Impact of an Automated Multimodality Point-of-Order Decision Support Tool on Rates of Appropriate Testing and Clinical Decision Making for Individuals With Suspected Coronary Artery Disease : A Prospective Multicenter Study

ObjectivesThis study sought to evaluate the impact of a multimodality-appropriate use criteria decision support tool (AUC-DST) on rates of appropriate testing and clinical decision making.BackgroundAUC have been developed to guide utilization of noninvasive imaging for individuals with suspected coronary artery disease (CAD). The effect of a point-of-order AUC-DST on rates of appropriateness and clinical decision making has not been examined.MethodsWe performed a prospective multicenter cohort study evaluating physicians who ordered CAD imaging tests for consecutive patients insured by 1 large private payer. During an 8-month study period, each study site was granted exemption from prior authorization requirements by radiology benefits managers. An AUC-DST was employed to determine appropriateness ratings for myocardial perfusion scintigraphy (MPS), stress echocardiography (STE), or coronary computed tomographic angiography (CCTA), as well as intended downstream testing and therapy.ResultsOne hundred physicians used the AUC-DST for 472 patients (age 55.6 ± 9.6 years, 61% male, 52% prior known CAD) over 8 months for MPS (72%), STE (24%), and CCTA (5%). The AUC-DST required an average of 137 ± 360 s to determine the appropriateness category that, by American College of Cardiology AUC, was considered appropriate in 241 (51%), uncertain in 96 (20%), inappropriate in 85 (18%), and not addressed in 50 (11%). For tests ordered in the first 2 months compared with the last 2 months, appropriate tests increased from 49% to 61% (p = 0.02), whereas inappropriate tests decreased from 22% to 6% (p < 0.001). During this period, intended changes in medical therapy increased from 11% to 32% (p = 0.001).ConclusionsA point-of-order AUC-DST enabled rapid determination of test appropriateness for CAD evaluation and was associated with increased and decreased testing for appropriate and inappropriate indications, respectively. These changes in test ordering were associated with greater intended changes in post-test medical therapy.