"Proof" is not what most people think it is

Proof. You keep using that word. I do not think it means what you think it means.

That thought kept running through my mind as I perused an article appearing on an antivaccine website. Another thought that rant through my mind is that this was clearly not a scientist of any sort speaking. In science, in general, we don't speak of "proof." We speak of evidence and experimentation. Lawyers speak of proof, as in "proof beyond a reasonable doubt." Scientists speak of evidence in shades of gray, because most evidence is on a continuum. Besides, the article asked the question, "Has 'science based medicine' proven vaccine-induced autism doesn't exist?" Obviously, Orac can't resist such an article, which is why when he saw an article on the anti-vaccine website Vactruth.com by someone named Jennifer Hutchinson entitled We've Shown Them the Proof it was only a matter of time before the application of a heapin' helpin' of not-so-Respectful Insolence would commence. Hutchinson is apparently the mother of a child named Jake and has written a book entitled Unlocking Jake: The Story of a Rabies Vaccine, Autism & Recovery.

I must admit, this was a new one on me. Of all the vaccines out there, I hadn't heard of the rabies vaccine ever being blamed for autism before, but I guess you learn something new every day. What I didn't know is that you can apparently "recover" a child from allegedly vaccine-induced autism through a combination of treatments including biomedical interventions (which in autism are mostly pure quackery), developmental optometry (a highly dubious set of tests and interventions that includes vision therapy), and that quackery to end all quackeries, homeopathy.

Yes, any time someone touts homeopathy as a treatment for autism, speaking of "proof" is a bit beside the point, don't you think, regardless of what you might think of developmental optometry and biomedical interventions? "Proof" clearly isn't what Hutchinson thinks it is, and unfortunately the rest of her post completely fits the pattern of someone who believes in homeopathy in that it relies almost entirely on anecdotes rather than science, epidemiology, and clinical trials. In fact, it revels in its reliance on anecdotes, as you will see, castigating scientists for not considering them to be valid evidence. I haven't said it in a while; so I'll say it again: The plural of "anecdote" is not "data." In other words, just because there are thousands of poorly documented anecdotes does not magically transform them, by sheer weight of numbers, into valid scientific data. Yet, Hutchinson apparently thinks it does, as she starts by describing her attendance at Jenny McCarthy's infamous "Green Our Vaccines" rally from nearly four years ago:

I remember Jenny saying that she had spoken with more than 50,000 parents, and they all had the same story. Their children were developing normally, had an MMR or a DTaP or some other shot, and then something went wrong.

Which is followed by:

Most of all, I remember Jenny's words: Their proof. Those are powerful words. If you're the parent of a child with autism, you have your proof that vaccines can cause or trigger autism. There's a lot of proof out there. For anyone who is willing to see it. Unfortunately, that doesn't include our government and most of our doctors. I'm not saying they will admit that there could be a problem with vaccines--far too much money tied up in the vaccine program. Way too much to lose. But I have to wonder, what would they consider proof?

Proof. You keep using that word. I do not think it means what you think it means.

Hutchinson then goes on to a false equivalency in terms of evidence. She laments the sentiment expresse in a USA Today article a while back that "Science has spoken when it comes to the theory that some childhood vaccines can cause autism. They don't, the Institute of Medicine concluded three years ago." She then moves on to saying that "thousands of parents of vaccine-injured children have spoken," too, as if their anecdotes constituted reliable scientific evidence. Again, they don't. Not that that stops Hutchinson from laying down this howler that is a profound misinterpretation of what Steve Novella wrote a few years ago:

So what if it's "anecdotal evidence"? Anecdotal evidence can be a start, right? After all, they are personal experiences, which "may be the first indication that there is a meaningful biological effect in play." With proper documentation, they can be useful for "pointing the way to future research." The findings or conclusions should then be "verified by controlled prospective clinical studies." [2]

What could qualify more as a personal experience--and a more reliable one--than a mother who carries her baby inside her body for nine months, gives birth to him, and then watches him around the clock, catering to his every need? A mother who knows what every look and sound mean because she knows her baby better than anyone in the world. She wakes at night, her full breasts leaking, just seconds before the baby begins to cry. Her body is in tune with his body.

