Accurate and Precise
This is the text version of my presentation to the Institute of Medicine Panel today. I delivered my comments remotely, because a fever has kept me bedridden for three days. I tried to speak as naturally and extemporaneously as possible, so there are some differences between the spoken and written versions. I also submitted a much longer version with more details and references to the Public Access Folder. Update January 31, 2014: Video of my talk has been posted.
My name is Jennifer Spotila and I have been disabled by ME/CFS for more than nineteen years. I’ve been involved in advocacy for most of that time, including past service on the CFIDS Association Board, and appointment to the FDA’s Patient Representative Program. I write about politics and other ME/CFS issues at occupycfs.com. My perspective on this IOM study is informed by all of that experience, and I believe the one thing you must keep in mind throughout this study is to be as accurate and precise as you can in order to create sensitive and specific diagnostic criteria.
You are facing an enormous challenge, the result of decades of inaccuracy and imprecision in the definition and diagnosis of our disease.
Multiple names and definitions have been used in the last 30 years, and you should go back as far as the 1950’s to examine the descriptions and definitions of ME. Each definition carries with it a rationale, an associated description of the disease, and a set of limitations. Prevalence rates vary because each definition draws a different circle around – or within – a patient population.
Another challenge is that there are no gold standard biomarkers for the heterogeneous Fukuda population. Despite that, we are very close to establishing one or more diagnostic marker, and a number of you have been responsible for that important research. But the breadth and weaknesses of the case definitions have been a huge obstacle to achieving this.
Finally, as you no doubt realize already, there are competing schools of thought on case definition. Does the mixed bag of definitions describe one disease or more than one disease? How do we identify a more homogenous cohort? What should we call it? Who is competent to diagnose it? We do not agree on the answers to those questions because there are no easy answers.
As in any controversial subject, there are potential pitfalls that could complicate your work.
First and foremost is the fatigue paradigm. Severe fatigue lasting longer than six months is an extremely common symptom experienced by 4-5% of the general population. One study of newly diagnosed MS patients found that nearly 30% had been diagnosed with severe fatigue or CFS in the three years prior to the MS diagnosis. Because it is based on fatigue, the Fukuda definition has been used as a wastebasket for people with unexplained fatigue, consigning them to medical purgatory when they may have more treatable conditions.
It might help to draw a comparison to chronic pain. While pain conditions are researched across diseases to identify common mechanisms and treatments, we do not define and diagnose them that way. We do not diagnose fibromyalgia, vulvodynia, and migraines as one disease that is subtyped based on where the pain is located in the body because chronic pain is simply too common a symptom. I believe the same is true for the fatigue paradigm. Severe fatigue is simply too common a problem to form the basis of an accurate and precise case definition, regardless of how you try to slice it into subtypes.
Second, as Charmian pointed out, it is imperative to recognize which definitions and exclusionary conditions were used in each research study. If you do not take this into account, you will not be able to sort through the evidence with any resulting clarity.
Third, I believe it would be a mistake to lose sight of the impacts your report will have in the real world. As Lori and Pat have said, you will have an impact on research and clinical trials. You will have an impact on policy. You will have an impact on clinical care. The stakes are very high, but there is a solution.
Accurate and Precise
I believe it is possible to create diagnostic criteria that is both sensitive and specific for ME/CFS. To do so, you will need to be as accurate and precise as possible.
Start by setting aside the fatigue umbrella, because severe chronic fatigue is common to many diseases, and should not be used as the foundation of a precise case definition. Instead, focus on the core symptoms of disease. Across multiple studies and surveys, post-exertional malaise – not fatigue – has emerged as the key and most disabling feature of this disease. PEM is an exacerbation of multiple symptoms after mental or physical activity, and can be measured through both self-report instruments and objective biological tests. It is also notable for its use in distinguishing between people with ME/CFS and those with major depressive disorder and other illnesses with a fatigue component. Another core symptom identified in the research is cognitive impairment, particularly memory and concentration problems. Unrefreshing sleep and autonomic symptoms also rise to the top. There are many other symptoms as well, but your diagnostic criteria will be more meaningful if you focus on the core features.
Another approach that will help you succeed is to use frequency and severity thresholds, in addition to a list of symptoms, to identify a more precisely defined patient population. In one study, 34% of healthy controls reported at least 4 out of 8 Fukuda secondary symptoms. When higher cutoff thresholds for frequency and severity measures were used, that number dropped to 5%. If you combine a core symptom set with thresholds for frequency and severity of those symptoms, I believe you will be able to establish accurate diagnostic criteria that have high sensitivity and specificity.
Precision is the most important tool at your disposal, and yet it will not remove all ambiguity and uncertainty because there is so much we do not know about this disease. Three hundred years ago, cancer could only be diagnosed when it advanced to externally palpable tumors. The definition of cancer has evolved from the most severe and easily detected form of the disease to the sophisticated detection and subtyping methods of today.
Your case definition is part of an iterative process, like all case definitions. Most criteria broaden over time, such as the 2011 revision of the diagnostic criteria for Alzheimer’s Disease. But that only happens after diagnostic criteria have been tested and validated on the more severe forms of disease. Precision based on what we know, is the first step.
In summary, you face the enormous challenge of creating diagnostic criteria for a disease with conflicting case definitions, no gold standard biomarkers, and competing views about how to define the problem. You must avoid multiple pitfalls, including the high prevalence of the symptom of chronic fatigue in the general population and the effects of those competing case definitions on research results, and never lose sight of the very high stakes we face. But there is a way for you to overcome all of these challenges, and that is by being accurate and precise. If you set aside the fatigue umbrella and focus on the core symptoms of this disease, if you establish frequency and severity threshold requirements, and if you recognize that your diagnostic criteria will be part of an iterative process of refinement, then I believe you can create diagnostic criteria that will advance clinical care and research, instead of putting us further behind: a definition that is sensitive and specific for ME/CFS.
I have submitted more details in writing, including references, for your review. Many of my fellow advocates have done the same, and I hope you will review that material with the same attention you have given us this afternoon. If you have any questions about what I’ve presented or if I can assist you further, here is my email address: jspotila AT yahoo DOT com. Patients are an essential resource for your task, and I hope you will involve us to the fullest extent. Thank you for the opportunity to speak to you today.