Author + information
- Vivek Y. Reddy, MD∗ (, )
- Ronald L. Akehurst, DSc and
- Stacey L. Amorosi, MA
- ↵∗Helmsley Trust Electrophysiology Center, Icahn School of Medicine at Mount Sinai, One Gustave L. Levy Place, Box 1030, New York, New York 10029
We appreciate the thorough review of our paper (1) by Drs. Singh and Wijeysundera and wish to clarify our analysis based upon their feedback.
Drs. Singh and Wijeysundera question whether our analysis was based solely upon a Watchman stroke risk of 0.95. We would like to emphasize that the value 0.95 reflects post-procedural ischemic stroke risk and that our analysis incorporates the 5 periprocedural strokes experienced in the PROTECT-AF (Watchman Left Atrial Appendage Closure Technology for Embolic Protection in Patients With Atrial Fibrillation) trial. In our model, patients who experienced procedural strokes incurred increased costs, decreased quality of life, and increased risk of subsequent stroke. Furthermore, the impact of ischemic stroke risk was evaluated in the sensitivity analyses. The variable “LAAC [left atrial appendage closure] relative risk of ischemic stroke” did not appear in Figure 2A (1) because only the top 10 variables having an impact on model results were reported. Readers will note risk of ischemic stroke to be the ninth variable listed in the comparison to non-warfarin oral anticoagulant agents (NOACs) (Figure 2B) (1). We agree that future analyses should consider event rates from both PROTECT-AF and the PREVAIL (Prospective Randomized Evaluation of the Watchman LAA Closure Device In Patients With Atrial Fibrillation Versus Long-Term Warfarin Therapy) studies.
We acknowledge potential differences among individual NOACs. However, most clinicians view NOACs as a group, with relatively minor differences among agents. Furthermore, policy and health care decision making most commonly consider therapeutic class effects. As such, the NOAC meta-analysis (2), which synthesized pivotal trial data for the 4 U.S. Food and Drug Administration-approved NOACs, was an ideal source for safety and efficacy data.
Finally, we agree upon the importance of selecting a model time horizon suited to the patient population under study. Recently published cost effectiveness analyses of stroke risk reduction in nonvalvular atrial fibrillation (NVAF) used lifetime analyses ranging from 20 to 30 years. Similarly, we reported lifetime results and demonstrated LAAC was the most cost-effective therapy relative to warfarin and NOACs. We questioned the relevance of these lifetime results for some decision makers, given the reduced life expectancy of many Medicare patients with NVAF. In an effort to be as informative as possible, and in accordance with good research practices published by the International Society for Pharmacoeconomics and Outcomes Research (3), we reported cost-effectiveness ratios at various time horizons. Nonetheless, we strongly support the caution provided by Drs. Singh and Wijeysundera and acknowledge that results may be misinterpreted if an inappropriately short time horizon is used, especially when a device-based procedure is compared to chronic pharmacologic therapy. However, we observed that once cost effectiveness was achieved, it was maintained over the lifetime horizon. This bolsters confidence in the results and appropriateness of reporting time to cost effectiveness.
Please note: Drs. Reddy and Akehurst are paid consultants for Boston Scientific. Ms. Amorosi is an employee of Boston Scientific.
Bruce D. Lindsay, MD, served as Guest Editor for this paper.
- 2016 American College of Cardiology Foundation