Embedded Clinical Decision Support in Electronic Health Record Decreases Use of High-cost Imaging in the Emergency Department: EmbED study

Abstract

Objective: The objective was to evaluate the impact of evidence-based clinical decision support tools integrated directly into provider workflow in the electronic health record on utilization of computed tomography (CT) brain, C-spine, and pulmonary embolism (PE). Methods: Validated, well-accepted scoring tools for head injury, C-spine injury, and PE were embedded into the electronic health record in a manner minimally disruptive to provider workflow. This was a longitudinal, before/after study in five emergency departments (EDs) in a healthcare system with a common electronic health record. Attending ED physicians practicing during the entire study period were included. The main outcome measure was proportion of CTs ordered by provider (total number of CT scans of a given type divided by total patients seen by that provider) in aggregate in the pre- and post intervention period. Results: There were 235,858 total patient visits analyzed in this study with an absolute decrease of 6,106 CT scan ordering for the three studies. Across all sites, there was greater than 6% decrease in utilization of CT brain and CT C-spine (–10%, 95% CI = –13% to –7%, p < 0.001; and –6%, 95% CI =–11% to –1%, p = 0.03, respectively). The use of CT PE also decreased but was not significant (–2%, 95% CI = –9% to +5%, p = 0.42). For all CT types, high utilizers in the pre-intervention period decreased usage over 14% in the post-intervention period with CT brain (–18%, 95% CI = –22% to –15%, p < 0.001), CT C-spine (–14%, 95% CI = –20% to –8%, p = 0.001), and CT PE (–23%, 95% CI = –31% to –14%, p < 0.001). For all three studies, the average utilizers did not change their usage practices. For CT brain, the low utilizers also did not increase usage but for CT C-spine and CT PE usage was increased (+29%, 95% CI = 10% to 52%, p = 0.003; and +46%, 95% CI = 26% to 70%, p < 0.001, respectively). Conclusion: Embedded clinical decision support is associated with decreased overall utilization of high-cost imaging, especially among higher utilizers. It also affected low utilizers, increasing their usage consistent with improved adherence to guidelines, but this effect did not offset the overall decreased utilization for CT brain or CT C-spine. Thus, integrating clinical decision support into the provider workflow promotes usage of validated tools across providers, which can standardize the delivery of care and improve compliance with evidence-based guidelines.