Evaluating an article in the Journal of Medical Ethics.

At White Coat Underground, PalMD considers an article from the Journal of Medical Ethics. The article (L. Johnson, R. B. Stricker, "Attorney General forces Infectious Diseases Society of America to redo Lyme guidelines due to flawed development process," Journal of Medical Ethics 2009; 35: 283-288. doi:10.1136/jme.2008.026526) is behind a paywall, but Pal was kind enough to send me a copy.

Pal writes:

I have a strong interest in medical ethics, although I'm not an ethicist myself. Still, I'm generally familiar with the jargon and the writing styles. This piece reads like no ethics article I've ever seen. It is basically an advocacy piece for the concept of chronic Lyme disease, and starts from an entire set of problematic assumptions.

I know very little about Lyme disease, and I have no horse in the race in the controversy about whether chronic Lyme disease exists, if so what it is, and how it might be treated.

I am, however, an ethicist with a strong interest in ethical issues connected to the building and use of scientific knowledge. So I told Pal I'd have a look at the article.

Here's the abstract:

Lyme disease is one of the most controversial illnesses in the history of medicine. In 2006 the Connecticut Attorney General launched an antitrust investigation into the Lyme guidelines development process of the Infectious Diseases Society of America (IDSA). In a recent settlement with IDSA, the Attorney General noted important commercial conflicts of interest and suppression of scientific evidence that had tainted the guidelines process. This paper explores two broad ethical themes the influenced the IDSA investigation. The first is the growing problem of conflicts of interest among guidelines developers, and the second is the increasing centralisation of medical decisions by insurance companies, which use treatment guidelines as a means of controlling the practices of individual doctors and denying treatment for patients. The implications of the first-ever antitrust investigation of medical guidelines and the proposed model to remediate the tainted IDSA guidelines process are also discussed. (283)

From this abstract, I would expect that the authors' strategy is to use a specific case (the antitrust investigation against the IDSA Lyme disease guidelines) to explore conflicts of interest in the medical milieu -- both the potential effects of competing interests on medical experts working to characterize a disease and effective ways to treat it, and the ways the interests of insurance companies might impinge on the interests of physicians and patients. Conflicts of interest and questions of just distribution of medical care and resources are topics of interest for medical ethicists.

It's harder to know what to expect from the discussion of "implications of the first-ever antitrust investigation of medical guidelines". Plausibly, there might be some consideration of whether expert knowledge constitutes a de facto monopoly of sorts, and whether legal remedies for these monopolies might lead to better or worse knowledge-driven medical policies.

From the abstract, we move on to the article itself. Here's what the first page of my copy looks like:

It's about as marked up as draft of a student paper (and indeed, I would need much larger margins to accommodate all the comments I wanted to write). Here is a brief accounting of some of what jumped out at me.

Johnson and Stricker write that:

One group of doctors, the Infectious Disease Society of America (IDSA), issued Lyme treatment guidelines in 2000 that promoted a narrow definition of the disease.¹ These guidelines excluded most Lyme patients and denied them treatment for their illness. The IDSA viewpoint has been adopted without question by government agencies, insurers, medical boards and hospitals. As a result, the diagnosis and treatment of Lyme disease has been suppressed, with profound healthcare implications for patients, their families and their communities. (283)

The claims here strike me as needing, at minimum, some clarification. If IDSA proposed a definition of Lyme disease, then arguably Lyme disease patients are persons whose illness falls within that definition. Of course, Johnson and Stricker have already said that the definition of Lyme disease is controversial; maybe "most Lyme patients" refers to the majority of persons who were being treated for what was identified as Lyme disease prior to the 2000 IDSA guidelines. However, to say that "the diagnosis and treatment of Lyme disease has been suppressed" by the acceptance of these guidelines suggests that Johnson and Stricker are committed to a different definition of Lyme disease than the narrow one in the IDSA guidelines. Otherwise, they would have used more neutral language here, for example:

These guidelines did not include within the accepted definition of Lyme many of the patients who had been identified as suffering from Lyme disease, and accordingly offered no recommendations for the treatment of their illnesses ... As a result of the adoption of the IDSA definition and treatment guidelines, patients who were before identified as Lyme patients have no clear diagnosis or course of treatment, a situation with profound healthcare implications for patients, their families and their communities.

