Problems with International Consensus Criteria

This is probably a simplistic question but, why do they think that one million previously sane, active, happy people would all of a sudden choose a life of misery, isolation and disgrace?

To my mind, Dr. Jason's work is sometimes lacking in professional propriety. I'm not sure how you can claim to have revised the work of others, as he claims to have done in “The development of a revised Canadian myalgic encephalomyelitis-chronic fatigue syndrome case definition.” At best, you can propose your own modified version.

I'm not sure either how you maintain your role as objective arbitror, validating definitions, while at the same time promoting your own competing measures. At what point does conflict of interest become an issue?

I've read the paper, and as far as my memory serves me, there is absolutely no mention of anything like 'psychosomatic' or 'somatization' etc, or anything else along those lines. He simply reports the finding that a proportion of patients had comorbid psychiatric illness, based on the SCID. (Using the various SF-36 subscale scores, there was no increase in any psychiatric presentation, but there was when using the SCID.) Jason doesn't infer anything from the results.

It's a fairly short paper, and I've selected a couple of paragraphs that I think are the most significant from the conclusion (I'm not sure how much I can legally post, so I might delete some of this later):

Perhaps he has an agenda, but I can't see any harm in exploring the issues surrounding the ICC. I think the more research carried out on it, the better.

But I do worry that he might be approaching this issue the wrong way around.

He seems to be implicitly exploring the possibility (this is my own interpretation, and it might be wrong) that the ICC selects a heterogeneous cohort, because it includes a high number of comorbid psychiatric patients.

But I see two problems with this:
1. That the ICC may select patients with more severe neurological symptoms, and these neurological symptoms can either be mistaken for primary psychiatric symptoms, or can cause secondary psychiatric symptoms.
2. That the ICC may actually select a more homogeneous group of patients, a large proportion of whom have secondary psychiatric symptoms.

It does seem like a messy area though, in general. It's one that needs sorting out. Have there been many studies comparing and contrasting the neurological symptoms of ME patients with psychiatric symptoms? The psychiatric lobby would never have done this because they prefer to believe that all the symptoms are psychiatric, and that there are no neurological symptoms. So maybe it's a neglected area of research.

Watching the FDA conference videos, I noticed that some patients were talking about 'blacking out', which is a symptom that I don't read about very often. At the end of the conference, some of the organisers said the discussion regarding symptoms had been very educational, and that although they knew quite a lot about ME, they felt that they were only just beginning to get an insight into what it's really like to have ME. A very positive outcome, but sad that the authorities are only just beginning to grasp the subject.

Unfortunately the results of the SCID are pretty meaningless until he reveals how he treated symptoms like inability to read, watch TV, or socialize, as well as insomnia, hypersomnia, mild or severe cognitive problems, feeling weak, and being totally drained after minimal physical activity. The DSM-IV specifically lists these as indicating depression when the patient is to shy to acknowledge feeling depressed. And with the SCID it's very much up to the researcher/psychiatrist to decide whether or not to attribute them to a biological disease and therefore not interpret them as indicating depression.

Without those details, this research is quite useless.

Yes, and why would people who can't obtain a diagnosis have formed false illness-beliefs about having ME?

Part 1 of 3: Summary of the paper.

Brown, A. A., Jason, L. A., Evans, M. A., & Flores, S. (2013). Contrasting case definitions: The ME International Consensus Criteria vs. the Fukuda et al. CFS Criteria. North American Journal of Psychology, 15(1), 103-120.

Jason Brown et al used the SCID (Structured Clinical Interview for DSM-IV) to establish Axis I psychiatric diagnoses. It appears that the SCID was administered before the medical assessment. The authors note the potential problem of wrongly assigning medical symptoms to psychiatric categories in the assessment of symptoms which overlap between CFS and psychiatric disorders, state that the SCID helpfully allows for clinical judgement in the assignment of symptoms, and cite a psycho-diagnostic study (Taylor & Jason, 1998) which validated the use of the SCID in a sample of patients with CFS. I am not sure if the SCID is good enough (especially if done before a ME or CFS diagnosis), but for its potential faults, the SCID is still better than some other methods. The rates of psychiatric comorbidity were 27% for Fukuda defined CFS and 61.5% for ICC defined ME, and the difference was highly statistically significant (p=0.001). There are no details about which psychiatric diagnoses were made in either group, which is unfortunate.

From Table 1 (Demographics, Psychiatric Characteristics, & Onset Issues), there were no statistically significant differences between the two groups for age, gender, ethnicity, marital status, work status (including employment and disability payments), education level, viral attribution, causal attribution type (degrees of psychological vs physical). There was a significant difference (p=0.05) for sudden onset defined as less than 1 month vs gradual onset, with 26% sudden onset for CFS and 44.4% sudden onset for ME.

