Are the Canadian Consensus and International Consensus Criteria Broken? Study Suggests Both May Select for More Psychiatric Patients
The Canadian Consensus and International Consensus Criteria are much beloved by the chronic fatigue syndrome community and it’s easy to see why; after decades of dealing with a vague definition, people with ME/CFS could finally see themselves in a definition for the first time. Created from within the community by ME/CFS practitioners, researchers and advocates, the general consensus has been that these criteria couldn’t replace the Fukuda definition too quickly, and that doing so will solve some of the research problems this disorder faces. They have been a breath of hope for many.
Two DePaul University studies lead by Dr. Jason’s DePaul University team suggest, however, that these definitions could actually make things worse. While they do capture a more ME/CFS-like group they appear to be plucking out a subset of patients with more psychiatric disorders.
The Study (this study has not been published on PubMed yet, the abstract is provided below.)
Contrasting Case Definitions: The ME International Consensus Criteria vs. the Fukuda et al. CFS Criteria Abigail A. Brown, Leonard A. Jason, Meredyth A. Evans & Samantha Flores DePaul University
Study Abstract: This article contrasts the Myalgic Encephalomyelitis International Consensus Criteria (ME-ICC; Carruthers et al., 2011) with the Fukuda et al. (1994) CFS criteria. Findings indicated that the ME-ICC case definition criteria identified a subset of patients with more functional impairments and physical, mental and cognitive problems than the larger group of patients meeting the Fukuda et al. (1994) criteria. The sample of patients meeting ME-ICC criteria also had significantly greater rates of psychiatric comorbidity. These findings suggest that utilizing the MEICC may identify a more homogenous group of individuals with more severe symptomatology and functional impairment. Implications of the high rates of psychiatric comorbidity found in the ME-ICC sample are discussed.
This study took patients from a variety of sources including physician referrals around the Chicago area, selected 114 who met the Fukuda criteria and then assessed whether they met the ICC criteria. They then assessed each groups symptom severity and functionality using a variety of questionnaires, took their heart rate and then analysed the data to see what types of patients met the two criteria.
Better Scoring System
One thing Jason has going for him is an excellent symptom questionnaire (DePaul Symptom Questionnaire -DSQ). Brilliant in its simplicity, the DSQ is a retrospective head slapper. In past surveys a severe headache once a month counted the same as a severe headache daily; all you could mark was that you had a severe headache. Since Jason looks at both frequency and severity he can rank symptoms according to how significant they really are. ((Frequency score x 33.3) x severity score then /100)). In this survey a moderate headache occurring regularly will get a much higher score than one occurring a couple of times a month.
More Functionally Disabled – check!
The ICC did a superb job of picking out more functionally disabled patients; Jason’s survey found almost 50 symptoms which were significantly worse in ICC selected patients than the Fukuda patients.
More ME/CFS-Like – check!
When I talked to Dr. Jason he noted that the creation of the Fukuda definition was a reaction to the concern that the Holmes definition was bringing in too many people with psychiatric diagnoses. The cure for that problem was simply to reduce the number of required symptoms from eight to four. That worked, Dr. Jason thought – the number of psychiatric diagnoses did drop – but where Fukuda et. al erred was in allowing any combination of the eight symptoms to qualify you for CFS.
That meant you could qualify for ME/CFS without having any of the core symptoms of this disorder; post-exertional malaise, cognitive problems and unrefreshing sleep. That almost certainly meant the Fukuda definition was allowing at least some people into studies who did not have ME/CFS.
The ICC was designed to fix that fault and it appears that it did. The group meeting the ICC criteria were particularly cognitively and exertion challenged and many cognitive scores and the post exertional malaise score were significantly higher than in the Fukuda group . This suggested that the group meeting the ICC criteria was indeed more ME/CFS-like than the group meeting the Fukuda criteria – an important accomplishment.
Not Over Representing Psychiatric Disorders…..Fail
One of the concerns regarding the Fukuda definition was that it was so vague that people with other disorders who did not have ME/CFS were being included in studies. Ironically, Jason’s study suggests the ME-ICC may be bringing in significantly more people with psychiatric disorders than the Fukuda definition ever did, and for the opposite reason – it’s too specific.
Jason pointed to fatigue studies which indicate that the more symptoms a definition requires, the more likely it will select for people with more psychiatric disorders; in particular, somatization disorder. Like the defunct Holmes definition the ICC requires that eight symptoms be present and that may be too many.
It’s not that ME/CFS patients don’t have a lot of symptoms; they have a lot of symptoms but unless a definition focuses on a disorders ‘ordinal’ symptoms; ie that groups of symptoms that are particular to that disorder – it risks classifying other disorders as ME/CFS.
Even though the CCC and ICC did require several ordinal symptoms Jason’s study suggested the large number of symptoms required still caused them to select out people with more psychiatric disorders. Since Jason’s study focused on CFS patients its not clear that the ICC would also inadvertently bring in people with mood disorders who do not have ME/CFS but it’s at least theoretically possible.
Because people with more severe illnesses are probably more like to have mood disorders an increased rate of psychiatric disorders might be expected in this more functionally disabled group but the percentage of psychiatric disorders was so much higher in the ICC group (61%) compared to he Fukuda group (27%) that more than illness severity must have been in play. (An earlier study found increased rates of mental (and physical) illnesses in Canadian Consensus Criteria derived patients vs Fukuda Definition derived patients as well. )
This study was small and needs to be validated. The higher percentage of gradual onset patients suggests that while physician referrals were included the study included patients not typically seen in ME/CFS specialists office. They are, however, probably more reflective of the population at large.