Johnson and Stricker are not neutral with respect to the IDSA definition of Lyme disease. Indeed, footnote 1 cites an article by Johnson and Stricker describing treatment of Lyme disease (presumably treatment that might be ruled out by adherence to IDSA guidelines), and Johnson and Stricker's institutional affiliations are listed as the California Lyme Disease Association and the International Lyme and Associated Diseases Society, respectively.

In examining whether the IDSA acted ethically in promulgating their Lyme disease guidelines, ought Johnson and Stricker to acknowledge as a potential conflict that they are committed to a different definition of Lyme disease?

If someone comes to a conclusion that differs from your own, it is not prima facie evidence that they came to this conclusion improperly. Indeed, one can arrive at incorrect conclusions without behaving unethically.

There still may be grounds for suspecting conflict of interest from members of the IDSA panel. But in an exploration of conflict of interest as an ethical problems, one would expect the authors to disclose their own interests -- financial and non-financial -- in this matter.

More from the article:

When the IDSA sought to revise its Lyme treatment guidelines in 2006, patients and their treating doctors tried to obtain a voice in the guidelines process. The IDSA denied their requests for participation and limited the guidelines panel to researchers known to have a bias against the diagnosis and treatment of Lyme disease. The IDSA panel then selectively ignored evidence of persistent infection or clinical improvement in patients with persistent symptoms of the disease. (283)

Here, there is no substantiation of the claim that the IDSA denied these requests for participation. Nor is evidence presented that the panel was designed to include only researchers "known to have a bias against the diagnosis and treatment of Lyme disease" (presumably on the basis of Johnson and Stricker's preferred definition of Lyme disease rather than the definition from the 2000 IDSA guidelines). Nor, for that matter, do Johnson and Stricker cite the papers that present the "evidence of persistent infection or clinical improvement in patients with persistent symptoms of the disease" that the 2006 IDSA panel purportedly ignored.

If the facts of the case matter to the ethical analyses of it, it is distressing that these facts are so poorly documented, and that the authors don't take care to present fairly the competing sides of the controversy.

[T]he IDSA's ability to further suppress the debate through its role as the gatekeepers (on editorial boards and as peer reviewers) for medical journals, medical conferences and grand rounds supported the antitrust argument when patients started complaining that they were being denied care for their disease. ...

The antitrust theory is that when a medical society has monopoly power such that its guidelines can take on the force of law, as the IDSA guidelines do, it has an obligation to ensure that the development process is fair, does not exclude competing points of view, and that its panel is not driven by commercial conflicts of interest. (283)

I don't think there's much disagreement that medical and scientific decisions should not be driven by commercial conflicts of interest. It's a harder question whether we can tell if someone who has commercial interests that create a potential conflict is in fact driven by those interests (and if so, to what extent).

What I find more interesting, and problematic, here is Johnson and Stricker's unexamined premise that it is a bad thing that medical experts have a certain kind of monopoly. Indeed, their monopoly is recognized by the state: you can't practice medicine without being properly trained and licensed.

Because of their specialized training, physicians and medical scientists have an expertise that arguably puts them in a better position than the state to promulgate disease definitions and treatment recommendations, and to evaluate research reports. Neither lay people nor elected officials can be assumed to have the expertise to make these decisions. Asking the state to make the rules about which submissions to a medical journal ought to be accepted seems like a really bad idea -- so why would asking the state to interfere with the judgment of medical experts in such "gate keeping" seem like a good idea?

It at least requires a positive argument to justify it, one that takes account of the potential benefits and harms of of such intervention compared to the benefits and harms of the present system.

