Two roads diverged in a wood, and I—
I took the one less traveled by,
And that has made all the difference
— Robert Frost
Lyme politics have been on my mind recently, due to the efforts to reconvene the IDSA panel that was responsible for creating a very restrictive definition of Lyme Disease and it’s treatment. The question I am concerned with in this post is how thinking, caring people could write down something so narrow and incomplete. The divided politics of this situation have never adequately explained to me how we got here.
Now, there is a lot of evidence that the individuals who wrote the IDSA guidelines are biased*, and that is certainly a possible explanation for what happened. In fact, as a result of investigations into this bias, a new panel has been convened. To the dismay of many individuals with Lyme disease, the new panel, although lacking conflicts of interest, still seems unlikely to make major changes to the IDSA treatment guidelines. Yet its web page states that its members are “committed to considering all points of view” and the panel is taking public comments until April 3, 2009.
Let’s for a moment take the optimistic viewpoint that the members of the panel are truly committed to this. Let’s suppose that they chose medicine because they are compassionate individuals who want to help people, that they are smart and capable of understanding scientific literature, and that they lack bias. Why do so many doctors choose to believe in a version of Lyme disease that excludes so many of us?
For one possible answer to this, I ask you to consider the Wikipedia page on Lyme disease. In particular, consider the discussion that the authors of that page, who’s viewpoints cover the entire spectrum regarding Lyme Disease, have had about how to define and treat Lyme Disease. I find this discussion fascinating because it lays out for the reader just how easy it is to take one road instead of another when it comes to Lyme Disease. In particular, Wikipedia has a policy regarding reliable sources (opinions represented must be drawn from credible, published material with authoritative authors), and this policy often comes up in discussion on the Lyme discussion page.
For example, consider the following discussion of the clinical trials involving antibiotics. Before we even look at the discussion, I must point out that only frequency knowledge (statistical data) is being discussed — process knowledge** would require the author to interpret sources, which they avoid to comply with Wikipedia guidelines on No original research. The Wikipedia article does have an (incomplete) section on the pathophysiology of Lyme disease. This discusses some of the process knowledge regarding possible mechanisms by which the spirochete that causes Lyme Disease may persist despite an immune reaction, may cause ongoing symptoms, and so on. None of this information is explored in the discussion of long term antibiotics either in the article or the discussion page.
Instead, the discussion centers around what one phrase would best summarize the results of all the clinical trials — “no [consistent] benefit” “mixed results” (the most generous even discussed) and so on. Eventually the authors settled on the following summary: “Of four randomized controlled trials of long-term antibiotic courses in patients with ongoing symptoms, two found no benefit, and two found inconsistent benefits and significant side effects and risks from further antibiotic treatment.” Later on in the article, the authors give more details on the studies and conclude: “Most medical authorities, including the [IDSA] … have concluded that there is no convincing evidence that Borrelia is implicated in the various syndromes of “chronic Lyme disease”, and recommend against long-term antibiotic treatment as ineffective and possibly harmful.”
To me, this discussion is a mirror of what I fear and expect will happen in the discussions of the IDSA panel. Multiple viewpoints are represented in the discussion. For example, one author says: “in the Krupp study, the significance of intention to treat analysis for benefit to fatigue was p<0.001. And even with all possible combinations of the 7 patients lost to follow-up considered, it remained p<0.05. That is a consistent benefit for fatigue in this study (fatigue was the sole clinical criteria for patient selection).” The argument is civil, and everyone seems to enter it and leave it believing that it is important to maintain a neutral point of view, base all claims on reliable sources, and so on. Yet despite this, the outcome seems almost pre-ordained: The lack of incontrovertible frequency data demonstrating a reliable cure for chronic lyme disease seems to lead to a single conclusion. No matter that the null hypothesis (that antibiotics don’t help) also hasn’t been proven. Somehow, the clinical experience of many doctors with thousands of patients and the science providing process data on how Lyme and its co-infections may persist all don’t seem to matter.
*The attorney general of Connecticut launched an antitrust investigation into the panel which uncovered many conflicts of interest. This eventually led to a new panel being convened, and the panel members were announced last month. Pamela Weintraub speaks in some depth about this history on her blog, Emerging Diseases. She then goes on to point out some of the evidence that supports a broader and more nuanced view of Lyme Disease, evidence that we all hope will be represented in the new guidelines.
** For an explanation of frequency knowledge versus process knowledge, see my prior post on this