Share |





Questioning the Evidence in Evidence-Based Guidelines

Michael Weber, M.D.
July 22, 2004

In a recent article published in the Journal of the American Medical Association (JAMA), two medical economists, Michael Fischer and Jerry Avorn, claimed that if practicing physicians could be persuaded to prescribe drugs for treating hypertension strictly according to recently published guidelines there could be savings of up to $ 1.2 billion in the nation's annual drug bill for treating this condition.

This claim was based on replacing newer, more expensive drugs, which are apparently popular with physicians and patients, with older less expensive products: in particular, generic diuretics. The stimulus for what these authors referred to as "evidence-based prescribing" was a government-sponsored set of guidelines for treating hypertension known as the Seventh Report of the Joint National Committee on Prevention, Detection, Evaluation and Treatment of high blood pressure.

This document, usually referred to as JNC 7, in turn is derived - at least as far as its main drug treatment recommendations are concerned - from a large clinical trial known as ALLHAT that had been sponsored by the same government agency (the National Heart, Lung and Blood Institute, NHLBI) that produced the guidelines.

The fundamental problem that Fischer and Avorn face in making their argument is that it depends largely on JNC 7 deliberations that were political rather than scientific, and upon the heavily criticized conclusions of the ALLHAT study, perhaps one of the most controversial hypertension studies ever published.

A Flawed Study
The ALLHAT study was carried out in 33,000 hypertensive patients. It compared an older drug, the diuretic chlorthalidone, with two newer drugs: an angiotensin converting enzyme (ACE) inhibitor, lisinopril, and the calcium channel blocker, amlodipine. Since all these drugs, if properly used, are similarly effective in reducing blood pressure, the primary question of ALLHAT was: for the same blood pressure effects with each of these drugs, will there be differences among them in their ability to protect people from heart attacks and other major events? When the study was over, its officials claimed that the diuretic, in fact, was superior to the other drugs in preventing some cardiovascular outcomes; and since it was also cheaper, it should be the preferred first-line treatment for hypertension.

Fischer and Avorn, in summarizing the top-line results of ALLHAT, unwittingly identified - as the first major result of the trial - one of its most serious blemishes: a design flaw that caused significantly unequal blood pressure effects among the three drug groups that invalidated any comparative claims about their effects on clinical outcomes. Because the study was biased in such a way as to produce greater blood pressure reductions in patients getting the diuretic, claims that this drug might be better in preventing stroke or heart failure - quite apart from serious doubts in this low-budget study concerning the accuracy of major diagnoses - could not be validly considered.

A New Question Arises
More to the point, despite the blood pressure inequality that favored the diuretic, the primary outcome of the study - fatal and non-fatal heart attacks - was virtually identical among the three treatment groups, leading to a further obvious question: if the blood pressure effects had been the same for all drugs, is it possible that patients getting the diuretic might actually have experienced more heart attacks than in the other groups?

The answer came quickly. Only a few weeks after ALLHAT was announced, the New England Journal of Medicine published another hypertension trial known as the Australian Study that addressed this critical issue. In a large group of Australians with hypertension, almost all of whom were white as compared with the more diverse American composition of ALLHAT, a diuretic and an ACE inhibitor had virtually identical blood pressure effects, but the combined rate of death and major cardiovascular events was actually higher in the diuretic group.

Compromised Guidelines
The JNC 7 guidelines, upon which Fischer and Avorn relied so heavily, were created out of a troubling conflict of interest. These guidelines were sponsored by the NHLBI, which not only hand picked the members of the writing committee, but in essence provided its agenda and conclusions. Needless to say, half the members of this committee were also officials of the ALLHAT study and committed to the Agency's pre-determined decision to promote diuretic therapy.

To the credit of the independent members of JNC 7, they argued during the Committee's deliberations that data to support the universal use of diuretics in hypertension did not really exist. As a compromise, the committee came up with the undefined term "most" to describe those patients for whom a diuretic could be appropriate, but insisted on adding the recommendation that other drug classes - including a variety of newer agents - also be considered.

Notably, other guidelines that were published within weeks of the JNC 7 document, gave clearly different recommendations on drug choices despite relying on the very same clinical trials and evidence that had supposedly guided JNC 7. For example, the recommendations of the International Society of Hypertension in Blacks, the principal organization devoted to the care of hypertension in African Americans, argued simply that clinicians should assess the needs of each individual patient and then select those drugs considered to be most appropriate. Likewise, the guidelines of the European Society of Hypertension, based on a particularly detailed and rigorous review of all available data on treatment outcomes in hypertension, advocated the selection of drugs most appropriate to each patient's situation.

Late Breaking News
It is disappointing, of course, that Fischer and Avorn accepted so uncritically the output of ALLHAT and JNC 7. A review of the contemporary medical literature would surely have told them that an international array of experts in hypertension had expressed serious misgivings regarding the validity of ALLHAT and JNC 7.

