OBJECTIVES

: Detection of paroxysmal atrial fibrillation (pAF) is important for optimal prevention of recurrent stroke. This study aimed: 1. to evaluate if inpatient monitoring for pAF detection is cost effective compared with outpatient monitoring and no monitoring; 2. to determine how many consecutive days of inpatient monitoring is most cost effective.

METHODS

: Markov model and two-level Monte Carlo simulation were applied to simulate stroke progression, and to estimate the lifelong cost and quality adjusted life years (QALYs) for patients hospitalized due to ischemic stroke (IS) or transient ischemic attack (TIA) presenting in sinus rhythm. Seven disease states were considered in the model: TIA, minor IS, major IS, recurrent minor IS, recurrent major IS, intracranial hemorrhage or hemorrhagic stroke, and death. Both primary clinical study and secondary database studies were conducted to obtain key model parameters. Costs were estimated from patients’ perspective, which included pAF monitoring cost, medication cost, post-stroke healthcare cost and care giver’s cost, acute care cost for recurrent stroke.

RESULTS

: The lifelong cost and effectiveness was S$34,575 and 5.68 QLAYs for patients without pAF monitoring; S$34,688 and 5.70 QLAYs for patients with 24h outpatient monitoring; S$34,621 and 5.71 QLAYs for patients with continuous inpatient monitoring. The incremental cost effectiveness ratio (ICER) of 1-day inpatient monitoring was S$1683.17 per QALY gained. There is a 77% chance that continuous IP monitoring up to 9 days is the most cost effective strategy given that the willingness to pay is S$ 50,000 according to Monte Carlo simulation.

CONCLUSIONS

: Inpatient pAF monitoring is cost effective compared with outpatient or no monitoring. Continuous inpatient pAF monitoring up to 9 days is the most cost effective strategy for hospitalized ischemic stroke patients.