July 31st, 2013
European Heart Guidelines Based on Disgraced Research May Have Caused Thousands of Deaths
Despite a 2-year-old scandal discrediting key evidence, current guidelines relying on this evidence have not been revised. As a result of physicians following these guidelines, some researchers say, it is possible that thousands of patients may have died each year in the U.K. alone. It is unlikely that a true understanding of the damage will ever be known.
Current European Society of Cardiology guidelines recommend that beta-blockers be given to many patients having surgery for noncardiac reasons to protect the heart during surgery. (U.S. guidelines are somewhat less aggressive in their endorsement of perioperative beta-blockade.) The guidelines, which were published in 2009, were based on analyses of the available trials. The strongest evidence came from the DECREASE family of trials, which appeared to strongly support perioperative beta-blockade, and one other large trial, POISE, which raised concerns that beta-blockers might lead to an increase in deaths. When the ESC committee combined all the data, they found a neutral effect on mortality but a strong benefit due to significant reductions in non-fatal MI and stroke with beta-blocker use. This was the basis for the strong recommendation in the ESC guidelines.
In 2011, however, faith in the reliability of the DECREASE trials was shattered as a result of a scientific misconduct scandal centering on the principal investigator of the studies, the now disgraced Dutch researcher Don Poldermans. The issue was further complicated because, in addition to his key role in the trials, Poldermans was the chairman of the committee that drafted the guidelines.
Now, a group of U.K. researchers, led by Darrel Francis, have published in the journal Heart the results of a meta-analysis of the remaining non-DECREASE trials that tested perioperative beta-blockade. With the removal of the DECREASE trials, the findings were strikingly different from the earlier analyses. In a combined population of 10,529 patients taken from nine trials, there was a statistically significant increase in the risk for death in the group randomized to beta-blockers:
- 162 deaths in 5264 patients randomized to beta-blockers versus 129 deaths in 5265 patients randomized to placebo, for a 27% increase in the risk for death (RR 1.27, CI 1.01- 1.60, p=0.04).
In sharp contrast, the two randomized placebo-controlled DECREASE studies found a 58% reduction in mortality associated with beta-blockers (although this was not statistically significant).
The analysis did find some benefits associated with beta-blockade, including a statistically significant reduction in non-fatal MI. However, it also found a significant increase in hypotension and stroke. In the new meta-analysis, any beneficial effects are clearly overshadowed by the most important finding of an increased risk for death.
Each year in the U.K., according to the Heart authors, 2.5 million procedures are performed for which this treatment is recommended in the current guidelines. Based on a 27% increase in the risk for death associated with perioperative beta-blockade, they calculated that as many as 10,000 deaths might be caused by physicians faithfully following the guidelines. In an interview, Francis said that there is no way to reliably assess the true extent of the possible damage, either in the U.K. or elsewhere. The estimate of 10,000 deaths is based on the limited available data.
The Heart authors concluded:
Patient safety being paramount, guidelines for perioperative β-blocker initiation should be retracted without further delay. Future guidelines should be accompanied by a commitment from named individuals to retract them immediately if the advice given is later revealed to be harmful.
Routine initiation of β-blockers for this indication should not be recommended, except in the context of RCTs which should be designed carefully, conducted honestly and reported truthfully.
The chair of the ESC Clinical Practice Guidelines Committee, Jose Luis Zamorano, said that the ESC is currently revising the guidelines, with a new version expected in the first months of 2014. He said the ESC was taking the meta-analysis “very seriously” and would “convene an urgent task force to decide whether further actions are required.”
The new meta-analysis is “interesting but not surprising,” said Sripal Bangalore, the author of a 2008 meta-analysis that raised early questions about both the European and U.S. guidelines. He said the paper is concordant with the POISE trial and his own meta-analsysis. He said it was “difficult to prove” the estimates of the number of deaths that may have been caused by the guidelines.
Leslie David Hillis, the chair of the ACCF/AHA CABG guidelines, said that the Heart paper data and conclusions were “accurate” and that the findings were potentially important.
Sanjay Kaul said that he generally agreed with the conclusions of the Heart paper that the evidence does not support strong recommendations for perioperative beta-blockade, but he said that “the mortality evidence against beta-blockade is not robust.” In his own practice, he said, he does “not initiate beta-blockers to modify perioperative risk even in high-risk patients. I also do not hold beta-blockers prior to surgery in those who are on them long term.”
Mariell Jessup, speaking on behalf of the American Heart Association, said that she thought “the issue of how we use beta-blockers in the perioperative period is very complex and probably requires more trials, irrespective of what we think of some of the trials in the past, and requires great care as we move forward.” She said this issue was now the subject of intense discussion by the joint ACCF/AHA guidelines committee.
Some defenders of perioperative beta-blockade have argued that although initiating beta-blockers on the same day as surgery may be dangerous, a slow and gradual introduction of the drugs for as long as 30 days before surgery may be beneficial. However, according to Francis, this strategy has never been tested in a clinical trial outside of the DECREASE family. In addition, even if this strategy proved safe, it would likely not be feasible in most cases, since perioperative beta-blockade falls under the purview of the anesthesiologist, who generally first sees the patient on the day of the procedure.