Best of the Web
- Gregg Backs Medicare Efficiencies in Reform Bill; Says They Would Cut $3 Trillion over 20 years; Should Be Used to Balance Budget, Not Cover Uninsured – Health Affairs Blog (3/15/10)
- Steps to Stop Heart Attacks in Diabetics Fail, Studies Say - NYTimes.com (3/15/10)
- Third-Party Reviews of Devices Come Under Scrutiny at the FDA - WSJ.com (3/15/10)
- Conflict-of-Interest Policies: A Detailed Look - MedPage Today (3/15/10)
- As universities tighten ethics policies, drug firms turn to private physicians to promote products - JSOnline (3/15/10)
- Obama: Right on Substance, Risky on Politics - Joseph White in TheFiscalTimes.com (3/15/10)
- Romneycare Sure Looks Like Obamacare - FrumForum (3/15/10)
- Can Incremental Health Reform Provide A Path Forward? - Gail Wilensky on KHN (3/15/10)
- Health care bill still lacks House votes, congressman says - The Boston Globe (3/15/10)
- Life at All Costs: Dying Abroad - TheFiscalTimes.com (3/15/10)
Recent Articles
Is It Really the End of This Bloody Mess?
I suspect the trial lawyers will still have something to say about that.
A large clinical trial published Thursday in the New England Journal of Medicine confirmed that Trasylol (aprotinin), an anti-bleeding medication designed to reduce major bleeding during heart surgery, significantly increased the risk of death compared to two lysine analogues, which are much cheaper alternatives. The trial was stopped last October when it became apparent the drug was killing some people unnecessarily.
Patients on the slab for heart surgery were given aprotinin to reduce the need for blood transfusions. But, as this latest trial (called BART) found, they had a 50 percent higher risk of death in the first month after the operation. In an accompanying editorial, Wayne Ray and C. Michael Stein concluded that "in all likelihood, this is the end of the aprotinin story."
If so, this final chapter closes the book on a story largely written through the courageous and pioneering work of Dennis Mangano of the California-based Ischemia Research and Foundation, whose data-mining research and epidemiological review of outcomes among post-operative patients first brought the dangers of this drug to light. His story was featured on 60 Minutes earlier this year.
(And for some more background on how Bayer HealthCare, the drug's manufacturer misled the FDA, see this GoozNews post and this New York Times story.)
In the wake of this latest trial results' publication, the FDA announced that Bayer had notifed the agency that it would begin removing the remaining aprotinin stock from the U.S. market. Currently, the FDA limits access to aprotinin to investigational use according to procedures described in a special treatment protocol. The FDA also said that it is reviewing the BART data to reassess the appropriateness of this special treatment protocol.
In the editorial, Drs. Ray and Stein suggested that the aprotinin controversy lasted as long as it did because other trials of the drug tested it against a placebo instead of other available drugs. They also pointed out that although blood loss during surgery is of undisputed clinical importance, it is a surrogate endpoint.
They stated that, as with many other drugs, evaluation of anti-bleeding medications must include effects on death and other clinical endpoints, and not rely on surrogate endpoints (for more on that issue, see this GoozNews post.
Meanwhile, the results of the POISE trial, published online May 12 in The Lancet, call into question current guidelines that recommend use of beta blockers preoperatively in patients undergoing noncardiac surgery. These guidelines were based on previous studies that were small and showed conflicting results, but nonetheless adherence to the guidelines is used in quality assessments and hospital rankings.
In POISE, 8351 patients with cardiovascular disease or at high risk for cardiovascular disease were randomized to Toprol (metoprolol), a beta blocker, or matching placebo, given 2-4 hours before surgery and continued for 30 days. The patients who received metoprolol had fewer cardiovascular events but more deaths and strokes than the placebo group (all-cause mortality 3.1 percent vs. 2.3 percent, stroke rate 1 percent vs. 0.5 percent).
"What POISE says is that in the dosing we used, we see beta blockers have substantial risk in the perioperative setting," lead researcher Dr. Philip Devereaux told Heartwire, "and until someone demonstrates with a clear and large randomized controlled trial that an alternative dose is both effective and safe, it's just not rational, not in people's best interests, to be assuming -- that's how we got into this trouble in the first place."
He points out that if even 10 percent of physicians followed the guidelines, the POISE results mean that in the past decade 800,000 people have died and 500,000 have had a major stroke because they were given preoperative beta blockers.
In the cases of aprotinin and beta blockers in noncardiac surgery, many patients died or suffered injury because a drug or class of drugs were used without sufficient evidence, in the form of large randomized trials, showing that the benefits outweighed the risks that didn't show up in the limited clinical trials given to the FDA at the time of the drugs' approvals. Physicians need to weigh the strength of the available evidence carefully before adopting the latest new drug or treatment strategy.
-- by PM
- gooznews's blog
- Login or register to post comments
Comments
- MKirschMD commented on Unintended Consequences
- AnnieAppleseed commented on The Mandate Opt-Out Safety Valve
- MKirschMD commented on 'Acute Cost Bubble' from Reform's Medicaid Expansion
- Kate Murphy commented on Medicare, Medicaid, Private Insurance and Economic Logic
- petico commented on The Oxymoronic Columnist
Comments
This article brings up some questions regarding the increasing reliance on the recommendations of guidelines committees. Individual physicians may be discouraged from exercising their own judgement and evaluation of available evidence in considering the potential risks and benefits of a specific intervention. This is particularly the case when adherence to guidelines is used as a basis for assessing quality of care; that there should already be any liability attached to not following guidelines recommendations is not without danger, as this could be extended so as to make adherence to the guidelines almost coercive.
Controlled randomized clinical trials remain the most appropriate ways to resolve clinical uncertainty. The profusion of guidelines committees as a source of clinical recommendations can be seen as an indication, not only of a drift away from conducting well designed randomized trials, but also as a disincentive to physicians to exercise their independent judgement. Admittedly it costs far less to convene a guidelines committee than to conduct a large clinical trial, but this saving is clearly made at the expense of the public interest.
None of the above indicates that expert recommendations cannot be valuable. Of course there are instances when they are necessary. But they should never be allowed to replace appropriate trials as a guide to clinical decision making. Admittedly guidelines committees do consider clinical trials in formulating their recommendations, but these are not infrequently those that do not address the issue under question directly, or if they do so, it is with insufficient power. Perhaps an example from HIV medicine will illustrate this. When best to start antiretroviral therapy in the course of HIV disease seems to be known by various guidelines committees, but not from information provided by appropriately designed and powered trials that directly address this issue. By now, such trials could have been completed and a general answer known.But no appropriately powered trial has addressed this important question.
I believe there is now some tension between looking to guidelines committees and conducting appropriate controlled clinical trials in moving forward to best serve our patients. Maybe now is the time to address the general issue of how we now choose to deal with clinical uncertainty, and consider the means to do this that best serve the public interest, even if this should involve expenditure on clinical trials. In the process the role and relevance of guidelines committees should be clarified. The links of many members of guidelines committees to industry is not without relevance to this assessment.
You shouldn't pit guidelines against clinical trials. Good guidelines should be based on large definitive clinical trials, preferably ones that use other available treatments in the control arm. Alas, too many guidelines, written by professional society committees and others who have extensive ties to industry as Dr. Sonnabend points out, are based on weak single-arm or observational studies, or very small controlled clinical trials. That's why we need a comparative effectiveness agency in this country to definitively lay out the evidence, and seek as best as possible to objectively define state-of-the-art care.
Thanks for writing.