May 03 2010

More Anti-Science at the Huff Po

One of the challenges of trying to be scientific, and an honest intellectual, is that judgment is often required in assessing a claim or topic. The problem with relying upon one’s judgment is that it is fraught, even overwhelmed, with personal bias. The “default mode” of human behavior (which means most people do this most of the time) is to construct an elaborate rationalization for what we already believe, and want to believe. The more intelligent we are, the more sophisticated and elaborate our rationalizations – giving more confidence in our conclusions, but not necessarily deserved.

The solution to this problem is to develop a specific intellectual skill set – knowledge of the many and various ways in which we bias our thinking and the constant application of this knowledge to our own beliefs. In other words, we need to be skeptical, especially of ourselves. But not just skeptical in attitude, systematically skeptical of the process of our own thought. But since this is necessarily self-referential (we can bias our assessment of our biases) it is also necessary to check your beliefs and thinking against other people, people with different perspectives – from different backgrounds, areas of expertise, and cultures.

The opposite of this approach is to be insular, to have a self-contained belief system that feeds on itself but which is completely disconnected from logic and reality. Humans seem to have an unfortunate penchant for falling into such self-contained belief systems, cults being the ultimate expression of this tendency. Conspiracy theories are another manifestation.

One manifestation in particular I want to discuss today is the “demonization” of a person, belief, or system that we don’t like. You can make anything sound sinister and broken, if you exaggerate and emphasize the flaws and weaknesses of a system and ignore or downplay the virtues. For example, if you really wanted to make a case that democracy is broken, you could point out the corruption in the system, the power of lobbyists, the failures of the election system and the electoral college. The 2000 election would stand as a prime example. You could then conclude that we need to reject democracy for something else – anarchy, or perhaps a benign dictator.

This is exactly what Dana Ullman has done to science at the Huffington Post. He essentially says that science is broken and not to be trusted – therefore trust my pseudoscience (homeopathy) and help this new medical science to replace what we have now. It is a remarkable piece of propaganda.

His primary strategy is to point out weaknesses in the practice of science (not the underlying philosophy or methods), as if this is news to anyone. Science is a human endeavor, and is subject to all the frailty of humans. Skeptics and science educators (like Carl Sagan) are the first to point this out. The application of science can fail in a hundred ways, which is why we need to be vigilant, exacting in our evaluations, and rely upon consensus of evidence and consensus of interpretation. Ullman, of course, simply does not like the consensus of logic, evidence, and opinion that homeopathy is pure bunk, and so he rails against the practice of science as if it will save him and his preferred belief.

There are too many examples for me to dissect every one here, so I will give some salient examples. He writes:

Modern medicine uses the double-blind and placebo-controlled trial as the gold standard by which the effectiveness of a treatment is determined. On the surface, this scientific method is very reasonable. However, serious problems in these studies are widely acknowledged by academics but remain unknown to the general public. Fundamental questions about the meaning of the word “efficacy” are rarely raised.

He contradicts himself here – academics acknowledge this problem but it is also rarely raised? This is a case of what we call – making shit up. Ullman clearly has insufficient direct familiarity with clinical research, where fundamental questions about efficacy are constantly raised and discussed. I have actually been involved with designing clinical trials, and a great deal of time, thought, and care has always been given to exactly what we would measure, and what it means. Do we follow a clinical marker of disease, net clinical outcomes, overall quality of life, or all of these things? In most trials multiple endpoints are followed in order to blanket the notion of “efficacy” as much as possible. We want objective measures to know that something biological is happening, clinical measures to know that there is relevant benefit, and quality of life measure to know that patients will actually feel better. We also consider how long we need to follow subjects – are benefits sustained or temporary? Frankly, saying that clinical scientists do not question the notion of efficacy is profoundly ignorant and naive – the kind of thing someone writes when they are trying to build a case against science, not fairly assess it.

Then we discover what Ullman is really getting at.

For instance, just because a drug treatment seems to eliminate a specific symptom does not necessarily mean that it is “effective.” In fact, getting rid of a specific symptom can be the bad news. Aspirin may lower your fever, but physiologists recognize that fever is an important defense of the body in its efforts to fight infection. Sleep-inducing drugs may lead you to fall asleep, but they do not lead to refreshed sleep, and these drugs ultimately tend to aggravate the cycle of insomnia and fatigue, while conveniently (for the drug companies) tend to create addiction. Long-term safety and efficacy of many modern drugs for common ailments remains unknown, despite the high hopes and sincere expectations from the medical community and the rest of us for greater certainty.

Again – more nonsense.  A great deal of ink, as they say, has been spilled exploring all of these questions in the medical literature.  A low grade fever is not a problem, but a high fever can cause damage to the brain, lead to seizures in young children, and should absolutely be treated. Fever is not essential for the body to fight off an infection.

