EASE: Evaluating an AI driven stress echocardiography system
Research type
Research Study
Full title
EASE: Evaluating an AI driven stress echocardiography system (Ultromics EchoGo)
IRAS ID
315284
Contact name
Daniel Frings
Contact email
Sponsor organisation
London South Bank University
ISRCTN Number
ISRCTN15113915
Clinicaltrials.gov Identifier
Clinicaltrials.gov Identifier
NA, NA
Duration of Study in the UK
2 years, 4 months, 31 days
Research summary
Research summary - Evaluating the impact of new technology on populations at smaller scale is crucial before full-fledged implementation for all. This research aimed at examining the impact of EchoGo Pro, an artificial intelligence driven stress echocardiography analysis system designed to support diagnosis of cardiac conditions including cardiac ischaemia and potential coronary artery disease. EchoGo Pro provides clinicians with decision support system by generating diagnostic recommendations for the identification of coronary heart disease and risk of heart failure, etc.
The potential impacts of the technology in public health settings are (i) the autonomous nature of the system increases medical decision consistency between providers, (ii) improved diagnostic accuracy, (iii) support of earlier intervention and (iv) increased confidence to less experienced clinicians in making diagnostic decisions. These in turn increase speed and capacity to process tests, identify at-risk patients, and provide more preventive treatment and ultimately have fewer acute cardiac events amongst screened patients. This technology may also boosts the confidence of clinicians, especially junior doctors. These improvements in turn should benefit public in terms of good value and affordable health gains and, potentially, reducing health inequity.
Relying on a combination of interviews and focus groups and comparisons of quantitative health outcomes, our research intends to answer four questions; a) Is EchoGo Pro acceptable to patients and providers? Q2: Is EchoGo Pro safe and does it reduce health inequality? Q3: Does EchoGo Pro positively impact health problems through early or more frequent detection and intervention? Q4: Does the expenditure to implement EchoGo Pro can be justified by the improvements in health care decision making and ultimately patient outcomes and patient experience? Using financial metrics as a proxy for ‘value’ -what would an EchoGo Pro service be worth from the perspective of a patient, service manager/ or member of staff working in cardiac services?
Lay Summary - EchoGo Pro [EGP] is anAI technology supporting diagnosis of cardiac arterial disease[CAD]. We conducted a mixed-method evaluation including bench-testing, qualitative interviews, secondary analysis of a RCT dataset, surveys of clinicians and the general public and health economic analysis. We explored what processes were used to deploy EGP, the acceptability of the technology to clinicians, staff and patients, objective alignment between EGP and clinicians and the extent to which EGP delivers value. Our findings suggest that cardiologists are prepared to adopt AI cardiac tools,includingEGP,and they are perceived to be potentially acceptable. However, current technical limitations(especially high scan rejection rates)and clarity of the operational value of the offer remain a challenge. Deployment in Trusts proved difficult in the current evaluation and our findings suggest active support from NHS transformation leaders and clinicians is vital. Our findings suggest EGP increases diagnostic sensitivity without reducing specificity, and it is likely that wider deployment would lead to more people receiving appropriate CAD diagnoses. However, care must be taken to monitor for an increase in false positivesrates.These impacts appear to be stable across a variety of demographics. In terms of health economics there is still work to be done around establishing an effective price point (from a health economic perspective) which is also commercially viable.In conclusion, we would argue that, should price and deployment issues be addressed, EGP provides a viable method of increasing CAD diagnosis accuracy when used as a decision-making aid by experienced cardiologists
Has the registry been updated to include summary results?: Yes
If yes - please enter the URL to summary results: https://gbr01.safelinks.protection.outlook.com/?url=https%3A%2F%2Fclick.pstmrk.it%2F3ts%2Fosf.io%252Fu2vxf%252F%2FNBTI%2FpmC5AQ%2FAQ%2F416e3e1e-a026-4533-b192-67fb555adaff%2F2%2FX6EmBj9cgl&data=05%7C02%7Cbrightonandsussex.rec%40hra.nhs.uk%7Cd9a1e55403ad4e4969ae08dd057d7334%7C8e1f0acad87d4f20939e36243d574267%7C0%7C0%7C638672758541760866%7CUnknown%7CTWFpbGZsb3d8eyJFbXB0eU1hcGkiOnRydWUsIlYiOiIwLjAuMDAwMCIsIlAiOiJXaW4zMiIsIkFOIjoiTWFpbCIsIldUIjoyfQ%3D%3D%7C0%7C%7C%7C&sdata=oKYObG5pEovlR7pt60lSKzQDaBHJnmxOIAvsSZwFjsk%3D&reserved=0
If no – why not?:
Did you follow your dissemination plan submitted in the IRAS application form (Q A51)?: Yes
If yes, describe or provide URLs to disseminated materials: Papers are in review, workshops have been undertaken (see below). 1 international conference paper delivered
If pending, date when dissemination is expected:
If no, explain why you didn't follow it:
Have participants been informed of the results of the study?: Yes
If yes, describe and/or provide URLs to materials shared and how they were shared: A variety of workshops aimed at key stakeholders help inform the final findings and study conclusions.
If pending, date when feedback is expected:
If no, explain why they haven't:
Have you enabled sharing of study data with others?: Yes
If yes, describe or provide URLs to how it has been shared: For elements where data can be shared; https://gbr01.safelinks.protection.outlook.com/?url=https%3A%2F%2Fclick.pstmrk.it%2F3ts%2Fosf.io%252Fy9jvc%252F%2FNBTI%2FpmC5AQ%2FAQ%2F416e3e1e-a026-4533-b192-67fb555adaff%2F3%2FovR9mRpVmR&data=05%7C02%7Cbrightonandsussex.rec%40hra.nhs.uk%7Cd9a1e55403ad4e4969ae08dd057d7334%7C8e1f0acad87d4f20939e36243d574267%7C0%7C0%7C638672758541785611%7CUnknown%7CTWFpbGZsb3d8eyJFbXB0eU1hcGkiOnRydWUsIlYiOiIwLjAuMDAwMCIsIlAiOiJXaW4zMiIsIkFOIjoiTWFpbCIsIldUIjoyfQ%3D%3D%7C0%7C%7C%7C&sdata=rNkPADJqxBLKhD%2Fsh3DjCOWGrPvxkmKt4GyTwEyKA00%3D&reserved=0REC name
London - Brighton & Sussex Research Ethics Committee
REC reference
22/PR/1068
Date of REC Opinion
21 Sep 2022
REC opinion
Further Information Favourable Opinion