Effect of Computerized Clinical Decision Support on the Use and Yield of CT Pulmonary Angiography in the Emergency Department
This study was performed to assess the impact of an evidence-based clinical decision support (CDS) tool on the use and yield of CT pulmonary angiography (CTPA) studies ordered in the emergency department (ED) for the diagnosis of acute pulmonary embolism. The study took place between October 1, 2003, and September 30, 2009, at a 793-bed quaternary care facility with 60,000 annual ED visits.
Chest CT pulmonary angiography is used to diagnose acute pulmonary embolism. The use of CTPA in the emergency department setting increased precipitously in the 5 years leading up to this study. If CTPA was being overutilized, then a CDS tool could potentially modulate the appropriate use of this test, thereby increasing care quality and decreasing the number of patients unnecessarily exposed to medical radiation. The study was performed to demonstrate whether a CDS tool could decrease the number of patients undergoing CTPA and increase the yield of positive findings.
The research team developed an evidence-based algorithm to assist physicians in the use of CTPA and implemented this decision support tool into the ED radiology computerized physician order entry system. This algorithm used blood levels of D-dimer (elevated, normal, not evaluated) along with patient history and level of clinical suspicion for pulmonary embolism (high, intermediate, or low) to guide physicians in ordering a CTPA. Along with clinical guidance, the clinical decision support tool included a link for additional information and a link to give feedback. Data was gathered from the electronic health record using diagnosis codes and billing information. Radiology reports positive for acute pulmonary embolism were identified using natural language processing.
From October 2003 to August 2007, the use of CTPA increased 82% to a maximum of 26.4 examinations per 1000 patients. In the 2 years following the CDS intervention, use of CTPA decreased to 21 examinations per 1000 patients. Preintervention, the yield of positive studies was 10% overall. At the beginning of the intervention, the first quarterly yield was 5.8% with a 69% increase in yield to 9.8% at the end of 2 years.
The implementation of the CDS tool decreased CTPA utilization 20% and increased positive yields over the 2 year postintervention period. The increased utilization prior to the intervention was not unique to this institution, as other studies had shown as much as a 5-fold increase in the utilization of CTPA in the ED setting during the same time period. This particular study did not have a control group, so it is difficult to say with certainty that the statistical differences noted after the intervention were solely due to the CDS. Other reasons suggested included decreased use of x-ray studies due to increased awareness of the risk of medical radiation and the increased use of postsurgical thromboprophylaxis during the same time period. The authors felt their study was successful because they targeted a very specific clinical indication and diagnostic modality, which allowed them to incorporate very specific evidence-based guidelines.
A well-planned, targeted CDS tool based on high quality evidence-based guidelines can achieve the objective of improved quality of care and decreased costs.
The authors felt that their intervention was successful because their rollout strategy utilized a multifactorial approach to changing the facility’s culture and the practitioners’ attitude toward the use of CTPA to rule out pulmonary embolism. Their approach included a multidisciplinary team, physician champions, and an education campaign explaining the evidence behind the recommendations incorporated into the decision support tool.
- Raja, A. S., Ip, I. K., Prevedello, L. M., Sodickson, A. D., Farkas, C., Zane, R. D., … Khorasani, R. (2012). Effect of computerized clinical decision support on the use and yield of CT pulmonary angiography in the emergency department. Radiology, 262(2), 468–74. doi:10.1148/radiol.11110951 Retrieved from http://pubs.rsna.org/doi/abs/10.1148/radiol.11110951