Behavioral Interventions Do Not Cut It: CT Utilization in Emergency Department Patients With Negative Age-Adjusted D-Dimer

D-dimer testing is used in the Emergency Department (ED) to exclude venous thromboembolism (VTE), specifically in patients deemed to have a low or moderate pre-test probability of a VTE diagnosis. Baseline D-dimer values increase with age, further reducing the specificity of this test. Use of an age-adjusted D-dimer (AADD) allows clinicians to safely avoid unnecessary advanced imaging, easing crowding in the ED. This study investigated whether we could increase utilization of AADD by implementing an electronic health record behavioral nudge.

This is a pre-post retrospective observational study of a 9-hospital health system in Pennsylvania. Baseline AADD utilization was collected. Following this, the study team provided education in a weekly Epic email update and reinforced at regional staff meetings the safety and improved sensitivity of the AADD and provided a documentation smart phrase to support its use in the medical record. Cases were identified using the SlicerDicer feature in Epic. CT results were abstracted by a team of trained chart reviewers blinded to the study hypothesis using a standardized abstraction tool.

Pre-intervention, July 2022 to June 2023, comprised 7562 patients older than 50 years who were evaluated for pulmonary embolism (PE) with a D-dimer and/or CT Pulmonary Embolism (CTPE). 3640 patients were evaluated post intervention from October 2023 to March 2024. CTPE was performed in 274 patients (32.3%) with a negative age-adjusted D-dimer that was greater than or equal to the standard D-Dimer cutoff (0.5 µg/mL FEU) in the pre period and 122 (30.1%) patients in the post period (p=0.428). In the overall timespan, AADD had a sensitivity of 96.2% (95% CI 94.4-98.1%, specificity of 14.4% (13.4-15.4%), positive predictive value 9.3% (8.40-10.2%) and negative predictive value 97.7% (96.5-98.8%). Corresponding values for traditional cutoff were 99.0% (98.0-100%), 5.6% (4.9-6.3%), 8.7% (7.9-9.6%), and 98.4% (96.8-100%)

Rates of CT utilization in patients with a positive traditional D-Dimer value but negative AADD did not change after education and deployment of a behavioral nudge. This study serves as a large-scale retrospective clinical validation for AADD. We hope to use these findings to support the direct report of AADD in the laboratory results to see if this improves CTPE utilization.