Atrial fibrillation (AF) is a major clinical and public health problem. AF is associated with an increased risk of stroke, embolic events, heart failure exacerbation, and all-cause mortality. Most of the evidence on the consequences of AF has focused on patients with persistent AF. However, the most common type of AF in the US is paroxysmal AF (PAF) which is often asymptomatic and unlikely to be identified except in subjects undergoing continuous cardiac monitoring. Hence, the natural history and health consequences of PAF and its progression towards persistent AF remain incompletely understood. Recent efforts focusing on understanding the role of PAF in triggering strokes and other major adverse events have been inconclusive and current evidence is insufficient to quantify the risk of stroke and other major adverse events associated with PAF episodes due to limited follow-up and sample size. To better understand the short- and long-term effects of PAF on the risk of death and other cardiovascular events, we will assemble a unique dataset of ~238,000 patients with dual-chamber implantable cardioverter defibrillators (ICDs) using remote monitoring implanted from 2006-2017 and with no persistent/permanent AF at the time of implant but who are at particularly high risk for PAF. This dataset represents the largest cohort of individuals with continuous monitoring of atrial arrhythmic events and is the result of an unprecedented combination of data from the three major ICD manufacturers (Medtronic, Boston Scientific, and St. Jude Medical), which collectively have 95% of the US ICD market. Reliable automated algorithms for detecting PAF episodes of duration >30 minutes are available in all dual-chamber ICDs which record time of onset, duration and heart rates. We will link PAF episode data with patient characteristics at the time of implant from the National Cardiovascular Data Registry and with hospitalization outcomes data from the Medicare utilization files and the National Death Index with up to 12 years of follow-up (~990,000 person-years of follow-up) through well-established linking methodologies using indirect data elements. The primary endpoint will be ischemic stroke (expected number of events ~4,700) and the secondary endpoints will be myocardial infarction (~8,600 events), other embolic events (~4,500 events), hospitalization for heart failure exacerbation (~20,500 events) and all-cause mortality (~96,700 events). We hypothesize that PAF episodes increase the short-term and the long-term risk of the primary and secondary endpoints and that faster rates of progression toward persistent AF are also associated with an increased risk of outcomes. The scope and size of this study will allow us to describe the consequences and progression of PAF and to quantitatively estimate, for the first time, the short- and long-term increase in risk associated with PAF episodes. These results will have important implications for clinical risk assessment and health policy.
The risk of clinical events associated with episodes of paroxysmal atrial fibrillation (AF), the most common form of AF in the United States, is unknown as most studies of AF and clinical outcomes have focused on persistent AF. To better understand the short- and long-term effects of paroxysmal AF on stroke and other major adverse events, we will assemble a unique dataset of ~238,000 patients, the largest to date, with dual-chamber implantable cardioverter defibrillators using remote monitoring implanted between 2006- 2017 with continuous monitoring and recording of AF events. By linking this database to the National Cardiovascular Data Registry, Medicare utilization files and the National Death Index, we will be able to establish how paroxysmal AF affects the risk for stroke and other major adverse events with important implications on patient management and health policy.