Potentially Very Negative Consequences
If the data is even close to being correct psychologists in the UK and Europe would start finding much higher rate of psychiatric disorders in ME/CFS suggesting that in a worse scenario the ICC could inadvertently become a Trojan horse, so to speak, that helped get ME/CFS more identified with psychological disorders. What a shock that would be.
It’s also possible and, one would think probable, that the ICC would select out patients who showed more physiological abnormalities as well but given the attitudes of the research community the focus could still shift to a psychological one.
The loss of these definitions would be difficult to take given how enthusiasm they’ve generated but thankfully finding a definition that doesn’t have highlight psychiatric issues but still defines an ME/CFS group should not be that difficult. In fact it’s already been done.
Minor Tweaks Needed
If the ICC needs to be tweaked at least it’s not that far off; it’s got core characteristics of ME/CFS embedded in it; it simply needs to pare down the number of other symptoms it requires for a diagnosis. Jason found that an ME definition based on Ramsay’s criteria was pretty darn close; it picked out a more severely ill group of patients than the Fukuda definition without increased rates of psychiatric illness. That definition required sudden onset, post-exertional malaise, one neurocognitive symptom and one autonomic symptom.
A Pathway to a New Definition
You need a pathway – Dr. Jason
The jist of all this is that Dr. Jason believes that an ME/CFS definition needs to be a) relatively brief (probably around 4 symptoms) and b) derived empirically; that is, derived through statistical analysis. Not only is the subject matter too complex but an empirically derived definition would gain the trust of researchers and pharmaceutical companies. None of the at least eight definitions (Holmes, Fukuda, Oxford, Empirical, Pediatric, Hyde, London, Canadian Consensus, International Consensus, Myalgic Encephalomyelitis) created for ME/CFS have been empirically derived but Jason has produced a pathway to generate one.
He Has A Plan
It’s a four-step process.
(1) define the core symptoms of the illness and operationalize them (ie find ways to get valid data on them by connecting physiological indices to self-report measures). Jason has noticed that in many data sets, as well as in many ME case definitions, post-exertional malaise, cognitive issues and unrefreshing sleep are core, fundamental symptoms and his group is working to operationalize those and other symptoms and issues associated with ME and CFS.
PEM is found in other disorders but it is a critical and core symptom of ME. Jason has validated ways to measure post-exertional malaise using a questionnaire specifically developed for ME/CFS (called the ME/CFS Fatigue Types Questionnaire.) In that study, controls did not have this symptom whereas patients did. Unrefreshing sleep is common in healthy people and it occurs in almost all patients with ME. Many ME theorists have also noted neurocognitive symptoms such as memory and concentration problems. Throw the three symptoms together and you may have the start of a definition for ME. ( Pain is a core symptom for many but not all people with ME.)
Jason noted one of his graduate students is currently trying to operationalize sudden vs gradual onset. The problem is actually quite complex one; how sudden is sudden? What if you had a series of colds that just got worse? At which point do you determine you have ME/CFS? What if you got sick, got better and then got sick again?There are many issues like this with our case definitions, and to increase reliability, we need ways to better operationalize these concepts.
(2) use the symptom questionnaire (DSQ) across many groups to check validity of the core symptoms. Jason’s already got about 500 DSQ questionnaires answered; the DSQ is being used in the CAA Biobank, the CDC multi-center physician project, the Bested Clinic in Vancouver and others and will soon be integrated into RedCap, which will make it accessible to thousands of researchers around the world. This instrument allows an investigator to reliability indicate which case definition a patient meets, and then it is possible to statistically compare and contrast the different definitions, including the Canadian ME/CFS criteria as well as the ME-ICC criteria.
(3) All current case definitions are based on consensus methods, but statistical approaches using data mining and factor analysis can be employed to identify core symptom domains.
(4) Come up with research and clinical definitions. The research case definition is the tightest and needs to occur first. Then the clinical case definition with looser criteria to include more individuals who really have the illness but might have just missed making the research criteria. The goal of the research criteria is to includes as broad swath of ME/CFS patients as possible without including other types of patients. Given that focus, the research criteria will probably always be a bit too constrictive; ie it will miss some patients but so long as it gets a representative sample it will be fine. The clinical criteria will pick up the rest of the patients.
The CCC and ICC moved the field forward by identifying and requiring that core symptoms be present and by identifying more functionally disabled groups. Two DePaul University studies lead by Dr. Jason’s team, however, suggest that requiring eight symptoms be present may cause the criteria to select out a patient subset with more psychiatric disorders. Their studies suggest that a shorter definition that includes the core symptoms may be more appropriate. Dr. Jason has worked out a pathway to produce a statistically derived definition.
If the DePaul studies are correct this is a painful moment for a community that placed so much hope on these definitions. These definitions were never going to be last word, though, for ME/CFS. By highlighting the need for an ME/CFS definition to focus on the hallmark symptoms of this disorder, the CCC and ICC have played an enormous role. Nothing in this study suggests they are not excellent clinical definitions; the question is whether they’re suitable for research purposes.
Given the continuing search for biomarkers and more and more evidence of subsets, we may and hopefully will be talking more and more about changes in ME/CFS definitions over the next couple of years. Hopefully we will.
Like the blog? Make sure you don’t miss the latest on ME/CFS and FM treatment and research news by registering for our free ME/CFS and Fibromyalgia blog here.