Prevalence of Undiagnosed Atrial Fibrillation in Elderly Individuals and Potential Cost-effectiveness of Non-invasive Ambulatory Electrocardiographic Screening: The ASSERT-III Study

Background: In approximately 10% of patients with implanted pacemakers or defibrillators, previously unrecognized atrial fibrillation (AF) is detected within 3 months. It is unknown whether elderly patients without implanted devices have a similar prevalence of undiagnosed AF using non-invasive ECG monitoring, and if this approach to screening in this population is cost-effective.

Methods: Individuals ≥80 years old attending outpatient clinics without a history of AF and with hypertension and one additional risk factor underwent 30 days of continuous ECG monitoring with an option for an additional 30 days of monitoring if no AF was detected. The primary outcome was AF ≥ 6 min. Cost-effectiveness to prevent stroke was estimated using a Markov model based on observed AF detection rates and data from the published literature.

Results: Among 129 patients enrolled, 100 initiated monitoring for an average duration of 36 ± 21 days. The proportion of patients that completed at least 30 days of monitoring was 59%. Average age was 84 ± 3 years and mean CHADS-VASc score was 4.5 ± 1.2. AF ≥ 6 min was documented in 14%, ≥6 h in 8%, and ≥24 h in 3%. One week of monitoring costed $50,000 per quality-adjusted life-year-gained, 30 days and 60 days of monitoring costed $70,000 and $84,000, respectively.

Conclusions: Continuous non-invasive ECG monitoring is feasible in elderly patients. Undiagnosed AF is present in many elderly individuals, with 1 in 7 having episodes lasting ≥6 min. One week of monitoring may be cost-effective for stroke prevention in this population.