Objective: To evaluate the usefulness of toxicologic studies on the management of children with suspected ingestions.
Design: Prospective, consecutive case series.
Setting: Two tertiary care children's hospital emergency departments.
Patients: All children < or =18 years of age presenting with a suspected ingestion.
Study design: Pediatric emergency physicians completed a 14-point questionnaire on each identified patient that included demographics, signs, and symptoms, and if applicable, the extent of drug analysis performed. Pre-test and post-test utility values were determined by the ordering physician using an 11-point scale (0 = least valuable, 10 = most valuable). Physicians also assessed how positive or negative drug analyses affected patient management.
Results: Two hundred twenty patients met study criteria. Median age was 5 years, with males making up 53% of patients. Drug analysis was ordered in 72% (158/220) of cases, with 59% of these tests obtained for a history of ingestion and 27% obtained for altered mental status (AMS). The most common suspected ingestions were acetaminophen and cold preparations. Seventy-eight of 158 (49%) patients had positive toxicology tests, with 17 unsuspected findings. Patient management was affected in 53/158 (34%; 95% CI, 27-41%) cases. Unsuspected findings affecting management were found in only 4/158 (3%; 95% CI, 1-6%) cases. Significant differences in pre-test and post-test utility values occurred for serum assays (mean difference +0.4, P = 0.008), patients presenting with AMS (mean difference -0.8, P = 0.005), and patients having a negative drug test (mean difference -0.5, P = 0.003). Although negative drug analysis gave the physician reassurance in 39/80 (49%; 95% CI, 38-60%) cases, patient management was altered in only 8/80 (10%; 95% CI, 5-18%) cases.
Conclusions: Seventy-two percent of children presenting with suspected drug ingestions had toxicologic analysis performed as part of their evaluation. Analysis was most valuable to physicians when evaluation of overdoses required serum drug levels. Qualitative urine drug screens provided minimal useful information. Unexpected findings on urine drug screening leading to changes in management were uncommon.