Mar 24 2008

Responding to Dana Ullman

Recently homeopath Dana Ullman has been gracing the comments section of this blog. I always welcome contrary views – it sparks great debate and always seems to deepen my understanding of what advocates are saying on all sides of an issue. Recently he has brought up some points that could use a detailed response, sufficient to fill a separate blog entry, so here it is.

In response to my blog entry on A Golden Age of Quackery and Antiscience, Dana Ullman wrote:

Quackery is commonly defined as the use of unproven methods by practitioners who claim impressive results and who charge a lot of money. When you consider how much of conventional medicine is not evidence-based, and when you consider how much they wear the guise of “science” and how much they charge for their services (and drugs), we ARE living in the golden age of quackery…conventional medical quackery.

This is a common reply by advocates of one or more unscientific medical modalities – the claim that mainstream medicine is also not sufficiently scientific. This usually is presented as a tu quoque logical fallacy – which means “you too” and is the counter of a legitimate criticism with the claim that others (often the accuser) also suffers the same vice. This is irrelevant to the validity of the criticism.

But that aside, let’s examine what Dana Ullman is saying – that modern medicine actually represents a golden age of medical quackery. This is absurd on its face. Medicine over the last several hundred years has been under constant evolution toward a greater founding in science and evidence. Medicine is more scientific and evidence-based now than it has ever been.

It is difficult to estimate the degree to which mainstream medicine is science-based, because this is not a black-and-white issue, but credible estimates are that 78% of what physicians do has a solid grounding in science and evidence (which is much more than the bogus figure of 15% often cited by CAM proponents). I discuss this question in more detail here.

I also think we can (and will) do better than this, because mainstream medicine is based upon a scientific standard of care. Evaluating treatments and diagnostic modalities are built into the system. Medicine is self-critical, as any science should be. Another strength of the system is that there are different specialties. So, for example, the surgeons that perform a procedure may be largely dependent upon referral from a non-surgical specialist who treats the same problems with different modalities. This provides an important cross-check to the system. There are also academics who typically do not make their money from practicing medicine and whose job it is to question and improve the current science and standard of care.

My overall impression (admittedly from the perspective of an academic) is that while there are many specific criticisms one could make, the overall system works and is heavily science-based.

The biggest problem facing the scientific basis of modern medicine today are those who are trying to promote so-called “alternative” modalities that are not science-based. In so doing they are trying to redefine what is science, or to change the rules of scientific medicine to allow for patent nonsense, or to eliminate the science-based standard of care so that anything goes.

It is also critical to point out that while we are talking about the scientific basis of treatments in mainstream medicine, we are generally referring to treatments that are reasonable and make sense when considered in the context of our scientific knowledge of biology. But we also recognize that this is insufficient to conclude that a treatment works, so we also demand high quality clinical evidence of safety and efficacy, but sometimes we need to make treatment decisions when definitive clinical evidence just doesn’t exist. CAM proponents will then compare this situation to theirs, in which they are advocating (as Dana Ullman does) for a treatment modality (in his case homeopathy) that has no scientific plausibility in the first place – not chemical, physical, or biological. The image of splinters and beams come to mind.

Dana Ullman specifically address surgery when he writes:

And don’t forgot those surgeons…as much as I respect their work, let’s not fool ourselves into thinking that it is based on double-blind placebo controlled trials. Oh, I see, this “gold standard” is not applicable to every medical treatment, nor is it or should it necessary be expected from every unconventional medical treatment either.

This is a straw man. I never claimed (nor, to may knowledge, has any serious commenter) that surgical interventions are studied with double-blind placebo controlled trials. This is simply not possible – for one because it is generally considered to be unethical to perform sham surgery; to tell a patient that you have operated on them when all you did was open them up, do nothing, and then close them back up.