Except that a lot of mothers aren't nearly as in tune with their babies as Hutchinson seems to think they are. How many times have we heard the story that the mother is utterly convinced that her child's regression started right after vaccines but then, when her story is more closely examined, that is found not to be the case? In the Autism Omnibus hearings, experts examined videotapes of the child of one of the claimants (Cedillo) and noted clear signs of autism well before the child regressed or had received the vaccines blamed for autistic regression. In Andrew Wakefield's infamous case series from 1998, timelines were--shall we say?--fluid, to say the least.

Hutchinson places too much faith in the observational abilities of human beings. We actually suck at it. Confirmation bias leads us to remember things that fit with our preexisting beliefs and forget or downplay the rest. As Simon and Garfunkel once sang, a man hears what he wants to hear and disregards the rest. We confuse time lines, remembering events at different times from when they actually happened. We miss subtle changes that a trained expert would be able to pick up on. We saw this with Cedillo and in many other cases. Hutchinson discusses a scenario in highly emotional language about a mother observing her child being "sucked into the black hole of autism" (note the imagery; antivaccinationists frequently describe autism in language that suggests their "real" child is "lost" rather than the child right in front of them) and asks rhetorically, "Would this scenario be any different if a doctor or a scientist observed this baby from day one?" and "And what if he had documented it all with journal entries and videos?" The answer is: Almost certainly yes, because most people suck at observation.

In fact, one major reason we need science in the first place is because we as humans suck at observation. Scientists know this. That's why science is a system, a process, that seeks to minimize human bias and human cognitive quirks that lead them to leap to the wrong conclusions from observation. For example, if a person gets better on his own but had started taking something right before that, he will conclude that what made him better was whatever remedy he had started taking. That getting better on one's own, depending on the circumstances, is called regression to the mean, and concluding that the remedy worked even if it was a homeopathic remedy (i.e., water) is due to confusing correlation with causation.

The same is true of vaccines and autism. Children receive a lot of vaccines around the time that autism is commonly diagnosed. By random chance alone, we would expect a significant number of children to "regress" sometime within reasonably close proximity to vaccination (say, two weeks). It's impossible to tell whether such an association might be causative or not examining one child. Population-based studies are required to determine whether regression occurs more frequently in close temporal proximity to vaccination more often than would be expected by random chance alone. Hutchinson uses the example of a peanut allergy as an example of who physicians will listen to parents when it comes to a hypothesis about what is causing their child's health issues. What Hutchinson doesn't consider is that scientists and physicians have listened to parents who thought that vaccines caused their children's autism. Because of parental concerns, scientists have carried out multiple large studies designed to try to detect a correlation between vaccination and autism. Large, well-designed studies (as opposed to the crappy studies presented as "proof" by the antivaccine movement) have failed to find even a whiff of a trace of a hint of a correlation between vaccination and autism. Although it's impossible ever to "prove" (there's that word again!) a negative, the preponderance of evidence is very consistent with the conclusion that it is far more likely than not that vaccines do not cause autism and that, even if they did, their contribution would be incredibly small, given that studies involving huge numbers of children have failed to find even a weak correlation.

Proof. You keep using that word. I do not think it means what you think it means.

Hutchinson concludes with an oldy moldy from the antivaccine movement, likening vaccines to cigarettes causing cancer. She trots out the same old nonsense we've heard from Dr. Jay Gordon before about how doctors used to advertise for cigarettes and dismiss them as being unhealthy but were wrong. So they might be wrong about vaccines, as well, right? At least that's what Hutchinson asks. Here's the problem: All of those things with cigarettes happened before the large epidemiological studies were done to demontrate a link between smoking and lung cancer. Also, there were indications that smoking caused cancer as early as the the late 1930s, but they came from--of all places--Nazi Germany. Then there was a campaign of misinformation and denialism by tobacco companies that contrubted largely to the delays in admitting the deleterious health effects of smoking. In any case, regardless of how long it took to get the evidence or why we might have ignored early evidence, once the studies were done and the conclusion was solid, the Surgeon General warned people about cigarettes. Now here's the difference: The studies have been done in autism and haven't found anything. It's highly unlikely there is anything to find; science has moved on.

People like Hutchinson, unfortunately, have not.