Predicting Outcomes of Cardioversion for Atrial Fibrillation
Research type
Research Study
Full title
Predicting Outcomes of Cardioversion for Atrial Fibrillation
IRAS ID
322557
Contact name
Dhiraj Gupta
Contact email
Sponsor organisation
Liverpool Heart & Chest Hospital
Duration of Study in the UK
1 years, 0 months, 1 days
Research summary
Electrical (direct current) cardioversion (DCCV) is a procedure utilising an electric shock to restore normal rhythm to the heart. This is commonly performed for atrial fibrillation (AF), which is the most common abnormal heart rhythm.
Whilst frequently successful, many patients have early recurrence of their abnormal rhythm. We would like to study which factors may predict unsuccessful outcomes by analysing data from our electronic patient record over several years. This will include factors such as age, sex, body mass index, medical conditions, medications, imaging parameters, etc.
We will also utilise machine learning techniques to analyse the patient's ECG to see if this can accurately predict outcomes. To the best of our knowledge, this has not been tried before.
This may involve sending pseudonymised copies of the ECGs (all identifiable data removed and a code number assigned) to an external company called PMCardio. This company uses an AI application which can digitise the ECG - i.e., convert the image into a series of numbers which allow recreation of the trace. These data will be sent back to us and fed into our own novel machine learning algorithm in order to attempt prediction of outcomes. If this is not feasible (discussions in progress), an alternative approach is to utilise the ECG image itself in a machine learning model called a 'convolutional neural network'.
Additionally, we would like to study long-term outcomes of patients undergoing cardioversion - such as subsequent need for catheter ablation procedures, pacemaker implant, adverse outcomes and death. Factors which may predict adverse outcomes would be beneficial to understand when referring patients for this procedure.
We have around 1300 cardioversion cases over a 6-year period which could be analysed in this way.
REC name
North West - Greater Manchester West Research Ethics Committee
REC reference
23/NW/0238
Date of REC Opinion
10 Aug 2023
REC opinion
Favourable Opinion