What Dana Ullman is doing is implying that if you cannot do a placebo-controlled trial, then you are not scientific, and this is the same as using magic water as a treatment. However, placebo-controlled trials are not the only way to study a treatment. First, let us recognize that often surgeries are based upon a solid understanding of anatomy, physiology, and pathology. Surgery is often done to fix an obvious anatomical problem, or acute problems that have a well-known natural history.

Let us consider, for example, appendectomy. If an appendix becomes seriously infected in can become filled with puss and then eventually burst, spilling bacteria into the abdomen which will cause a widespread infection that is often fatal. Removing the appendix before it bursts is such an obvious intervention one must consider the ethics of withholding this intervention for a study. Also, we can compare outcomes from appendectomy to what we know historically about the natural history of untreated appendicitis. This is scientific evidence.

Surgical interventions can also be compared to best medical management – for example cardiac bypass vs anticholesterol and heart medication. And many surgical procedures are studied without blinding, which can still provide useful information if the outcome measures are very objective (like stroke or death). Again – there simply is no valid comparison between this and using modalities that are based upon pre-scientific notions.

Dana Ullman also wrote:

The “team” on which I play is the Hippocratic team, for I honor his “First, do no harm” dictum. I also play on the evidence-based medicine team, not just evaluating short-term but long-term results. I am critical of the use of polypharmacy unless there is evidence that multiple drug regiments have evidence (99% of the time, they don’t).

The statement about “not just short-term but long-term results,” in the context of this discussion, seems to imply that mainstream medicine focuses on short-term results. This is patently not true. There are many long term follow up studies, and it is generally recognized when this is needed.

But let me address the meat of this statement – that polypharmacy, using multiple drugs at once, is not studied “99%” of the time. I don’t know where he gets this figure, but the figure is not as important as the deep misconceptions behind it. In fact, polypharmacy is part of most phase III clinical trials – it is built into basic trial design.

Here’s how: subjects are randomized to either the treatment or placebo arm of controlled trials. The purpose of randomization is to equalize any potential confounding factors that may affect the outcome. Large numbers of subjects are required for randomization to achieve this. Subjects in these large efficacy trials are real patients who have the disease in question, but also have all the other things that real patients have. They are often taking other drugs for other reasons. So the new drug in question is often studied in combination with other drugs that the subjects happen to be taking. And since the net clinical effect is what is measured, the effects of polypharmacy are taken into consideration in such trials.

It is true that certain patients are excluded from specific trials (all trials have their exclusion criteria). This will often include the absence of a major disease that can affect the outcome being followed. It will also typically exclude any medication that can cause the same effects as the one being studied – but this is not always true. Often the only requirement is that there are no changes to the medications during the trial period.

The bottom line is that new drugs are studied in subjects who are taking other drugs, and this provides information about polypharmacy. Also, a great deal is known about so-called drug-drug interactions. Pharmacists are actually quite good at predicting how drugs will interact, and this is information clinicians must master as well. This information is also part of FDA approval and is documented in the official package insert for all medications.

Finally – specific combinations are studied when there is a reason to believe that the combination will have either a synergistic benefit or might produce a negative effect. (I often combine medications that have complementary mechanisms of action – but then of course you have to know their mechanism of action.) But such combinations are the vast minority of all potential combinations – the permutations of all drugs out there is vast and it is literally impossible to test even a tiny percentage of them.

Therefore it is grossly misleading to say we do not have scientific information about polypharmacy, and the 99% figure is meaningless.

Dana Ullman finishes:

There is a good reason that homeopathic medicines are used by hundreds of millions of people today and that historically, it has been used by many of the most respected physicians, scientists, and cultural heroes of the past 200 years.

Ah. The argument ad populi. The logical fallacies are piling up. Blood letting was used for a couple thousand years as well. Millions of people can get it wrong. To deny this is to be supremely naive about the nature of medicine, the placebo effect, cultural beliefs, and the history of medicine. It is possible for a therapy to be widely popular while being utterly useless. Homeopathy, if anything, is proof of that.

43 responses so far