The beginning of the new year will mark the 15th anniversary of Science-Based Medicine as a blog. In that time we have published 4,328 articles covering just about every topic in the science of medicine. Hopefully we have at least added a useful voice to the much needed conversation surrounding how best to apply the latest science to the practice of medicine.
It is, of course, very difficult to assess the impact of SBM. There is no way to know what the world would have been like had we not engaged in this project. There is no controlled experiment. We must make due with anecdotal evidence. We frequently hear from practitioners who find our articles and our perspective useful in their practice, even changing how they practice medicine. We are often a resource for the media who would otherwise have only promotional or gullible voices on topics involving pseudoscience in medicine. And we hear from many health care consumers who were able to make a much more informed decision before turning to “alternative” interventions. I like to think we even influenced the FDA and FTC when they were reconsidering their regulation of homeopathy. We also engaged with the NIH regarding their funding of pseudoscientific research.
I wish we had accomplished much more, but promoting SBM has been a consistently uphill battle. One of our unmet goals is to essentially evolve evidence-based medicine (EBM) as the standard within medical academia into something more closely resembling SBM. Perhaps the most frequent question we get here is “What’s the difference?” I’m always happy to answer this question because it is at the core of what we do.
EBM understandably focuses on clinical evidence to answer clinical questions – but it does so with too narrow a focus. In practice EBM tends to ignore prior probability, plausibility, or face validity (“Does the claim make basic scientific sense?”). It also inadequately considers the overall pattern of evidence in the research literature (although this has significantly improved over the last decade and a half). An interaction I had with a respected researcher is emblematic of the problem. They explicitly endorsed not considering prior probability when addressing medical scientific questions, because often what we think is plausible or true isn’t, and we need to rely on experimental evidence regardless of how plausible we think something is.
This approach makes sense, but only within a narrow range of plausibility. The problem is that this approach assumes a few things that are demonstrably not true – that everyone is playing fair, that all proposed hypotheses do not violate the basic laws of physics or everything we know about biology, and that clinical evidence is ironclad. The reality is that we are now living in a world where demonstrable pseudoscience has infiltrated all the institutions of medicine, where there are players that are actively trying to change the rules of good science to promote their beliefs or products, and we now have a great deal of information about the many layers of complexity of clinical evidence. We also know specifically how prior probability affects the statistical analysis of clinical evidence – it is literally necessary to interpret clinical evidence.
Further, those who are promoting pseudoscience in medicine have become expert at exploiting these weaknesses in EBM to slip their nonsense through the cracks. SBM seeks to close those cracks. There are countless examples I could give, but I will simply refer to the most recent example I wrote about here, a recent systematic review of acupuncture in back pain during pregnancy. The authors followed the rules of EBM (although honestly, not well) in order to argue that acupuncture essentially works for low back pain in pregnancy, but that further study was needed. Meanwhile, when I did a more SBM review of the study I think I convincingly argued that it shows acupuncture does not work for back pain in pregnancy. At the very least the data presented does not support the conclusion that it does work (just read the original article to see why). The difference in the ultimate conclusion is stark, and it showcases the difference between SBM and the ways that EBM can be exploited to promote a worthless treatment based on unscientific notions.
Where do we go from here? That’s always a moving target, but generally we would like to continue to grow the SBM community. We need more practitioners, researchers, academics, and regulators who truly understand the principles of SBM and how EBM is currently being abused to blur the lines between science and pseudoscience in medicine. Also, while we have had many authors on SBM over the years, we rely on a core of dedicated SBM authors and editors, and it would be helpful to expand this group.
Meanwhile, we will keep plugging away, trying to make medicine and the world in general a little more scientific, one article at a time.