Key result
A 4-year lookback optimally estimates AF incidence, revealing a ~31% mortality drop despite rising prevalence.
Why the study?
Estimating AF incidence in administrative databases depends on the retrospective lookback period used, but the optimal lookback period remains unclear.
What is the optimal lookback period for estimating atrial fibrillation incidence in administrative databases, and what are the long-term trends in AF burden and outcomes?
Cohort (n=1,558,389)
Yes
What is the optimal lookback period for estimating atrial fibrillation incidence in administrative databases, and what are the long-term trends in AF burden and outcomes?
A 4-year lookback period is optimal for estimating AF incidence in administrative databases, revealing a stable incidence but increasing prevalence and declining cardiovascular mortality over 14 years in Korea.
Answers are grounded in this paper and cite their sources.
Supports 4-year lookback for AF incidence estimation in claims data; extends evidence on rising prevalence with stable incidence and better outcomes.
Background/Introduction The prevalence of atrial fibrillation (AF) is increasing worldwide, contributing to a significant healthcare burden. Long-term surveillance is essential to understand evolving trends in epidemiology of AF. The estimation of AF incidence in administrative databases is dependent on the choice of lookback period—the retrospective search window used to distinguish new-onset cases from pre-existing cases. However, the optimal lookback period for estimating AF incidence remains unclear. Purpose This study aimed to (1) determine how variations in lookback periods influence AF incidence, (2) establish an appropriate and reproducible lookback period for accurate epidemiological assessment, and (3) investigate the long-term trends in the incidence, prevalence, and clinical burden of AF in Korea from 2006 to 2021, Methods This retrospective cohort study utilized the nationwide claims database, including individuals aged ≥20 years diagnosed with AF from 2002 to 2021. Incident cases were identified using a minimum fixed 4-year lookback period to ensure consistency in incidence estimates (Figure 1A). A lookback period was deemed suitable if the calculated incidence rate stayed within 10% of the reference rate derived from a 10-year lookback period. All-cause mortality and cardiovascular outcomes including ischemic stroke, intracranial hemorrhage, myocardial infarction and heart failure were evaluated using hazard ratios (HR) with 95% confidence intervals (CI). Results Among 1,558,389 patients diagnosed with AF, the prevalence increased from 0.83% in 2006 to 2.25% in 2021, with the highest burden observed in individuals aged ≥80 years. Standardized AF incidence rates remained relatively stable, ranging from 2.49 per 1,000 person-years in 2006 to 2.33 in 2021. The fixed 4-year lookback period was identified as the most appropriate (Figure 1B). The incidence rates varied significantly when different lookback periods were applied (Figure 1C). Over the 14-year period, the risks of all-cause mortality and cardiovascular outcomes significantly declined: mortality (HR 0.69, 95% CI 0.67–0.70), intracranial bleeding (0.57, 0.53–0.62), ischemic stroke (0.73, 0.70–0.76), myocardial infarction (0.53, 0.48–0.58) and heart failure (0.79, 0.76–0.83). Conclusions This study highlights the critical importance of selecting an appropriate lookback period when estimating AF incidence in administrative databases. These findings emphasize the necessity of transparent methodologies to improve the accuracy of AF surveillance and guide future healthcare policies.
No takes yet. Share an insight, caveat, or question.
Kim et al. (2025) conducted a cohort in Atrial fibrillation (n=1,558,389). Calendar year (14-year period) was evaluated on All-cause mortality (HR 0.69, 95% CI 0.67-0.70). A fixed 4-year lookback period was optimal for estimating AF incidence, while over a 14-year period, all-cause mortality in AF patients declined.
Synapse has enriched 10 related papers. Consider them for comparative context: