The PARADISE study aims to develop and validate prediction tools to identify patients at risk of Atrial Fibrillation (AF) after cardiac surgery.
Atrial Fibrillation (AF) is a common abnormal heart rhythm. AF causes the heart to beat irregularly and sometimes very rapidly. About 30-50% of patients develop AF after heart surgery. These patients stay longer on the Intensive Care Unit (ICU) after surgery, are more likely to develop complications and have a higher risk of dying. Avoiding AF is important. Some drugs, including beta blockers and amiodarone may help prevent AF if given after surgery. However, these may also lead to complications (such as lung damage). It is therefore important to identify which patients are most likely to benefit from these treatments (i.e., where the benefits outweigh the risks). There are existing tools designed to predict the risk of suffering AF after heart surgery. However, they are unreliable and therefore not used in clinical practice. A modern, reliable risk prediction tool is needed. The PARADISE study will develop and test new prediction tools to identify which patients are most at risk of developing AF after heart surgery. The investigators will focus our tools on those patients who most commonly develop AF, such as those who have had surgery to repair a valve or blood vessel in their heart. To do this the investigators will: * Review the medical literature and assemble a panel of medical experts to create a list of known factors that affect patients' risk of AF after heart surgery * Use a large UK general practice database (CALIBER) to see whether the investigators can find new risk factors. * Ask the expert panel to agree a list of known and new risks factors to be included in the prediction tool. * Develop two new prediction tools using an existing American cardiac surgery database (the Partners research Database). The first will be used before surgery, the second immediately following surgery. Two models are needed as events during surgery may alter the risk of AF. * Test how reliably our new tools predict which patients suffer AF after surgery, with data from large UK (United Kingdom) NHS (National Health Service) heart centres, one US Hospital (Brigham) and a UK clinical trial (Tight-K). * The investigators will work with two charities (AF Alliance and StopAfib) to share our results with patients and the wider public.
Study Type
OBSERVATIONAL
Enrollment
13,684
Not applicable as observational study
Critical Care Research Group, Nuffield Department of Clinical Neurosciences, University of Oxford
Oxford, Oxfordshire, United Kingdom
Model discrimination (c-statistic) to predict Atrial Fibrillation in external data set
Model discrimination (c-statistic) to predict Atrial Fibrillation in external data set
Time frame: Within 7 days of cardiac surgery
Model calibration (intercept) to predict Atrial Fibrillation in external data set
Model calibration (intercept) to predict Atrial Fibrillation in external data set
Time frame: Within 7 days of cardiac surgery
Model calibration (slope) to predict Atrial Fibrillation in external data set
Model calibration (slope) to predict Atrial Fibrillation in external data set
Time frame: Within 7 days of cardiac surgery
Additional model performance metrics to predict Atrial Fibrillation in external data set
Model positive and negative predictive values, sensitivity and specificity to predict Atrial Fibrillation in external data set
Time frame: Within 7 days of cardiac surgery
Candidate risk factors for inclusion in new onset atrial fibrillation prognostic models
Candidate risk factors for inclusion in new onset atrial fibrillation prognostic models, identified through Systematic literature review and analysis of the CALIBER database using statistical and machine learning methods. For pre-operative model, the investigators will include patient information available up to the time of surgery. For the post-operative model, the investigators will also include patient information available up to 12 hours after surgery.
Time frame: Within 7 days of cardiac surgery
This platform is for informational purposes only and does not constitute medical advice. Always consult a qualified healthcare professional.