Reduced Computed Tomography Use in the Emergency Department Evaluation of Headache Was Not Followed by Increased Death or Missed Diagnosis.

introductionThis study investigated whether a 9.6% decrease in the use of head computed tomography (HCT) for patients presenting to the emergency department (ED) with a chief complaint of headache was followed by an increase in proportions of death or missed intracranial diagnosis during the 22.5-month period following each index ED visit.

methodsWe reviewed the electronic medical records of all patients sampled during a quality improvement effort in which the aforementioned decrease in HCT use had been observed. We reviewed notes from the ED, neurology, neurosurgery, and primary care services, as well as all brain imaging results to determine if death occurred or if an intracranial condition was discovered in the 22.5 months after each index ED visit. An independent, blinded reviewer reviewed each case where an intracranial condition was diagnosed after ED discharge to determine whether the condition was reasonably likely to have been related to the index ED visit's presentation, thereby representing a missed diagnosis.

resultsOf the 582 separate index ED visits sampled, we observed a total of nine deaths and 10 missed intracranial diagnoses. There was no difference in the proportion of death (p = 0.337) or missed intracranial diagnosis (p = 0.312) observed after a 9.6% reduction in HCT use. Among patients who subsequently had visits for headache or brain imaging, we found that these patients were significantly more likely to have not had a HCT done during the index ED visit (59.2% vs. 49.6% (p = 0.031) and 37.1% vs. 26% (p = 0.006), respectively).

conclusionOur study adds to the compelling evidence that there is opportunity to safely decrease CT imaging for ED patients. To determine the cost effectiveness of such reductions further research is needed to measure what patients and their healthcare providers do after discharge from the ED when unnecessary testing is withheld.