Use of Artificial Intelligence-Guided Echocardiography to assIst cardiovascuLar Patient managEment (AGILE-Echo)

Alice Springs Hospital, Princess Alexandra Hospital, Brisbane, Australia, Royal Perth Hospital, Nepean Blue Mountains Local Health District, Ochre Health

Heart Failure and valvular heart disease are disproportionate problems in rural and remote Australia (RRA). Echocardiography is the best imaging investigation, and essential for management, but access to this essential test shows huge geographic variations, primarily because of dependence on expert acquisition. This trial seeks to demonstrate the effectiveness of artificial intelligence-based echocardiography for triage and management of patients with known or suspected heart disease in RRA.

Heart Failure (HF) and valvular heart disease (VHD) are disproportionate problems in rural and remote Australia (RRA) relative to the rest of the country, due in part to an ageing rural population and to the frequency of rheumatic heart disease in the Aboriginal community. Late diagnoses can lead to avoidable hospital admissions and expense to the Australian health system. Echocardiography is the imaging investigation of choice, and a cornerstone of management, but access to this essential test shows huge geographic variations in Australia. The primary reason for this is the dependence of this technique on expert acquisition. Artificial intelligence (AI) has now been harnessed to optimise echocardiographic image acquisition, and secure, cloud-based storage enables remote measurement and interpretation. This trial seeks to demonstrate the effectiveness of AI-based echocardiography-guided triage and management of patients with known or suspected heart disease in RRA. This study will involve the conduct of a world-first randomised controlled trial of AI-testing and early intervention to detect early stages of HF and VHD, select appropriate management, reduce admissions and preserve functional status and quality of life. The study will be conducted with partipants in RRA, aged 40 years and older with at least one HF risk factor and recruited through clinic and community outreach in four sites with the involvement of remote outreach from i) Alice Springs Hospitals, ii) Nepean Hospital to Dubbo Hospital and Western NSW, iii) Princess Alexandra Hospital to Roma, Charleville and Western Queensland, and iv) Perth Aboriginal communities in partnership with the Royal Perth hospital and the Derbarl Yerrigan Health Service. Approximately 1200 individuals at risk for HF and VHD will be screened and followed up. The study will be conducted in partnership with Aboriginal community partners.

AI-guided echo acquisition performed using a desktop echo machine (uSmart 3300, Terason, Burlington, MA) with AI software (Caption Health, Brisbane, CA) obtained by a non-expert (eg. clinic nurse, registrar general practitioner or physician) at the remote clinic, usually on the day of clinic review. Images will then be uploaded onto a secure cloud and downloaded at the core lab for measurement, interpretation and reporting.

Standard echo acquisition by a sonographer either at a referral hospital, or by a visiting team - usually necessitating a delay. Measurement, interpretation and reporting will happen as usual.

Inclusion Criteria: Age >45 years eligible for Medicare exercise intolerance or cardiovascular (CV) risk factors Exclusion Criteria: Known HF or HVD situations where cardio-protection is already indicated (eg. known CAD) comorbid conditions with life expectancy <2 years inability to provide written informed consent