Prognosis, adverse events and characteristics of individuals at high risk for atrial fibrillation and heart failure: population-based linked cohort study of 3.5 million individuals

Background

Atrial fibrillation (AF) and heart failure (HF) are emerging cardiovascular epidemics,(1,2) and associated with adverse outcomes, but are often diagnosed late. Randomised controlled trials have investigated screening for AF and HF individually,(3,4) but screening simultaneously for both diseases in a high risk cohort may be more impactful on clinical outcomes. Individuals can be risk stratified for AF and HF using routine data but the outcomes of those at risk for both AF and HF are unknown.

Objective

To determine the clinical outcomes of patients identified as high risk for both AF and HF (herein referred to as ‘dual high risk’).

Methods

Using UK nationwide secondary care-linked primary care data we identified individuals aged ≥40 years without known AF or HF between January 1998 and 31 March 2021. Risks of AF and HF were estimated using the FIND-AF and FIND-HF models. High risk for each model was based on the pre-specified thresholds that optimised sensitivity, positive and negative predictive value when they were individually prospectively validated for AF and HF screening (FIND-AF≥0.0053, FIND-HF≥0.0125). Individuals were stratified as dual high risk, single high risk (high FIND-AF or FIND-HF only), and dual low risk. The primary outcome was a composite of HF event, stroke/TIA, myocardial infarction (MI), and cardiovascular death. We created Kaplan-Meier plots for each stratum and calculated cumulative incidence at five years and fit Fine and Gray’s models adjusting for competing risk of death for outcomes at five years.

Results

Of 3,101,727 individuals (mean age 64.5 years, 49.1% women), 790,534 (25.5%) were at dual high risk, 710,294 (22.9%) at single high-risk, and 1,600,899 (51.6%) at dual low risk. Compared with low risk and single high-risk individuals, those with a dual high-risk were older and more frequently were co-morbid (Table 1).

Dual high risk demonstrated the highest frequency risk of the primary outcome (cumulative incidence per 1000 persons at five years: 323.0, 95% CI 321.9-324.0) compared to single high risk (144.7, 95% CI 143.9-145.6) and low risk (59.0, 95% CI 58.7-59.4) (Figure 1), and higher hazard compared with low risk (HR 6.5, 95% CI 6.4-6.5) which was robust when adjusted for age and sex (aHR 6.1, 95% CI 6.0-6.2).

Individuals at dual high risk also had the highest frequency and hazard for each of the components of the primary outcome (HF event 61.8, 61.3-62.4; HR 9.1, 9.0-9.3; MI 178.8, 177.9-179.7; 5.2, 5.2-5.3; stroke/TIA 153.3, 152.4-154.1; 7.8, 7.7-7.9; cardiovascular death 63.7, 63.1.4-64.2; 18.0, 17.5-18.4).

Conclusions

Individuals at dual high risk for AF and HF frequently experience adverse events highlighting an opportunity for targeted early diagnosis and treatment.