Factors Associated With Imaging of Nontraumatic Headaches in Virginia Emergency Departments

To explore whether patient-reported demographics and hospital classifications were significant factors in the likelihood that a computed tomography (CT) scan of the head would be ordered for nontraumatic headaches.

Nonexperimental, cross-sectional analysis was performed on a database that included 18 279 patients presenting to emergency departments (EDs) in the Commonwealth of Virginia. Information included in the analysis was diagnosis of a generalized, nontraumatic headache; patient age and sex; the treating hospital's profit status (ie, for-profit vs nonprofit); status of the hospital as a teaching institution; hospital location (ie, rural vs suburban and urban); diagnostic procedures ordered; and patient's insurance coverage (commercial insurance, Medicare, or Medicaid). Descriptive statistics, chi-square, and binary logistic regression analyses were performed.

Approximately 10% of patients who presented to EDs with a chief complaint of headache received a diagnostic head CT scan. Patient age proved to be the only significant variable in the analysis (P ≤ .001). Hospital's teaching status, classification as rural or suburban and urban, and profit status were not significant factors in determining the likelihood of a head CT being performed, nor were any other demographic variables studied.

Radiologic procedure expenditures have increased more than any other physician service in the past 2 decades. The increase in CT use warrants further research. Although the analysis did not demonstrate that hospital characteristics were a significant factor in the ordering of head CT scans for nontraumatic headaches, continued analysis should be performed because resource use in EDs is not static.

The overall rate of CT usage in EDs in the United States has increased significantly. Understanding CT use rates is important to radiographers and policymakers when considering resource use in dynamic EDs.