma patients. Many scans are negative. Clinical predictors of positive abdominal CT scans would be beneficial in patient care.

Methods: A prospective study of 213 patients at a Level I trauma center presenting with blunt trauma who underwent abdominal CT scan. Indications for CT scan were analyzed statistically, using univariate and multivariate models.

Results: Univariate chi2 tests showed abnormal pelvis x-ray (p = 0.0002) and an intubated patient (p = 0.03) were predictors of a positive CT scan. When subjected to multivariate logistic regression, these two indications were significant predictors of a positive CT scan, abnormal pelvis x-ray (p = 0.0005, OR=6.6, 95% CI), and an intubated patient (p = 0.02, OR=2.6, 95% CI). Univariate chi2 tests also showed that alcohol intoxication was statistically significant predictor of a negative CT scan (p = 0.03).

Conclusion: Our data suggest that an abnormal pelvis x-ray and intubation are significant risk factors for a positive CT scan. Alcohol intoxication, mechanism of injury, and unreliable examination, without other associated indication for a scan, may warrant further study.

MeSH terms

  • Abdominal Injuries / complications*
  • Accidental Falls
  • Accidents, Traffic
  • Adolescent
  • Adult
  • Aged
  • Aged, 80 and over
  • Alcoholic Intoxication / complications
  • Analysis of Variance
  • Female
  • Hemoperitoneum / diagnostic imaging*
  • Hemoperitoneum / epidemiology
  • Hemoperitoneum / etiology
  • Hospital Mortality
  • Humans
  • Injury Severity Score
  • Intubation, Intratracheal
  • Length of Stay / statistics & numerical data
  • Logistic Models
  • Male
  • Middle Aged
  • Patient Selection
  • Predictive Value of Tests
  • Prospective Studies
  • Radiography, Abdominal / statistics & numerical data*
  • Risk Factors
  • Tomography, X-Ray Computed / statistics & numerical data*
  • Trauma Centers
  • Wounds, Nonpenetrating / complications*