An Over the Top Denunciation of the "Comparative Effectiveness Crowd"

Many have raised concerns that physicians too often favor the latest, most expensive, most high-technology, and sometimes most invasive treatments, contributing to the high costs and often bad outcomes of health care. One solution would be better research comparing the various possible treatments for particular problems. We recently posted about the controversy over such comparative effectiveness research, particularly Dr Scott Gottlieb's reasonable concerns with the approach, but failure to acknowledge similar and worse problems affecting the alternatives to comparative effectiveness research.

In a new commentary in the Washington Times, Robert Goldberg, "vice president of the Center for Medicine in the Public Interest," went over the top in his denunciation of what he called "the comparative effectiveness crowd."

The cost or comparative effectiveness crowd claims that, opposed to drug companies, it is guided purely by scientific and economic facts about which medicines are more effective for the money. And they only have our better health in mind — no matter what it costs.

Except that what guides these decisions smells more like bias or fear, not science. And both can kill.

His first objections to comparative effectiveness came from a set of personal anecdotes. The first one was

A couple of years ago my daughter was being treated for an eating disorder that required the use of Abilify, an atypical anti-psychotic. Public Citizen calls it a 'worst pill' that should not be on the market at all. Consumers Union calls a 30-year-old drug that causes cognitive impairment a 'Best Buy.'

The second one was

When my cholesterol nearly reached Barry Bonds levels, I went initially on a generic drug that was a Consumers Union best buy. My legs and arms cramped up in a week and my cholesterol levels were unchanged. I was gene tested and found that lovastatin, simvastin and Lipitor would cause muscle pain but that Crestor would not. The comparative effectiveness crowd hates gene testing because it identifies individual differences that screw up their ability to make sweeping judgments about what's best for everyone. Crestor is not a best buy and Public Citizen tells people not to use it.

Obviously, neither anecdote provided any justification for Goldberg's assertion that comparative effectiveness research "can kill."

Beyond that, I was at first bewildered by these anecdotes, which seemed unrelated to any actual comparative effectiveness research. Goldberg did not cite and I am not otherwise aware of any comparative effectiveness research relevant to the treatment of the clinical problems he mentions, the use of atypical anti-psychotics for eating disorders, and genetic testing to identify susceptibility to adverse effects of statin drugs for increased cholesterol.

But I then realized that these anecdotes were about Goldberg's apparent grudge against Public Citizen's Worst Pills web-site and Consumers Union's Best Buy Drugs site.

Regardless of what one thinks about these two sites, however, equating them to some "comparative effectiveness crowd" makes little sense. There are lots of people, including yours truly, who believe that comparative effectiveness research has some merit, but have no affiliation with either of these organizations.

But it gets worse. Goldberg then criticized what the "comparative effectiveness people" had to say about hypertension. His first point was about the fixed-combination drug Bildil (isosorbide dinitrate plus hydralazine, made by Nitromed.)

Black Americans have an 80 percent higher death rate associated with strokes due to hypertension than the general population. For years cardiologists have sought to find — in the absence of fully understanding the complex interaction of genetic, lifestyle and clinical factors — a treatment that could cut into the death rate. So the Association of Black Cardiologists cosponsored the research on a drug called Bidil. The FDA initially rejected the drug as a blood pressure treatment when it first tested on everyone. But sub-population and the sort of genetic analysis the comparative effectiveness people deride as "surrogate endpoints," suggested that it worked to reduce death in blacks. It retested the drug against placebo on blacks and demonstrated a 43 percent reduction in death.

Jerry Avorn, a Harvard Medical School professor, who is a high priest of comparative effectiveness research, when he isn't testifying on behalf of trial lawyers who sue drug companies for a living, derided the results. He called FDA regulation a "minimal standard that would be unacceptable anywhere else in research." He claimed since Bidil wasn't tested against other drugs the results were meaningless.

There is one glaring problem with this argument. Bildil was not tested as a treatment for hypertension. The recent controversial research on Bildil assessed it as a treatment for congestive heart failure (CHF), not hypertension. [Taylor AL, Ziesche S, Yancey C et al. Combination of isosorbide dinitrate and hydralazine in blacks with heart failure. N Eng J Med 2004; 351: 2049-2057. Link here.] That study was indeed performed by people in the Association of Black Cardiologists, and it found that Bildil reduced the death rate for patients with congestive heart failure by 43% compared to placebo. But congestive heart failure and hypertension are not the same (although admittedly hypertension is one of several possible causes of CHF). The Taylor et al study has nothing to do with Bildil as a treatment for hypertension. (See our blog post here on the Bildil controversy.) For that reason, the first paragraph by Goldberg above makes no sense. Thus, the remainder of his disagreement with Professor Avorn on this issue also makes no sense.

