Author + information
- Alexandra Ward, PhD,
- Shien Guo, PhD and
- J. Jaime Caro, MDCM⁎ ()
- ↵⁎Caro Research, 336 Baker Avenue, Concord, Massachusetts
Recently, Feldman et al. (1) published a cost-effectiveness analysis based on the COMPANION (Comparison of Medical Therapy, Pacing, and Defibrillation in Heart Failure) trial, comparing cardiac resynchronization therapy (CRT) pacemaker alone or CRT-ICD implantable cardioverter-defibrillator versus optimal pharmacological therapy (OPT) alone, projecting the implications of the trial over seven years. Hlatky, in an editorial commenting on this analysis, derived incremental cost-effectiveness ratios (ICERs) comparing CRT-ICD to CRT pacemaker alone (2). Although we agree that it is the ICER that is of interest, several aspects of the Feldman analysis are of concern for this derivation.
Forcing the utility to be equal at baseline yet accepting the nonsignificant differences observed at later time points creates a differential quality adjustment against CRT-ICD, and this leads to a doubling of the derived ICER. This is inappropriate as ICDs should affect only mortality, not quality of life (3). Had the values assigned postbaseline been equalized as could reasonably have been assumed, the quality-adjusted life-year (QALY) estimate for CRT-ICD would substantially increase and the ICER drop by half.
Given that most, if not all, of the difference between the two CRT modes is in mortality, a proper comparison would require careful, realistic modeling of the mortality and of competing risks. Although the model used by Feldman et al. is barely described in the study, the investigators state that they assumed a constant death probability per month. This sort of oversimplification has often been felt necessary with Markov models, but does not correctly simulate mortality and will not properly support predictions of the number of lives saved by CRT-ICD over longer periods of time.
Finally, Hlatky (2) cites as fact the $50,000 threshold for cost-effectiveness, although this value has no basis whatsoever. A more sophisticated approach to modeling is necessary to estimate the potential long-term health benefits of CRT-ICD relative to CRT alone and appropriately assess the incremental cost-effectiveness.
Please note: The authors are employees of Caro Research Institute (Concord, Massachusetts), a consultancy that has received grants for unrelated research from various pharmaceutical companies. Caro Research has also received grants for research from Medtronic, the maker of products potentially affected by this work.
- American College of Cardiology Foundation
- Feldman A.M.,
- de Lissovoy G.,
- Bristow M.R.,
- et al.
- Hlatky M.A.
- Carson P.,
- Anand I.,
- O’Connor C.,
- et al.