From Table 2 (SF-36 Subscales), which have “indicated sufficient psychometric properties as a measure of functional status in a CFS population” (Buchwald et al 1996), there were significant differences between groups for physical functioning, bodily pain, vitality, social functioning, but not for role physical, general health, role emotional, and mental health (somewhat contradictory to the previous finding of higher rates of psychiatric comorbidity in the ME group, but the ME group still had lower average scores for the latter subscales, just not statistically significant).

Table 3 (Significantly Different Symptoms), lists several dozen symptoms which are significantly more common-severe in ME than CFS. It is important to remember that these values are not n(%) but score(SD). It is a combined frequency x severity 0-100 scale. So the 23.99 (35.6) score for alcohol sensitivity in ME is not 24% of patients. It may be difficult to interpret how common alcohol sensitivity is, but this score is still rather low compared to scores for other symptoms listed. Alcohol intolerance may be affected by illness duration, something not studied in this paper. The authors also note that out of the remaining items not listed, most had directionally worse scores in the ME-ICC group.

Table 4 (Heart Rate & Cognitive Measures), the ME group had significantly higher HR when lying down, and after 2 minutes and 10 minutes after standing. The differences do not seem large enough to indicate that POTS is a major problem for most patients in either group, but there is not enough information to draw conclusions. The following symptoms are listed in the ICC but do not appear in Table 3 as showing significant differences: orthostatic intolerance (OI), neurally mediated hypotension (NMH), postural orthostatic tachycardia syndrome (POTS), and palpitations with or without cardiac arrhythmias. The cognitive measures (Trailmaking Test) were worse in the ICC group but the difference was not statistically significant (larger study samples may find differences to be significant). I had wondered if these results were presented together because the cognitive test was done while standing up, but that does not seem to be the case.

The following limitations need to be kept in mind: The data was collected before the ICC was published, so “questions used to tap the domains of the ME-ICC were not specifically designed for this purpose, and thus some symptoms were not ideally operationalized”. I am not sure if the new questionnaire used adequately captured the “PENE” category in the ICC as opposed to simplistic PEM. One item was missing completely (susceptibility to frequent viral infections with prolonged recovery periods), making it “easier for individuals to meet the Immune, Gastrointestinal and Genitourinary Impairments category, as participants were only required to have two symptoms rather than three from this category”. Also, homebound and bedbound patients were excluded. As the authors wrote, “Findings of the present study need to be replicated with larger samples of individuals and with questionnaires that are developed to specifically tap the domains of the ME-ICC.”

Part 2 of 3: Some commentary on the paper.

Judging from Cort's blog article, it sounds like the ICC passed two main tests (more functionally disabled and more "ME/CFS" like) but supposedly “failed” for psychiatric comorbidity. Others have commented that there are major issues with how psychiatric comorbidity is defined and interpreted, so the ICC has not necessarily failed anything.

Perhaps part of the problem may be with the notion that an accurate organic ME definition must fundamentally define a cohort with little or no overlap with the often vague and questionable criteria for psychiatric diagnoses, in order to better distinguish itself from "psychiatric CFS". However, in the descriptions of the ME outbreaks which people hold up as classic examples of genuine ME, depression and emotional instability were apparently common features, and after consideration of the evidence and the nature of patients' experiences or previous histories, these were presumed to have organic causes. Later in 1990, Dowsett et al also reported that about 98% of a sample of several hundred ME patients experienced "emotional lability" (includes frustration, elation, depression), although this does not necessarily mean that 98% of patients were diagnosed with a psychiatric disorder per se.

Inappropriate overlap and conflation with psychiatric diagnoses has been a serious problem, a tricky issue which Jason has been studying for some time. It is feared that a bad definition would include people with primary psychological causes for their symptoms. I think that fear is warranted, but, adequately defining and measuring the cardinal symptoms of ME is much more important than avoiding association with psychiatric comorbidity or requiring it to be lower than rates for Fukuda defined CFS. People with psychiatric symptoms but without cardinal features of ME should never be diagnosed with ME. People with the cardinal symptoms of ME should not be excluded from a ME diagnosis just because they also have symptoms which are captured in questionable criteria for psychiatric diagnoses.

Lupus erythematosus has numerous “neuropsychiatric” consequences, but trying to define LE by excluding those symptoms would not necessarily be more accurate. alex3619 mentioned "somatization disorder", which is largely dependent on assumptions made about the cause of symptoms, and I recall a study on Parkinson's disease which found that over 90% of patients would be diagnosed with somatization disorder if PD was ignored as an explanation for their symptoms. There has also been research in other areas which suggest that a general association exists between the number of physical symptoms and the presence of psychiatric comorbidity regardless whether the physical symptoms are explained by medical disease or medically unexplained (the associations are not identical but close enough to say a general association exists regardless of etiology). Furthermore, those with medical diseases are also more likely to report so-called “medically unexplained” symptoms, another arbitrary judgement on behalf of the clinician.

