Diagnostic yield of head CT in pediatric emergency department patients with acute psychosis or hallucinations

Background: Children presenting to the emergency department with acute psychosis or hallucinations sometimes undergo a head CT to evaluate for a causative lesion. The diagnostic yield of head CT in this scenario has not been reported. Objective: To determine the yield for head CT in children with acute psychosis or hallucinations. Materials and methods: We retrospectively searched the radiology report database over a 7.5-year period for head CT reports for pediatric emergency department patients using the following keywords: hallucination, psychosis, psychotic or “hearing voices.” All reports were categorized as normal or abnormal, and we reviewed and categorized the abnormal cases. We calculated the 95% confidence interval for abnormal CTs using the method of Clopper and Pearson. Results: We identified 397 pediatric emergency department head CTs. We excluded one non-diagnostic exam. We excluded 34 additional cases (which were all normal) because of clinical indications that might have independently triggered a head CT. Of the remaining 362 cases, 12 reports described abnormalities or variants and we reviewed them individually. Based on consensus review, four were normal, four had congenital malformations, three had encephalomalacia versus demyelination and one demonstrated cortical atrophy. There were no cases with actionable findings such as mass, hemorrhage, infection or hydrocephalus. The 95% confidence interval for a CT demonstrating causative findings was calculated at 0–0.82%. Conclusion: In the absence of concerning factors such as focal neurological deficits, evidence of central nervous system infection, trauma or headache, routine screening head CT might not be warranted in children presenting with acute psychosis or hallucinations.