Sleep inducing drugs are tricky to use well. First we need to distinguish sleep onset problems vs sleep maintenance problems. For sleep onset problems a sedative may be effective, and in fact will restore the sleep cycle. But it must be done in conjunction with what we call sleep hygiene – no napping during the day, avoid eating late at night, etc. Sure, some primary care doctors with insufficient training and experience with sleep disorders may just prescribe a sedative. That is a reason to improve education and practice, not chuck the whole system. Because those with more experience and training take a very sophisticated approach to sleep. Start with sleep hygiene. Then treat either sleep onset, maintenance, or both as needed. The goal is to reestablish a normal circadian rhythm and then wean off the medications. Addictive medications are to be avoided – some sleep medications are not addictive, but patients may become dependent upon them if they do not also work on sleep hygiene, or if they have another unidentified sleep problem. So if this approach does not work – then order a sleep study or perhaps consult with a sleep expert.

The point of going over sleep medicine is to contrast the reality to Ullman’s cardboard cartoon. His goal is not to understand or elucidate, but to confuse and poison the well. But also to support an alternative to the scientific view – the homeopathic view that all symptoms are good – they are the body’s way of fighting disease. Treating symptoms is therefore counterproductive. While this may be true is a minority of cases, this is not a general rule of medicine or biology. It is a homeopathic philosophy, nothing else, and is not based upon evidence. Pain does not help the body fight off any ailment that causes pain. Pain only warns us of a problem. So you would not want to numb someone to all pain, because then you cannot follow it as an outcome. But you do want to provide relief to improve quality of life. Pain should be treated – just with care and thought.

Further, there are two kinds of pain: nociceptive and neuropathic. Nociceptive pain is the body sensing tissue damage, and provides a protective and warning function. It should be alleviated, but not masked. Neuropathic pain serves no function, it is a malfunction of the nervous system, caused by trauma, disease, and sometimes even genetic mutations that alter nerve function. With neuropathic pain, the pain is the disease. There are specific medications that may reduce neuropathic pain – not regular pain medication, but neurologically active medication that suppress the abnormal pain generation or pathways.

Homeopaths like Ullman whitewash over all this nuance and complexity. To them, symptoms are good, and all symptomatic treatment is bad. And then they criticize science-based medicine from their own childish perspective.

Later on comes this gem:

Conventional drugs used today are so new that there is very little long-term research on them. There are good reasons why a vast majority of modern drugs used just a couple of decades ago are no longer prescribed: they don’t work as well as previously assumed, and/or they cause more harm than good.

Drugs used today have been available for a variety of time. Some drugs in common use have been used for decades. I routinely prescribe drugs that have been around for 40-50 years – so-called tried and true drugs. There are few surprises in store when using such drugs. Others have been around for 10 or so years – enough time for post-marketing experience to have been tracked and for doctors to have a good fix on long term efficacy and safety.

And sure, there are also new drugs always coming on the market. It is acknowledged that with such drugs we, by necessity, lack long term experience. And so they are used with more caution, while their effects in the real world are tracked and studied. Some drugs ultimately fail at this stage, while others go on to become established drugs.

Again we see the complex reality, and Ullman’s self-serving cartoon. I would also like to hear what an alternative to this system would be – how do we gain decades of experience with large numbers of patients without, you know, using drugs for decades in large numbers of patients. Ullman has his alternative – chuck the whole system and replace it with nonsense and pseudoscience – his particular flavor of pseudoscience. Trust in the gurus, not science.


To a physician, Ullman’s rant is a transparent and naive polemic. But to someone not personally familiar with the intricacies of modern medicine, who cannot put his claims into perspective, his case can seem superficially compelling. Reading his article is like watching Loose Change without any further background knowledge about 9/11. If you pile up a bunch of biased misinformation in one place, the shear size can seem impressive.

Ullman also entirely misses the big picture that medical science is self-critical and self-corrective (unlike, say, homeopathy which is insular and protective). All of the criticism he exploits to attack medicine are self-criticisms that academic physicians and scientists have and continue to raise. We are asking the questions – how long do we need to study a drug before we can confidently extrapolate its long term effects? Then we look back and ask – how is this working? Maybe we need to follow these drugs longer before approval. Maybe we need to follow different outcomes. Maybe we need a different strategy.

In that way medicine continues to scratch and claw forward. It’s a messy process, fraught with error, missteps, even bias and fraud. But just like democracy, it may be a frustratingly flawed system, but it is the best we have. And I would no more want to chuck science-based medicine for the wishful thinking and magic Dana Ullman is selling, than I would want to chuck democracy and accept a dictator.

19 responses so far