Again, from the article:

The increased concentration and centralisation of power in medicine was also an important factor leading up to the antitrust investigation. Twenty years ago, individual doctors had more autonomy than they do today in deciding whether to comply with medical guidelines. Medical malpractice and unprofessional conduct actions were always rooted in the standard of care, which was determined by the consensus of practising doctors, but guidelines became a vehicle for driving the standard of care. This transformed a very fluid environment where opinions in medical practice changed as individual doctors' viewpoints changed based on new information from the latest medical conference or grand rounds, the success or failure in treating patients or the latest study read by the doctors. The opinions of individual panel members on guidelines panels of powerful medical societies became a proxy for the opinions of doctors in the broader community. With "evidence-based" guidelines, someone other than the treating doctors selects, interprets and makes recommendations regarding treatment. (285)

In this discussion of the autonomy of the doctor, I would have expected to see a more careful discussion of the conditions that allow for a better informed exercise of this autonomy. The doctor's free choice of treatment options, after all, can have profound effects on the patient's health and well being. What kind of obligations does the doctor have to gather reliable information in order to bring about the best chances of a positive outcome?

Similarly, this would have been a good opportunity to consider the obligations taken on by panel members on guidelines panels as they undertake to draw on their expertise to develop recommendations on which other doctors will rely. Surely the weight given the conclusions of these panels would place duties on their members to present as accurately as possible both what is known and what is uncertain.

Johnson and Strickland, however, don't really examine the role of knowledge building in professional and scientific communities. Instead, they seem to treat autonomy as a good that trumps any call for knowledge established by agreed upon methodology and standards of evidence.

The problem arises when the ability of the doctor to select, interpret and exercise clinical judgment is replaced by an obligation to comply with guidelines that do not work or that are not appropriate for the individual patient. Population-based treatment approaches using average results leave individual patients who are the "norm" out of the equation. (285)

To me, there seems to be a tension between this passage on the one I've quoted right before it. What are the doctors who are changing their viewpoints "based on new information from the latest medical conference or grand rounds, the success or failure in treating patients or the latest study read by the doctors" doing if not coming up with something like a population-based approach to disease and its treatment that will inform their interaction with their next patients? Each patient is an individual, but surely the presumption is that doctors start by trying to understand their case in the context of earlier ones -- otherwise each new patient is terra incognita and the doctor must start from scratch.

From an ethical point of view, I think it's important here for the authors to unpack the doctor's "obligation to comply with guidelines". We know that doctors have obligations to their patients, and obligations to other members of their profession, but it seems crazy to argue that they also have a literal obligation to a set of definitions or recommendation on a piece of paper or a web page. Instead, the "obligation to comply with guidelines" must be a recognition of a duty to be guided by the collective judgment of one's professional community, or by the best available knowledge generated by experts in the field, or something along those lines.

Not unpacking this obligation makes it look unnecessarily stupid.

This article does have the occasional glimmer of insight. For example, Johnson and Stricker note that financial interests for physicians are not restricted to interests in the sales of pharmaceuticals but may also include interests in the licensing and sales of vaccines and diagnostic tests, consulting fees from insurance companies, and fees obtained as expert witnesses in litigation. They might even have built on this observation by considering what changes in our medical, scientific, and economic systems would be necessary to insulate those who create expert knowledge from financial interests tied to how that knowledge is used.

However, there are far more instances where they completely ignore obviously relevant ethical questions. An especially striking omission is a consideration of what rights a patient has. Does a patient have a right to have his or her ailment diagnosed -- even if it does not fit the accepted definition of any disease? Does a patient have a right to treatment for his or her ailment -- even if there is no research that supports a safe and effective treatment for it? These are important questions, and the structure of the health care system (including the health insurance industry) in the U.S. surely complicates them. But Johnson and Stricker don't even take these questions up, let alone provide satisfying answers to them.

In short, this was not an exemplary article in medical ethics. It wasn't even an especially coherent one. I hope that the bulk of the articles in the Journal of Medical Ethics rise far above the standard set by this one.