Is This Science or Censorship? Statin guidelines, heart disease, and threats to evidence-based publishing

P20_briefing_main_268646kThere are two parts to this story. The first involves a scientific debate about new guidelines for prescribing cholesterol-lowering drugs. The second is about threats to that scientific debate, as an industry heavyweight in support of the guidelines holds court in mass media to demand retraction of two research articles (published in the international medical journal, The BMJ) that argue for a more cautious approach.

Part One: New Statin Guidelines In Question

New guidelines for prescribing statins (i.e., medications like Lipitor, Crestor, Levacor, Zocor, Pravachol, and others) to help lower cholesterol and potentially prevent stroke and cardiovascular disease are one of many areas of contention among biomedical and other experts. In November 2013, the American College of Cardiologists (ACC) joined with the American Heart Association to disseminate new clinical practice guidelines identifying groups that would benefit from taking cholesterol-controlling drugs. At the heart of the problem is what constitutes “normal” versus “high risk” for heart disease.

The new guidelines recommend a statin for everyone with a 10-year risk of heart attack or stroke of 7.5 percent or higher, as determined by a new risk calculator. The calculator estimates risk based on comparisons to pooled data drawn from multiple community-based populations and classified according to sex, age, race, total and HDL-cholesterol, systolic blood pressure, whether a person is being treated for high blood pressure, has diabetes, or is a smoker. All else equal, age appears to be the most influential risk factor to make a person in good health eligible for a statin prescription. Nearly 80 percent of people aged 60 to 75, according to the new calculations, would be advised to take statins.

From what I’ve gleaned from the critiques of the new guidelines, there are four points of contention. First, extending statins to larger portions of healthy populations puts more people at risk for adverse side effects such as nausea, diarrhea or constipation, muscle or joint aches, memory loss or other neurological side effects, Type 2 diabetes, muscle damage, and liver damage. Such problems are documented in studies, but the rate of adverse events from statins is unclear. And we cannot forget the hundreds of deaths resulting from the use of cerivastatin, withdrawn from the market after causing lethal kidney failure. More research is needed so clinicians and patients alike will be able to evaluate potential risks and benefits more accurately.

Second, the risk calculations are based on data mostly from the 1990s that do not reflect the current lower rates of strokes and heart attacks. This fundamental error contributes to an overestimation of risk by 75 to 150 percent, depending on the population. The overestimation of risk extends statin eligibility to nearly 13 million more Americans who are currently at low risk of heart attacks and strokes. Although a 10-year risk of 7.5 percent should not mean an automatic statin prescription, as clinicians are advised to take other factors into account, researchers from the Center for Cardiovascular Disease Prevention at Brigham and Women’s Hospital cautioned in The Lancet that, “reliance on the new risk prediction algorithm could put many primary prevention patients on statin therapy where there is little trial evidence, while potentially denying statin therapy to other patients despite trial evidence of efficacy.” Whether statins actually reduce the overall risk of death or serious illness in people with no personal history of heart attacks and no known risks – a crucial insight — remains in question.

Third, the ACC put the risk calculator online when the committee knew it had problems. The co-chairman of the guidelines’ risk assessment working group, Dr. David Goff, Dean of the University of Colorado School of Public Health admitted that the group recognized the “potential for overestimates, especially at the high end of risk.” But when researchers evaluating the calculator raised concerns about the data and overestimation of risk, Sidney Smith, of the University of North Carolina Chapel Hill who led development of the guidelines said, “We intend to move forward with the implementation of these guidelines. It’s really important not to raise unnecessary alarm.” Moving forward with a flawed calculation of risk does not engender confidence even if the risk calculator is only one element of the new guidelines. It is even more alarming to consider that millions of healthy people, based on outdated statistics and with little evidence of benefits, would be advised to take pills with quite a range of adverse effects.

Fourth, 6 of the 15 panelists authoring the guidelines have recent or current ties to pharmaceutical companies that sell or are developing cholesterol-lowering drugs. While there is no direct evidence that these ties influenced the panel’s recommendations such conflicts of interest cannot be ruled out when the pharmaceutical industry is a known catalyst for increased medicalization and intensive drug marketing. According to the Centers for Disease Control, spending for prescription drugs in the United States was $234.1 billion in 2008, more than double 1999 expenditures. Almost half of Americans use one or more prescription drugs, and 10 percent use more than five. Margaret McCartney, a general practitioner in Glasgow, points out that, “improving population health should not simply be made the work of drug companies.”

