Cost-effectiveness of an insertable cardiac monitor to detect atrial fibrillation in large- or small-vessel disease ischaemic stroke in the USA.

objectivesTo evaluate the cost-effectiveness of insertable cardiac monitors (ICMs) compared to standard of care (SoC) to detect atrial fibrillation (AF) in patients with stroke of presumed known cause of large-artery atherosclerotic disease (LAD) or small-vessel occlusive disease (SVD) from a US payer perspective.

designA lifetime Markov model assessed cost-effectiveness of ICM versus SoC from a US payer perspective. Patient characteristics and AF detection rates were based on the STROKE AF trial (NCT02700945): 3-year diagnostic yield was 21.7% (95% CI 16.7% to 27.9%) for ICM and 2.4% (1.0%-5.7%) for SoC. AF detection resulted in a switch from aspirin to direct oral anticoagulant unless precluded by prior bleeding. Subsequent risks of ischaemic strokes (ISs) and bleeding events were modelled based on published literature. Costs and effects were discounted at 3% annually. Specific SoC short-term monitoring strategies (STMs) were explored as scenarios.

settingUS healthcare system perspective.

participantsHypothetical cohort of patients with IS believed to be due to LAD or SVD.

interventionsPatients received an ICM within ten days of the index stroke or SoC involving conventional follow-up.

main outcome measuresStroke and bleeding risk, mortality, health-related quality of life and healthcare cost and utilisation.

resultsICM was associated with a gain of 0.176 quality-adjusted life years (QALYs) compared with SoC per patient, representing a reduction of 53 strokes per 1000 patients. The lifetime incremental cost of ICM was $6736 per patient. This resulted in an estimated incremental cost-effectiveness ratio (ICER) of $38 346 per QALY gained, making ICM a cost-effective intervention at willingness-to-pay thresholds of $50 000-$150 000 per QALY in the USA. ICMs were also cost-effective compared with various individual STMs, with ICERs ranging from $29 814 to $38 941 per QALY gained. The mean probabilistic ICER across 5000 samples was $46 910 per QALY (97.5% CI$45 421 to $53 523). Results were sensitive to anticoagulant uptake on AF detection and underlying stroke risk. Model findings were robust to both probabilistic sensitivity analysis and sensitivity analyses where inputs tested were considered within plausible ranges, as ICM was found cost-effective in these analyses.

conclusionsICMs are highly likely to be a cost-effective diagnostic tool for secondary prevention of stroke related to AF in US patients with prior stroke attributed to LAD or SVD. However, further research is needed to understand the efficacy of secondary stroke prevention treatments in patients with stroke attributed to LAD or SVD and subclinical AF.