Then Goldberg addressed the ALLHAT trial, which really was at least close to a comparative effectiveness study of some commonly used anti-hypertensive drugs. Goldberg described part of its methodology thus,

That five-year, 42,000-patient trial studied the comparative effectiveness of older blood-pressure drugs (diuretics) against newer medicines in reducing heart attacks. Patients were allowed to switch to a combination of drugs only if they 'failed' on the older medicine first. 'Failed' is another word for having a stroke or heart attack.

That would have been a bizarre way to run a study. But it was not how ALLHAT was conducted. Goldberg's description of the ALLHAT methodology also appears to be nonsense.

In fact, patients whose blood pressure was not controlled on the assigned drug were given additional drugs.

The treatment goal ... was a DBP [diastolic blood pressure] of less than 90 mm Hg and an SBP [systolic blood pressure] of less than 140 mm Hg.

If participants did not meet the BP goal while taking the maximum tolerated dosage of the initial medication, an open-label Step 2 agent ... or an open-label Step 3 agent ... could be added.

[From The ALLHAT Officers and Coordinators for the ALLHAT Collaborative Research Group. Major cardiovascular events in hypertensive patients randomized to doxazosin vs chlorthalidone: the Antihypertensive and Lipid-Lowering Treatment to prevent Heart Attack Trial (ALLHAT). JAMA 2000; 283: 1967-1975. Link here.]

Furthermore, the study drug could be discontinued for a variety of reasons, not just stroke or heart attack. These included refusal of the patient to take the drug, a "symptomatic adverse event," excess blood pressure, too low blood pressure, and "other." [From Figure 1 in Wright JT, Dunn JK, Cutler JA et al. Outcomes in hypertensive black and nonblack pateitns treated with chlortalidone, amlodipine, and lisinopril. JAMA 2005; 293: 1595-1608. Link here.]

Goldberg's description of the results of ALLHAT was,

ALLHAT concluded that diuretics were more cost-effective in reducing death from all forms of heart failure, (not heart attacks) and were therefore cost-effective overall.

That was wrong too. The main result of ALLHAT appeared to be that the two newer, more expensive drugs (amlodipine and lisinopril) did not reduce the rate of any of the bad outcomes of hypertension measured, death, coronary heart disease, congestive heart failure, stroke, or renal failure, compared to the older drug, the diuretic chlorthalidone, in either black or non-black patients.

Goldberg's final criticism of Avorn was,

So, Avorn opposed approval of Bidil, which reduced deaths from stroke among blacks by 43 percent but is making money peddling the ALLHAT regimen to doctors even though it increases stroke related death among blacks by 40 percent. Can you say Tuskegee study?

This was, at best, a total non sequitur. Bildil, again, was not tested or advocated as a treatment for hypertension. It was a treatment for congestive heart failure. Furthermore, black patients given chlorthalidone had statistically significantly fewer strokes than those given lisinopril, but slightly (but not significantly) more strokes than those given amlodipine. Black patients and white patients given chlorthalidone had less CHF than black patients given either of the two other drugs. [From Wright JT et al.]

So Goldberg's attempt to link Avorn with the Tuskeegee experiment is, at best, further bizarre nonsense.

His final rant was really over the top, and was totally unsupported by any of his previous arguments.

Comparative effectiveness has nothing to do with patient wellbeing or better evidence. It's about hating drug companies and know-it-alls claiming they know better. Such studies are rigged against new drugs and designed to eliminate the individual differences that are at the heart of the next generation of personalized medicines. Drugs that Avorn, ALLHAT and Public Citizen claim are cost-effective could be hazardous to your health. Want to feel better or stay alive? Take a medicine advertised on TV instead.

We have noted before that Goldberg works for an organization, the Center for Medicine in the Public Interest, whose president is Senior Vice President for Global Health Affairs at the big public relations firm Manning, Selvage and Lee. Manning, Selvege and Lee has many big pharmaceutical accounts, as listed on the CommuniqueLive.com site. This suggests at least an indirect conflict of interest, favoring the interests of the pharmaceutical industry, that should have been revealed in this op-ed article.

But my main complaint about Goldberg's article was not that it amounted to stealth health policy advocacy. It was non-stealthy health policy nonsense. The article was riddled with gross errors about how the results of particular clinical research studies, how they were done, and the context in which they should be interpreted. Most of its arguments amounted to nothing more than, to be charitable, a series of non sequiturs.

It sadly appears that comparative effectiveness research now arouses such passion in the supporters of the pharmaceutical industry as to render some incapable of coherent argument. Why respectable main stream media outlets would publish the resulting gibberish is hard to tell.

Hat tip to Kaiser Daily Health Policy Report.

ADDENDUM (27 September, 2007) - Robert Goldberg left a comment below, which in my humble opinion, does not do much to support his position. In case there is any doubt that he was the actual author, he posted the same thing on the DrugWonks blog. My full reply is here.