Each of these issues warrants investigation. Independent scientific research should zero in on what the body of research says about the benefits and harms of statins, how these play out for people at varied degrees of risk, and what else we still need to know to clarify these issues for research, clinicians, and patients. The American College of Cardiologists and the American Heart Association have a clear responsibility to ensure that their guidelines are based on the best available evidence, to seek out evidence if it is not available, and to update the guidelines to reflect new data. Any potential conflicts of interest among those funding research, analyzing it, disseminating it, and profiting from its findings must also be disclosed and balanced against data transparency and thoughtful skepticism.

Part Two: Threatening Scientific Debate

Forbes called it a kerfuffle between an editor and a prominent researcher. The editor, Fiona Godlee: editor in chief of one of the most respected medical journals, The BMJ. The researcher: Rory Collins, head of the Cholesterol Treatment Trialists’ Collaboration, one of the groups responsible for the new statin guidelines. The challenge: Rory Collins’ insistence that The BMJ retract two articles published in 2013.

The first BMJ article, by John Abramson and colleagues, reanalyzed data from the CTT Collaboration that Collins directs and questioned the evidence behind proposals to extend the use of statins to people at low risk of cardiovascular disease. The second article, by Aseem Malhotra, questioned the role of saturated fat in heart disease. Both papers cited a study by Huabing Zhang and colleagues that found that 17.4 percent of the more than 107 thousand patients in their study had “statin-related events” (meaning adverse drug effects). The two BMJ articles reported this statistic incorrectly as 18-20 percent, as a single statement (one sentence) across two papers. This is a small, important error to correct. But it did not invalidate the papers’ analyses or conclusions.

When the error was reported, The BMJ and the authors published a correction, saying that the statements were removed from the papers and that although the citation was “based on statements in the referenced observational study by Zhang and colleagues, that ‘the rate of reported statin-related events to statins was nearly 18 percent,’ the article(s) did not reflect necessary caveats.” The BMJ editor Godlee then reiterated this point in a detailed editorial so that readers, the media, and the public who may benefit from statins would not be wrongly deterred because of concerns over side effects.

Despite the correction, Rory Collins still wants both papers to be retracted. After contacting The BMJ about it, he resorted to fear mongering in the media. Collins told the BBC that the (now-corrected) statement “put lives at risk” because it “overestimated the side effects of statins by more than 20 times, when large trials show that they are safe, well-tolerated, and very effective in preventing heart attacks and strokes.”

As stated previously, the body of evidence shows that statins tend to work well for high-risk people (i.e., those who have already had a cardiac event) in preventing further cardiovascular events, but statins cannot be recommended simply on the basis of high risk alone without regard for underlying clinical conditions. Further, there is no good evidence that statins work well as primary prevention in healthy people, and the frequency of side effects for both groups is unclear.

Collins ignores these important points in an interview with The Guardian, saying instead that, “Doctors worrying about the safety of cholesterol-reducing statins are creating a misleading level of uncertainty that could lead to the loss of lives.” The frightening accusation that people will die is a common tactic when scientific analysis potentially invalidates accepted protocols (or new ones being heavily promoted by industry).

The BMJ Editor Godlee has asked an independent panel to decide whether the error published in the two BMJ articles (now corrected, and secondary to the focus of the papers) is sufficient to warrant retraction – a fate reserved for cases of fraud, plagiarism, misconduct, or clearly unreliable findings. The panel will review and comment on the process by which the articles were published, how criticisms were raised and the journal responded, and whether either or both articles should be retracted. In the spirit of transparency, The BMJ will publish the findings as well as all of the documents the panel uses for its deliberations. The deadline for the report is July 31, 2014.

None of the push-pull about statins should surprise anyone who has noticed the abundance of pharmaceutical advertising. But public health must rely on scientific debate that involves healthy skepticism, data transparency, evidence based publishing, and ethical treatment rather than censorship and fear mongering in the public square.

First published on Psychology Today on May 30, 2014.


Be Sociable, Share!

Articles & Posts