Here are a couple of interesting bits. First of all, go to this blog, where Eddy Jenner, a clinician in Australia, blogs about his encounters with EBM at the bedside -- he provides a fascinating, well-reasoned and well-read perspective. His most recent post is illuminating particularly in the context of the current discussion. It reminded me that I am much more interested in general in improving everyone's understanding of how we do clinical research than debunking alternative approaches.
Next, I want to articulate why it is I think that chicken pox and HPV vaccines strike me as being less straight-forward than, say smallpox, polio or pertussis. There are 3 things to remember about medicine as a science:
1. With every intervention's benefit there is also a risk of an adverse event.
2. What we think we know today will be different in a decade.
3. It is the obligation of medicine first to do no harm.
None of the 3 statements is particularly controversial or new, and I think everyone can agree on them. So, what?
Here is the so what. Since what we know today is necessarily incomplete, it is quite probable that there are many risks to our treatments, which today we just do not have the data to understand, but will be known in the future. Alas, we do not have a crystal ball to see what is coming down the pike, so, being circumspect about what we are so sure about today should be the norm. For example, when the risk of serious complications from a disease is extremely high (think smallpox, polio, diarrheal diseases in Africa, etc.), then, if the treatment (vaccine) diminishes this risk substantially (how about no more smallpox?), and there is no immediate reason to think that the risk of that treatment is overwhelming, then the risk-benefit equation is hard to tilt away from the benefit. However, when the risks of the complications or death from a particular disease are not that high (relatively speaking, of course), then one really has to examine the risks of the intervention with a much finer lens. And indeed, here is a study on chicken pox vaccination and deaths that showed that from pre-vaccination period to post-vaccination period deaths declined on average from 145 to 66 per year. So, that's pretty good, if there were absolutely no deaths associated with the vaccine itself. But invariably, there are, and this is less a function of the vaccine safety than it is of the human substrate that is being injected. In fact if you check CDC Wonder's VAERS, you will see that on average there are about 14 annual deaths that may be related to varicella vaccination. Now you may say that the balance of this intervention is in the direction of the benefit, and you will be correct. But the magnitude of this benefit makes me a bit cautious. If we were talking in purely scientific terms, I would worry about the possibility of type I error, where the difference is really there by random chance. But we also know that we are not all rationality and science, and there is a lot of emotion involved in this debate. So, it is easy to see how these numbers lead to a variety of interpretations and opinions. And add to this the idea that it is not inconceivable that some other safety signal may come along in the future, and it becomes patently obvious why the issue is difficult to reduce to either ethics or idiocy when one is really committed to science and its principal nostrum of "first do no harm".