Use of ARIA in Risk Stratification for Chest Pain Patients Presenting to Emergency Departments Suspected Acute Coronary Syndrome

Acute Coronary Syndrome

Acute Coronary Syndrome, Acute Chest Pain, Fundas camera, Machine learning, Artificial Intelligence, Risk Stratification

The current assessment of patients with acute chest pain in the Emergency Department (ED) remains lengthy with the need for serial troponin. This contributes to overcrowding in the ED and work overload of clinical staff. These are associated with increased costs and adverse patient outcomes. The use of risk scores such at HEART score can be subjective and is not useful in risk stratification for those with higher risk (age and risk factors) to Major Acute Cardiac Event (MACE). Aim of Study: This study is designed to explore whether the use of Automatic Retinal Image Analysis (ARIA) can identify patients presenting with undifferentiated chest pain without the need for serial troponin test results in order to facilitate early and safely discharge and at high-risk MACE to receive early appropriate intervention. Hypothesis: ARIA or the combination with single troponin or HEART score can identify patients with undifferentiated chest pain presenting to the ED at low- and high-risk of adverse cardiac events within 30 days and 3 months after initial presentation. Procedure: The ARIA is a non-invasive and novel technology, it will be used to access the risk of acute coronary syndrome by analyzing of fundus (back of the eye) photo taken by a fundus camera. All subjects will be arranged to take a fundus photography (both eyes) by a conventional fundus camera, and capture the retinal photo. The images will be used to develop a risk stratification method for chest pain patients presenting to ED with suspected acute coronary syndrome (ACS). The fundus photography will be taken in the Emergency Department of Prince of Wales Hospital. The process takes about 5-8 minutes. Subject may feel discomfort for a short while at the time of photo taking due to flash exposure similar to ordinary camera flash, but the procedure is neither invasive nor painful. The fundus image will then be analyzed by computer algorithm developed by the research team. Apart from that, subject's medical history, ECG findings, age and sex, risk factors, and serial troponin levels will be recorded during their ED visit in order to work out the HEART score. Their disposal outcome from the ED will also be recorded. After 30 days, subject will be phoned to follow-up whether they have been readmitted into the hospital. If the subject have been readmitted, his/her investigation findings, diagnosis, treatment, disposal outcome, and length-of-stay will be recorded. The same follow-up process will be performed once more at 3 months after the subject has joined the study in his/her inital ED visit.

No related papers found yet.

None listed

Hong Kong