Impact of a Clinical Decision Support System on Computed Tomographic Pulmonary Angiogram Yield and Utilization in Hospitalized Patients.

Abstract

Objective: To determine whether point-of-order clinical decision support (CDS) based on the Wells Criteria improves CT pulmonary angiogram (CTPA) yield and utilization in hospitalized patients in an enterprise-wide health system and identify yield-related factors.

Methods: This retrospective IRB-approved cross-sectional study in an urban, multi-institution health system included hospitalized patients undergoing CTPA 12 months before and after CDS implementation (entire cohort). Chi-square test was used to compare PE yield in patients in whom providers overrode vs. followed CDS alerts after CDS implementation. It was also used to compare utilization and yield pre- vs. post-intervention. Univariate and multivariable regression analyses were performed on patient factors and post-CDS Wells scores to evaluate yield-related factors.

Results: For 2,429 inpatient CTPAs post-intervention, CTPA yield was significantly higher when CDS recommendations were followed (18.3% [250/1365]) compared to those overridden (14.2% [151/1064] (p<0.01). For 5,372 CTPAs in the entire cohort, there was no difference in PE yield before (448/2,943; 15.2%) vs. after (401/2,429; 16.5%) CDS implementation (p=0.20). However, in 340,146 admissions over the study period, a 7.4% relative decrease in CTPA utilization (17.5 to 16.2 CTPA examinations/1000 admissions before and after CDS, respectively; p=0.003) was observed.

Discussion: When CDS recommendations were followed, the yield of CTPA was significantly higher than when clinicians overrode CDS alerts. In addition, point-of-order CDS to reduce unnecessary CTPA in hospitalized patients resulted in a significant decrease in CTPA utilization post-CDS implementation with a modest though nonsignificant increase in CTPA yield.