So it is not surprising that rates of psychiatric diagnoses was higher in the ICC group than the Fukuda group. I do also wonder how much of the difference was due to crossing the border i.e. mental health in the ME group was only somewhat worse than the CFS group but it was enough to push many over the threshold for a full blown diagnosis. As long as the definition adequately defines ME and adequately excludes primary psychological explanations for symptoms, then chances are the increased psychiatric comorbidity is reflective of harsh realities and not a “failure” on behalf of the case definition. As others have already noted, another major part of the problem is with the psychiatric diagnoses themselves and what they are supposed to be describing or defining. Many if not most of the common methods for measuring and diagnosing psychiatric disorders presumed to be often present in ME and CFS suffer from the issue where physical symptoms and obvious consequences of the illness can be misleadingly counted as evidence towards a psychiatric diagnosis.

Personally I do suspect that psychiatric symptoms in ME can be due to the biological effects of the disease process combined with negative psychosocial environments (especially those arising from the impact of ME) and any pre-existing vulnerabilities which have nothing to do with developing ME, keeping in mind that patients are from all walks of life before being stuck down and we are not all perfect specimens of awesome living and psychological wellness before becoming ill.

I am not an expert on how ME “should” be defined exactly. I appreciate the effort for statistical analysis of questionnaires in order to identify hallmark symptoms (which seems to be potentially circular as it depends on the cohort being studied in the first place), but it needs to be supported by correlations with objective measurements, and reflect both the original and contemporary understanding of how ME is best defined. I do not know if Jason's research team will end up cracking the problem, but I think they are at least heading in the right direction with the emphasis on post-exertional symptomatology, neurocognitive symptoms, and autonomic symptoms. What about pain/malaise and intolerances? I would also like to see a further examination of the historical descriptions of ME and neurological signs.

It is a good sign that all these issues about case definitions are coming more into the foreground of discussion and research. It also appears that the obsession with poorly defined “fatigue” is weakening? Along with the misguided assumption that “fatigue” must be the most impairing symptom rather than other symptoms and/or the body's inability to compensate for the chronic symptom burden.

Part 3 of 3: More commentary on the issues involved.

parvofighter asked some good questions which I think deserve repeating:

#1 If we treat the ostensible (apparently so but not necessarily so) psychiatric comorbidity, will it improve function? I have noticed a trend in CFS research where psychiatric comorbidity is not really an important factor in disability or even in the response to CBT/GET (*cough*), so I doubt that treating it is going to improve functionality much on average. Perhaps some individuals will be helped more, as obviously someone with significant depression or anxiety etc would be better off without it, but I also suspect it may be harder to treat psychiatric comorbidity in ME or CFS.

#2 (how is psychiatric comorbidity being defined) and #3 (multisystem signs and symptoms conflated with somatization) I already commented on briefly and right now cannot add much more to what parvofighter wrote. Valentijin has written numerous sobering posts on the issue in recent months. We need to keep a keen eye on this issue, but that said, I am more comfortable having Jason et al doing this sort of research than Wessely et al, as judging from the latter's work they do a poorer job at considering or reducing conflation of CFS symptoms counting towards psychiatric diagnoses and have a stronger tendency to take psychiatric diagnoses at face value.

4# Which aspects of depression, if present, are featured? Good question. I have read numerous times about researchers and patients saying that depression in ME or CFS often does not seem to be the same as when stand alone. So there is something about depression criteria which captures some aspect of what some patients experience due to illness but do not really adequately define it. I have experienced a few temporary bouts of depression and it is just not the same.

Interesting about the previous SF-36 health survey results, where all the bars are very low except for Mental Health and Role Emotional which are relatively preserved. An unsurprising reality for patients, but also impressive that the CDC seem to be catching on that the symptom domains are predominately physical rather than affective. Also, in the Brown et al study, the role emotional and mental health subscales were the highest out of the eight.

Perhaps the main problems lie in how psychiatric comorbidity is being defined/measured and interpreted, and in peoples' expectations of what the ICC was supposed to do. Of course Wessely et al predicted such an association would occur with the CCC due to the higher number of required symptoms. Jason et al's earlier work demonstrated opposite results to the prediction but followup studies conversely provided support for it.

We do indeed need to challenge assumptions about psychiatric comorbidity, and also acknowledge that elevated rates would be unsurprising given the symptom burden and impairments, the possible neuroimmune mechanisms, and the disrespect/neglect involved. I understand why Cort has interpreted the results of the study as a major potential problem for the ICC, given the previous battles of organic ME vs “psychiatric CFS”, but I would disagree that higher psychiatric comorbidity equals a "psychiatric cohort", as long as the ME case definition is adequately reflective of the disease itself.

