Author + information
- Desmond J. Sheridan, MD, PhD∗ ( and )
- Desmond G. Julian, MD
- ↵∗Division of NHLI, C/o Catherine Enright, Guy Scadding Building, Royal Brompton Campus, Imperial College, London SW3 6NP, United Kingdom
In their response to our paper (1), Dr. Taylor and colleagues dismiss our suggestion that small sample size and publication bias may have contributed to conflicting evidence between published meta-analyses on mortality reduction attributed to cardiac rehabilitation. Their idea that the absence of funnel plot asymmetry and a nonsignificant Egger test exclude publication bias is unconvincing because these would not adequately reflect unpublished findings. Furthermore, we are aware of 2 studies which showed no effect on mortality that were not published and there may be others.
In our paper (1) we argued that the inclusion of clinical trials of rehabilitation carried out when survival of patients with coronary heart disease improved markedly is an important weakness of many meta-analyses of the effects of rehabilitation on mortality. We welcome that Dr. Taylor and colleagues now accept this. However, the “evidence” they offer in support of it (their Figure 1B) seems to us a good example of the concerns we also raised about a willingness to over interpret statistical analysis in the context of evidence-based medicine. This plot is claimed to show “a trend for a linear reduction in all-cause mortality over time,” and yet a virtually identical plot published in 2012 by these authors was used in support of the opposite view; namely, that there is “no strong evidence of a reduction in mortality effect over time” (2).
Dr. Taylor and colleagues state that cardiac rehabilitation is no longer advocated on the basis of a reduction in all-cause mortality. This would be a welcome change. However, in 2011 they claimed that rehabilitation reduced overall mortality by 13%. In their most recent advocacy for cardiac rehabilitation they sought to challenge clinical trial evidence which conflicted with a reduction in all-cause mortality due to rehabilitation (3), and their most recent publication obscures the issue by reporting a “survival benefit” without clarifying its basis (4).
We remain of the view that evidence about the mortality benefits of rehabilitation in patients with coronary heart disease is weakened by an over reliance on secondary research, based on studies of patients with markedly different clinical risks.
Please note: The authors have reported that they have no relationships relevant to the contents of this paper to disclose.
- American College of Cardiology Foundation
- Sheridan D.J.,
- Julian D.G.
- Taylor R.S.
- Dalal H.M.,
- Doherty P.,
- Taylor R.S.
- Anderson L.,
- Oldridge N.,
- Thompson D.R.,
- et al.