Purpose or Case Report: To analyze the effect of point-of-care ACR-clinical decision support (CDS) software embedded in the electronic health records (EHR) on CT ordering patterns in a quaternary care pediatric emergency center (EC).
Methods & Materials: This study was HIPAA compliant and exempt from IRB approval. All CT exams across the enterprise ordered on EC patients from 9/18/18 to 10/8/2019 were categorized by CDS software (n = 6907). ACR-CDS scores were recorded: Red (1-3; usually not appropriate), Yellow (4-6; may be appropriate), Green (7-9; usually appropriate) and a category of “no score”. From 9/18/18 to 8/7/2019, no feedback was provided to the ordering provider (n = 5833). From 8/8/2019 to 10/8/2019, best practice alerts (BPA) provided feedback to the provider (n = 1074). Any changes in imaging management and order behavior were recorded and compared between the pre-BPA and post-BPA activation. Chi-square test for proportions was conducted.
Results: Differences between the pre- and post-BPA percentages were statistically significant for Red (22% vs 17%), Green (53% vs 64%), and no score studies (11% vs 5%), respectively (p<0.0001). Yellow (15% vs 14%) studies showed no statistically significant difference (p = .841). For Red scores, 8% (n = 15) of studies were cancelled after receiving a BPA. No studies were changed to a study deemed more appropriate by the BPA. Reasons for bypassing the BPA included: Disagree with appropriateness score (17%) and Consulted with Other Specialist (16%). 29% provided free text reasons for bypassing the BPA. Analysis of free text reasons found that 52% (n = 25) of the orders could have chosen more suitable indications which would have resulted in a non-Red appropriateness score.
Conclusions: ACR-CDS recommendations decrease the ratio of clinically inappropriate to appropriate CT exams ordered by referring physicians. A small percentage of CT exams when CDS feedback stated “not inappropriate” actually led to a cancelled order. This baseline study will serve as a foundation to measure future incremental positive change in CT appropriateness after additional clinical educational outreach and improved EHR CDS user interfaces have been effected.