212 The Impact of Physician Risk Tolerance and Attitudes Toward Clinical Decision Rules on Head CT Utilization for Trauma Patients in the Emergency Department

Abstract

Despite the development of well-validated evidence-based clinical decision rules (CDR) designed to reduce variability and utilization of head CT in patients who present with minor head injury, there is concern for increasing and inappropriate use of imaging for these patients. Physician risk tolerance, including fear of malpractice as well as misgivings regarding the use and accuracy of clinical decision rules are frequently cited as contributing to suboptimal utilization of head CT. Our objective was to determine whether physician risk tolerance or attitudes toward CDR were associated with utilization of non-contrast CT of the head in patients with head trauma. Our hypothesis was that neither risk tolerance nor attitudes towards CDR would be associated with utilization. We prospectively surveyed 39 attending emergency physicians at a single urban tertiary level-1 trauma center to assess their risk tolerance using 3 previously validated scales: the Risk Tolerance Scale (RTS), Stress from Uncertainty Scale (SUS) and Malpractice Fear Scale (MFS). Attitudes toward CDR were assessed by modifying a previously validated survey designed to measure physician attitudes toward clinical practice guidelines. Physician demographic characteristics, including age, sex, years in practice, clinical hours worked, 3 or 4-year residency training, and prior lawsuit involvement, were also captured. We reviewed all non-contrast head CT scans performed for head injury over the prior 2 years and collected data on patient demographics, triage acuity and time of day at presentation. Using a logistic regression model, the influence of physician and patient factors on the decision to obtain imaging was assessed. Thirty-seven of 39 physicians (95%) completed the survey. A total of 7932 head CT scans for patients with trauma were performed during the study period. Increased head CT utilization was associated with increasing patient age (P<0.0001), higher triage acuity (P<0.0001), male sex (P<0.0001) and time of presentation (overnight greater than evening or day) (P<0.0012). Neither physician demographic and risk tolerance characteristics, nor attitudes toward CDR were significantly associated with utilization. However, the prevalence of CDR use for ordering head CT in patients with minor head injury was low: while 95% of physicians surveyed regularly used CDR when they ordered diagnostic tests, only 35% used them when ordering head CT for patients with minor head injury, and 43% felt CDR for minor head injury were cumbersome and difficult to remember. Utilization of head CT in patients with head injury is associated with patient factors but not physician risk tolerance or attitudes toward CDR, and the prevalence of CDR use for imaging in head injury is low. Future initiatives to reduce head CT utilization in trauma patients should focus on identifying low risk patients in which diagnostic imaging can be safely withheld by educating physicians on CDR and facilitating their use by integration of clinical decision support into electronic order entry software.