Author + information
- Sheldon M. Singh, MD∗ ( and )
- Harindra C. Wijeysundera, MD, PhD
- ↵∗Schulich Heart Program, Sunnybrook Health Sciences Center, 2075 Bayview Avenue, Toronto, Ontario M4N 3M5, Canada
The study by Reddy et al. (1) adds to the growing body of work evaluating the cost effectiveness of stroke prevention therapies from a U.S. perspective. We applaud their use of decision analytical models to inform practitioners and payers. We highlight 3 important considerations for readers of this work.
First, the Watchman stroke risk (Boston Scientific, Marlborough, Massachusetts) in this model was 0.95, which is remarkably lower than the value of 1.26 reported in the 4-year PROTECT-AF (The WATCHMAN Left Atrial Appendage Closure [LAAC] Device for Embolic PROTECTion in Patients with Atrial Fibrillation) study, presumably due to the exclusion of 5 periprocedural strokes. Despite this difference, a sensitivity analysis with the variable “Watchman relative risk of ischemic stroke” was not reported. We encourage the authors to report this as, in addition to the appreciable difference between the stroke risk values of 0.95 and 1.26, prior work has demonstrated a significant impact of ischemic stroke risk on Watchman cost effectiveness (2). Furthermore, a recent meta-analysis of Watchman trials reported an even higher (1.56) post-procedure ischemic stroke risk (3).
Second, efficacy and safety estimates derived from a meta-analysis of novel oral anticoagulants (NOACs) may obscure the potential differences between NOACs. In fact, a cost-effectiveness analysis conducted in Ontario, Canada by our group previously demonstrated that apixaban dominated the Watchman stroke risk (4).
Third, we question the theoretical underpinning of the concept of “time to cost-effectiveness.” A model’s time horizon should reflect the period during which meaningful differences between the interventions being studied can be expected—in the case of atrial fibrillation stroke prevention, typically a lifetime. By presenting results over different time horizons, the authors demonstrate the inaccuracies that may be observed when choosing an inappropriately short time horizon (5). We caution misinterpretation of these results to imply that a certain time period must pass before left atrial appendage occlusion transitions from being economically unattractive to cost effective; rather, the analysis illustrates that a short time horizon would be poor modeling practice and methodologically unsound.
Please note: The authors have reported that they have no relationships relevant to the contents of this paper to disclose.
Bruce D. Lindsay, MD, served as Guest Editor for this paper.
- 2016 American College of Cardiology Foundation
- Reddy V.Y.,
- Akehurst R.L.,
- Armstrong S.O.,
- et al.
- Micieli A.,
- Bennell M.C.,
- Pham B.,
- et al.
- Holmes D.R.,
- Doshi S.K.,
- Kar S.,
- et al.
- Sander B.,
- Krahn M.D.