De-Implementation of Low Value Imaging in Pulmonary Embolism (DELVE PE)

1 PROJECT SUMMARY 2 Two percent (2%) of all 120 million annual emergency department (ED) visits in the United States undergo computed 3 tomographic pulmonary angiography (CTPA) for pulmonary embolism (PE) every year. However, many of these are

4 avoidable. Unnecessary testing subjects patients to additional costs, ionizing radiation, and risk of anaphylaxis from contrast 5 media. Further, because false positive CTPAs are common (estimated at 5-26%), overdiagnosis and overtreatment are 6 growing problems in the United States. Despite guidelines recommending the reduction of low-value imaging for PE, the

7 volume of unnecessary imaging studies for PE in US EDs continues to rise, exposing patients to harm from overtesting and

8 overdiagnosis. Prior strategies have attempted to close this knowledge-practice quality gap but only had modest success,

9 possibly because changing provider behavior is challenging and most initiatives focused on isolated strategies. We propose

10 to develop, pilot, and evaluate a multi-dimensional `enhanced' audit-feedback strategy (EAF) to de-implement low-value

11 imaging in PE. The core strategy, audit-feedback, was chosen to target drivers of low-value imaging in PE discovered in

12 our foundational work: knowledge, peer pressure, emotion, and belief about consequences. We designed the strategy using 13 a novel audit-feedback theory, the Clinical Performance Feedback Intervention Theory (CP-FIT). The audit-feedback is

14 `enhanced' by complementary strategies to assist in action planning and empower clinicians to change their behavior. This

15 includes an aid for appropriate testing that guides clinicians through an unambiguous algorithm for evaluation of PE, an

16 educational podcast, and local champions. In Aim 1, we will develop, refine, and operationalize the de-implementation 17 strategy components. We will conduct cognitive testing of the prototypes of the audit-feedback reports, develop an

18 electronic health record (EHR)-integrated aid, record the educational podcast, and identify and train local champions. In

19 Aim 2, we will pilot alternative versions of the EAF strategy in 5 EDs to evaluate acceptability and appropriateness. We

20 will alter the delivery of the components of feedback (local champion versus departmental chair), educational podcast (with

21 or without a real patient narrative), and the aid for appropriate testing (static versus EHR-integrated aid) to ascertain the

22 acceptability, appropriateness, and added value of these iterations. We will pilot different versions of the strategy to evaluate

23 aspects of CP-FIT including complexity, social influence, and actionability, which CP-FIT suggests drive the audit-feedback

24 cycle and, ultimately, behavior change. We will use mixed methods to evaluate the strategy and select a refined strategy for

25 a multi-center trial. In this K23 application, we have proposed a detailed career development plan in which I will gain

26 methodological and technical expertise in advanced implementation science strategies, methods, and trial design. I am well

27 supported by an experienced team of mentors and advisors. At the end of this mentored career development award, I will

28 have a robust empirically-derived and theory-informed strategy and will be positioned to test this strategy in a multi-center 29 trial as an independent investigator. 30