Author + information
- David Feldman, MD, PhD, FACC⁎ ()
- ↵⁎Reprint requests and correspondence:
Dr. David Feldman, Department of Medicine/Cardiology, Physiology, and Cell Biology, The Ohio State University, Suite 200, Davis Heart and Lung Institute, 473 West 12th Avenue, Columbus, Ohio 43210-1252.
In this issue of the Journal, Elmariah et al. (1) bring forward a provocative and potentially contentious idea—that men and women are not the same and should be treated differently. Specifically, these investigators assert that for patients with advanced heart failure who may require a transplant, there should be different objective criteria for determining when a patient should undergo transplantation based on the patient’s gender. The application of the investigators’ conclusion would suggest that women are either genetically superior, possess different myocyte transduction mechanisms to compensate for the same degree of heart failure (1), or should receive inferior health care based on a number (metabolic stress test). The danger of this number is that it was derived from a single-center retrospective database and from a relatively small number of patients. More importantly, neither the investigators nor I understand the mechanism for the observed difference. This makes the results an observation, not an irrefutable fact. The danger in accepting the results in this study is not the validity of the observations, but the potential application of this study to clinical practice without the understanding of a mechanism. Potentially, the results might be applicable to clinical practice if they had been substantiated by a prospective randomized trial. This would not likely provide the mechanistic data required to understand the observations put forward in this article, but it would provide additional support for a meaningful concept that may take physicians years to understand, or possibly may never be understood. Although this is easy for me to request, it is far more difficult to orchestrate in clinical practice. An internal review board would certainly pronounce that such a study was unethical because it discriminates against a preselected group of patients. Of course, the only end point that would address the issues in the paper would be all-cause and cardiovascular mortality, with everything else perceived as supportive or damning evidence for this single assertion. Therefore, clinicians and scientists find themselves at an impasse. If you do not apply the results of this article to clinical practice, and there may never be a prospective randomized control trial, where are we with the concerns put forth by these investigators?
To state the obvious, all papers have their shortcomings and strengths. Many of the limitations of this specific study have been enumerated by the investigators in the discussion section of the article. As such, it would be redundant to recapitulate those issues again. However, there are a few additional concerns worth considering in a critical review of this well-constructed analysis: 1) It would be overly simplistic to suggest that any patient should or should not undergo transplantation based on any number. 2) The women in this study were younger than the men and had higher ejection fractions, and variables such as New York Heart Association functional class, standardization of the exercise protocol, prior exercise conditioning, and renal function were all unknown. All of these variables have been known to impact mortality in heart failure; hence, it is plausible that some of these variables may have confounded the results of this study. In addition, although there were an equal number of patients using beta-blockers, the dose that may affect mortality (1) for any given patient was unknown (controlling for heart rate may not be sufficient). 3) As a final concern, the percentage of patients who were women in this study was 28% of the total population analyzed; this may suggest a selection bias.
In contrast to these concerns, this article also has some unique features worthy of mention. Further analysis from these investigators showed that patients who had achieved anaerobic threshold (by respiratory exchange ratio or V-slope methodology) still had a similar disparity with concordant results from the published article. In addition, 167 women from diverse backgrounds were represented in this single-center study. This represents the largest number of women studied in this setting, albeit not a huge number. Furthermore, these investigators did account for some of the most probable confounding variables, including lung function, baseline medications (but not doses), age, race, diabetes, hypertension, atrial fibrillation, and heart failure etiology.
Unfortunately, the conundrum of how to interpret these results becomes even more complicated when one reflects on this subject for another moment. Investigators use statistical analysis as one of our most important tools for determining the relevance of a specific finding, but regrettably standard statistical analysis can fail any investigator with the best of intentions. Examples may include a statistically relevant p value with a result that may not be clinically significant or the relevance of the p value of 0.05 or 0.01 in any specific study. That is, these p values would suggest that an investigator would come to the same conclusion 95 or 99 of 100 times, or in the current study, 999 of 1,000 times. In many studies, investigators power their statistics for a specific outcome with hundreds or thousands of patients. In such a case, perhaps a p value 0.053 may be clinically significant and not statistically significant. Furthermore, if a trial has a total of 10 patients and there is a difference between groups with a p value of 0.001, the validity of the results would certainly require a larger study with power calculations, and an advisory panel. In contrast to patients before transplantation, women after transplantation may have a higher risk of death compared with men. Statistics suggest that women have a 5% higher rejection rate than men in the first year, and are a greater risk of death by generally having a higher antigen profile and being of shorter stature. These three findings are all associated with higher death rates with statistically significant p values. Further analysis of the results suggests that women have a five-year mortality rate that is 13% higher than that for men with a p value of 0.02 with a very large number of patients, but at the end of multivariate analysis and conditional modeling, these variables do not suggest that women are disadvantaged (4). This illustration may suggest that we unconsciously may perceive a given result or data set to fit our biases.
One of the most important contributions of this article is reflection on the larger questions that these investigators were attempting to address. One of the statistical assertions of this article is that patients should not undergo transplantation (or possibly receive heroic therapies) if they only have a 15% chance of dying in a one-year period. This is a mortality rate that would not be well tolerated for many other disease processes. Today, with such a devastating mortality rate for heart failure patients, we must consider whether we are asking the right questions, or whether we have become too comfortable with our current understanding of heart failure mechanisms, pathophysiology, and therapeutics. Perhaps we do have some of the correct medications and are using them in the wrong patients. Case in point, the one-year mortality rate for an acutely decompensated patient is increased with a hospital admission (5), and the hospital five-year mortality rate in many studies continues to be 50% (6). In addition, the investigators have proposed MVo2cutoff levels that are substantially lower than those currently used by many transplantation centers. The Elmariah et al. (1) paper focuses our attention on considering whether there is a general need to reassess the criteria for transplantation in the modern era of new medical therapies and mechanical device technology. The inferred subtext of this suggestion is that we may be performing transplantations unnecessarily or prematurely and thereby squandering a very limited resource. This paper forces us to think about uncomfortable issues such as how we interpret our current practice, what we do not know, and how much further we have to go (2,3).
↵⁎ Editorials published in the Journal of American College of Cardiologyreflect the views of the authors and do not necessarily represent the views of JACCor the American College of Cardiology.
- American College of Cardiology Foundation
- Elmariah S.,
- Goldberg L.R.,
- Allen M.T.,
- Kao A.
- Bristow M.R.,
- Gilbert E.M.,
- Abraham W.T.,
- et al.,
- MOCHA Investigators
- Wang T.J.,
- Evans J.C.,
- Benjamin E.J.,
- Levy D.,
- LeRoy E.C.,
- Vasan R.S.