Many people with heart failure (HF) develop an abnormal heart rhythm, known as atrial fibrillation (AF). AF often causes breathlessness and lower life-expectancy in people with HF. Current treatment of AF involves slowing the heart rate with medication. If people have ongoing symptoms, we try to restore normal rhythm using special drugs or an ablation procedure. An ablation procedure involves passing thin tubes through blood vessels in the leg to the heart to create small burns that block electrical signals causing AF.
A recent clinical trial showed that restoring normal rhythm earlier, within 12 months of AF diagnosis, either with drugs or ablation, can reduce the risk of death. Only a few people in the trial had HF. Therefore, we will assess how many people there are with HF and AF in the population who could be eligible for early treatment. This information will be useful for planning healthcare services.
Secondly, ablation is better at restoring normal rhythm than drugs, but we don’t know if ablation reduces deaths or hospital admissions in people with HF. This information would help patients and doctors better understand the risks and benefits of ablation when deciding on their treatment. We will compare rates of death and hospital admissions between people with AF and HF undergoing ablation compared with those receiving drugs. We will also study whether a patient’s gender or other medical conditions, such as diabetes or lung disease, might change the effects of AF ablation and should therefore be considered when planning treatment.
The EAST-AFNET4 trial suggested that early rhythm control improves outcomes in patients with AF. However, few patients in EAST-AFNET had HF. We will use primary and linked secondary care data to evaluate the proportion of patients with diagnostic codes for AF and HF in England between 2004-2024, who are eligible for early rhythm control and define the incidence of death and cardiovascular hospitalisations in this group.
Secondly, we will evaluate whether AF ablation is associated with better survival and lower hospitalisation rates than pharmacological rate control among patients with HF and AF. We will include patients who do not have an implanted cardiac device and those prescribed newer HF therapies, such as ARNI and SGLT2 inhibitors, for whom there is no high-quality evidence. To achieve similar treatment groups, we will calculate a high-dimensional propensity-score (hdPS) based on empirical variables across five dimensions (hospitalisations, procedures, diagnoses, prescriptions, laboratory values). Conditional Cox proportional hazards regression (stratified by matched-pair) will be used to compare the risk of each outcome in the hdPS weighted cohort. We will test additional models with biologically relevant investigator-selected covariates manually forced into the hdPS model, e.g. age, sex, use of antiarrhythmic and HF medications in the year before the index date (adjusted-matched model). A final model will include interaction terms to study effect modification between the exposure (ablation) and age, gender, time since first AF documentation and comorbidity burden.
This project will benefit patients by defining the generalisability of the EAST-AFNET trial and effect size of AF ablation in patients with HF in England, thereby filling an evidence gap needed to inform clinical practice guidelines and enable the design of future clinical trials to improve care for this population. By defining the size of the HF population eligible for ablation, this study will also support healthcare resource planning.
1. Proportion of patients with AF and HF eligible for early rhythm control.
2. Rates of all-cause mortality and CV hospitalisation in patients eligible for early rhythm control.
3. Rates of all-cause mortality and CV hospitalisation in patients undergoing ablation-based rhythm control versus pharmacological rate-control therapy.
Rosita Zakeri - Chief Investigator - King's College London (KCL)
Rosita Zakeri - Corresponding Applicant - King's College London (KCL)
Chloe Bloom - Collaborator - Imperial College London
HES Admitted Patient Care;Patient Level Index of Multiple Deprivation