Emergency department abdominal computed tomography for nontraumatic abdominal pain: optimizing utilization.

OBJECTIVES: To identify predictors of positive computed tomographic (CT) yield and to measure the impact of CT yield on the disposition of patients referred for computed tomography after presenting to an emergency department with nontraumatic abdominal pain.

MATERIALS AND METHODS: Computed tomographic reports, laboratory data, and emergency department and hospital records were retrospectively analyzed in 604 consecutive patients undergoing CT examinations. Computed tomographic yield was correlated to age, gender, leukocyte count, specified precomputed-tomography clinical diagnosis, and patient disposition.

RESULTS: Forty-eight percent of CT scans (298 of 621) had positive results. Computed tomographic results were positive in 76% of children (13 of 17) and 47% of adults (285 of 604) (P < .03) and in 45% of female patients (155 of 343) and 51% of male patients (143 of 278) (P < .2). Fifty-two percent of CT scans (223 of 426) with and 38% (75 of 195) without specified precomputed-tomography clinical diagnoses had positive results (P < .01). Fifty-eight percent of CT scans (161 of 278) with elevated and 40% of CT scans (135 of 336) with normal patient leukocyte counts had positive results (P < .001). Sixty-seven percent of patients (171 of 256) admitted and 35% of patients (127 of 365) discharged had positive CT results (P < .001). Computed tomography revealed unsuspected diagnoses in 27% of patients (165 of 621). Thirteen percent of patients (12 of 93) without any clinical predictors for positive CT yield were admitted after positive CT results. Thirty-eight percent of patients (104 of 273) with clinically suspected diagnoses requiring admission were discharged after negative CT results.

CONCLUSION: Clinical indicators of positive CT yield include pediatric age, leukocytosis, and a specified precomputed-tomography diagnosis. Positive CT results are a predictor for hospital admission. In one quarter of cases, computed tomography identifies clinically unsuspected diagnoses and thereby adds information important for patient management, even after clinical evaluation.