To be fair, though, Fischer and Avorn cannot be responsible for the fact that they have been blindsided by a very recent re-analysis of ALLHAT, presented as a "late breaking clinical trial," at the 2004 American College of Cardiology annual meeting, by the study's principal investigator, Dr.Curt Furberg. It was now revealed that, in fact, the apparent cardiovascular superiority first reported for the diuretic depended almost entirely upon the differential effects of the blood pressure medications in white and African American patients. Dr. Furburg now conceded that, at least as far as white and other non-black people are concerned, ACE inhibitors and diuretics can be considered as having equal benefits. And, even without Furburg's additional analysis, the calcium channel blocker - apart from the much debated and doubtful diagnosis of heart failure - seemed narrowly to have the best results of all. Other issues regarding the conduct of ALLHAT still remain, but it seems rather awkward for Fischer and Avorn that the ALLHAT conclusions have changed so suddenly.

Are Diuretics Really Cheaper?
In making his announcement, Furburg still claimed that the inexpensive nature of the diuretics continued to make them an attractive choice. But, as yet, there has been no information from ALLHAT to justify such an assertion. It is rather unfortunate that Fischer and Avorn, apparently without being able to fully examine the issue, were forced to accept on faith the unsupported statement about the cost of diuretics.

To be sure, diuretics are the cheapest of the available blood pressure medicines available, typically costing just pennies a day. But this is not necessarily the main part of their overall cost. Diuretics, when used in the type of relatively high doses that were tested in ALLHAT, commonly cause such disturbances as potassium depletion (which can cause danger to the heart), increases in blood uric acid (which can result in gout) and increases in blood glucose (which can lead to diabetes). These potential risks require that patients receiving this type of treatment get regular blood tests to check for these problems, and if they are found, patients will then need appropriate - and, at times, rather costly - remedies to deal with them. All of this, naturally, will also require increased doctor visits at a cost both to the patient and the health care system. So it can be unwise to make assumptions about the cheapness of the diuretics without fully analyzing the overall costs of their use.

What Is the Place of the Diuretics?
Let it be said immediately that diuretics are indispensable for the effective management of hypertension in many patients. When used as low-dose combinations with other blood pressure drugs they are especially useful in enhancing hypertension treatment, and are well tolerated. It is probably true to say that diuretics, prescribed in this logical and appropriate fashion, are being underutilized.

At the same time, this does not justify the primarily economic agenda of Fischer and Avorn to promote first-line full dose diuretic therapy in the majority of hypertensive patients. Quite apart from the questionable evidence used to support the clinical benefits of these drugs, and the absence of reliable data on their real costs, there are other factors to consider. For instance, one of the most dramatic findings of ALLHAT was that diuretics were 43% more likely than ACE inhibitors and 18% more likely than calcium blockers to cause the new onset of diabetes in people with hypertension. And while the original ALLHAT report shrugged off this finding by saying that the appearance of diabetes did not seem to produce adverse cardiovascular consequences, a long-term study in Hypertension, reported in the last few days, has shown that people developing diabetes during their hypertension treatment - typically associated with diuretic use - are almost three times as likely as other patients to get heart disease, strokes or other serious outcomes.

Where Does This Get Us?
When we consider all the discrepancies and uncertainties that underlie the information that Fischer and Avorn used in making their report, it is easy to understand why so many physicians - who carefully scrutinize the medical literature and take a keen interest in new developments - are so skeptical when it comes to accepting guidelines. It would be a real misfortune for our patients with high blood pressure if issues of cost, particularly when obtained so casually, are allowed to get in the way of providing the best treatment. In reality, the marketplace for hypertension drugs is highly competitive, and many of the newly available branded drugs cost very little more than older generics.

More important than government agencies, guidelines committees and even practicing clinicians are the real participants in hypertension: the patients themselves. Data concerning the probability that people with high blood pressure will continue to refill their prescriptions show that for newer drugs there is a better than 60% likelihood that patients will still remain on their treatment after one year. For diuretics, the number is only about 20%! This is an extraordinary example of people voting with their feet, and seems to say a great deal more about trends in treating hypertension than the outputs of academic economists and government officials.

It would be a shame if debatable cost considerations were allowed to drive people into using medicines they don't like, and even worse, into giving up their treatment altogether.

Michael Weber, MD, is professor of medicine and Associate Dean for Research, at the State University of New York, Downstate College of Medicine. He has published numerous research articles in the medical literature and has authored and edited ten books. Together with Dr. Suzanne Oparil, he is responsible for the widely used reference volume Hypertension. He was one of the founders of the American Society of Hypertension and has also served as its president. He is currently chair of the Society's Hypertension Specialists Program. He is also a Fellow of the American College of Physicians, the American College of Cardiology, and the American Heart Association.

home   spotlight   commentary   research   events   news   about   contact   links   archives
Copyright Manhattan Institute for Policy Research
52 Vanderbilt Avenue
New York, NY 10017
(212) 599-7000