Clinical trial design data for electrocardiogram artificial intelligence-guided screening for low ejection fraction (EAGLE)

The article details the materials that will be used in a clinical trial - ECG AI-Guided Screening for Low Ejection Fraction (EAGLE): Rationale and design of a pragmatic cluster randomized trial [1]. It includes a clinician-facing action recommendation report that will translate an artificial intelligence algorithm to routine practice and an alert when a positive screening result is found. This report was developed using a user-centered approach via an iterative process with input from multiple physician groups. Such data can be reused and adapted to translate other artificial intelligence algorithms. This article also includes data collection forms we developed for the clinical trial aiming to evaluate the artificial intelligence algorithm. Such materials can be adapted for other clinical trials.


Data
includes a clinician-facing action recommendation report with two versions e one for a negative result which requires no action, and the other for a positive result, which suggests ordering an echocardiogram. Fig. 2 is a sample email alert to clinicians when a positive screening result is detected. Fig. 3 is the baseline survey that will be administered to clinicians at the time of enrolment. Fig. 4 is the end-of-study survey that will be administered to clinicians in the intervention group at the end of the trial [1].

Experimental design, materials, and methods
The clinician-facing action recommendation report was developed over a period of four months (December 2018eMarch 2019). A multi-disciplinary team developed a prototype of the report using a user-centered iterative approach. The principal investigators of the project (a health services researcher and a cardiologist) drafted an initial prototype. The investigative team then identified major groups of clinicians who frequently order ECG (i.e., those in primary care, cardiology, emergency medicine, and anesthesiology) and introduced the tool to the leadership of these departments during face-to-face meetings. At these stakeholder meetings, the investigative team got a better understanding of their needs and solicited feedback on the new tool and the design of the report. The investigative team also asked the department leaders to suggest 3e5 practicing clinicians in each department to participate in the subsequent testing and refinement of the prototype. Two designers worked with practicing clinicians to conduct interviews and workflow observations. A series of Specifications Table   Subject Cardiology and Cardiovascular Medicine Specific subject area Heart failure Type of data Figure How data were acquired The data were obtained via the discussion within the investigative team and interviews with clinicians from a variety of specialties. The data were created by the investigators using simple software like Word and pdf. Value of the Data These data provide an example of how an artificial intelligence algorithm can be translated to practice and how to design a clinical trial to evaluate the value of the algorithm in routine clinical practice. Clinicians and researchers who are working on translating artificial intelligence algorithms to routine practice and who are designing clinical trials. Clinicians and researchers can use these materials as a start point and adapt to their own projects. Dear XX -Thank you for participating in the EAGLE trial to test the ECG AI-guided screening for low ejection fraction. This is an automated report to support your use of this new tool.
One of your patients (Patient's Name Clinic Number) had a positive screening result for reduced ejection fraction. Please review the patient's record and consider ordering an echocardiogram unless you feel an echocardiogram would be of low value within the patient's clinical context (e.g., a recent test showing normal EF or finding a low EF would not change management). Information on how to access the screening report and bill for the test can be found here (link to the FAQ hosted on mayo intranet).
If you decide not to order an echocardiogram, please let us know the reason here. This information will greatly help us refine the tool and inform future implementation efforts. I think that it is not likely that this patient has a low EF, so the potential cost/inconvenience/risk of echocardiogram is not justified · TTE ordered at an outside facility · Other, please specify_________

Survey Instructions
We would like to get a better understanding of your experiences using the AI-enabled ECG-based screening tool for left ventricular systolic dysfunction. This information will be used to further refine the tool and inform implementation strategies. The survey should take about 2-5 minutes to complete.
Last Name: First Name: Location: Care Team Name: :

Strongly Disagree Disagree
Neither agree or disagree Agree Strongly Agree 1. The screening tool provided valuable information I cannot obtain elsewhere. 2. I trust the information this new tool provides. 3. Clinicians in my care team have a shared understanding of the value of this screening tool. 4. I understand how to use the screening tool. 5. In the event of a positive screening result, I am well prepared to discuss with my patients. 6. I have sufficient resources to support the use of the new tool (e.g., education materials, training, and support from the investigative team and clinician champions). 7. Leadership and management adequately support the new screening tool. 8. The screening tool improved the care I provided to my patients. 9. I would like to continue using the tool when the trial ends. 10. I have no concerns related to costs for patients or the department.
If you answer "strongly disagree" or "disagree" to any of these questions above, can you provide some more details or explanation?
Is there anything else you want to tell us about your experience or any suggestions for improvement? prototypes were developed, tested, and revised based on these clinicians' feedback. The investigative team met regularly to discuss the iterations of the prototype and the clinicians' feedback. The prototype was also tested with five clinicians using real patient data and was then finalized based on the feedback. Other trial materials were developed by the multi-disciplinary team including physicians from cardiology and primary care, health services researchers, statisticians, and a study coordinator.