Thank you Bob, that clears up the question. Psychiatric comorbidity, unspecified, is not an issue. As others have said, we would expect more severe patients to have more psychological issues, which might allow a comorbid psychiatric diagnosis.

One thing that always concerns me though is if symptoms of ME are similar to symptoms of depression, how do they distinguish them? I suspect many of the instruments they are using should be called blunt instruments.

See pages 9 – 10 of the CCC Guidelines for “Differences Between ME/CFS and Psychiatric Disorders,” including “Depression” and “Somatoform Disorder.”

Note also, at the foot of page 8 of the ICP, a list of “Exclusions,” including primary psychiatric disorders and somatoform disorder, and of “Co-morbid Entities,” including Reactive Depression.

According to the list of exclusions, nobody diagnosed with a primary psychiatric disorder should have been given an ME-ICC diagnosis.

I am unable to get into a longer discussion about this issue now, and Ember posted something useful already, but your post reminded me of this comment from Cort on his blog article:

http://www.cortjohnson.org/blog/201...-for-more-psychiatric-patients/#comment-15123

A psychiatric assessment does need to be done before a ME diagnosis can be made. I just wonder how the research team, after acknowledging the problem of wrongly assigning medical symptoms to psychiatric categories, can best make such decisions if the psychiatric interview is done first without any knowledge about the patient's medical status.

One point following up on biophile's excellent posts: are they capturing depression, or that we are unhappy? How could we be happy with a potentially life long disabling illness that limits not only our physical function but our cognitive function? Is this being interpreted as depression? Sure we might get depressed, thats easy to see, but how much are they picking up not only physical and cognitive ME symptoms but just plain unhappiness? I suspect that many ME patients with "comorbid depression" might not be depressed at all but simply unhappy. Similar arguments apply to things like the effects of isolation and responses to societal abuse. I made the argument some years ago that anxiety is not necessarily dysfunctional in ME, its a survival feature that enhances our capacity to cope - we are anxious about the things that make us worse. This is not dysfunctional.

Its not just Lenny who has pointed out its possible to distinguish between depression and ME. Komaroff showed this objectively using spectral coherance EEG. Other docs have noted its easy to spot with a simple interview ... just ask the right questions. My question arises as to standard questionaires and how they may fail to make these distinctions. I do not know enough about the questionairre used by Lenny to be able to discuss how much they take this into account.

Leaving aside how reliably in such circumstances a distinction can be made between primary and secondary depression, that exclusion opens the door to psychs being able to avoid ever having to deal with the possibility of a primary ME (or CFS) diagnosis by simply diagnosing primary depression (or whatever) in the first place.

I don't see why primary depression and ME (or CFS) are mutually exclusive, nor why one should take causal precedent over the other, in the regular clinical setting. (The exclusion requirement is not a problem in a research setting where it is important to get a 'clean' sample, with minimal co-morbidities.)

I think you identified half of the issue in your post: the continued confusion between research diagnostic criteria and clinical diagnostic criteria. Generally speaking an exclusory condition can only be validly used clinically if it can explain all the symptoms. In a research setting it might be important to exclude such patients however. For example, like half of all patients under a CFS diagnosis (this was not research done on an ME definition) I have sleep apnoea. There is no way it can explain most of my symptoms though. Hence its not clinically exclusory. It might however explain a component of my fatigue and cognitive dysfunction, but only to a point. However more than a few of my comorbidities might rule me out from participating in future ME research as a test subject.

However disorders like depression and somatization are so symptomatically broad that almost anything could be claimed that was not clearly nailed down as a physical issue. This leads to the other half of the issue: the continued battle by those promoting psychogenic diagnosis to show that these diagnoses have validity.

I think in both cases (research and clinical diagnosis), psychiatric assessment needs to be done with an open mind. If the patient doesn't 'fess up to being depressed, it's very presumptuous to then use physical and cognitive symptoms to assume depression and shut the door on other possibilities.

At best all that they determine is that the patient has un-admitted depression OR an underlying biological problem causing the same symptoms. Or both. But it's very stupid to assume that physical and cognitive symptoms MUST indicate depression.

DSM sucks in general.

Thanks. I guess at some point I should find this study and read it.

Just thinking here, if someone is stigmatized, has lost all his friends, and is so disabled he can barely take care of himself -- why wouldn't he run a greater risk of developing a psychiatric issue -- especially if he is also suffering from inflammation in his central nervous system.

I don't see the problem.

We like to think of ME/CFS as a neurological condition.

Many neurological conditions result in neuropsychiatric symptoms that in the absence of the core diagnosis (e.g. Alzheimer's, lupus, encephalitis etc) could result in a psychiatric diagnosis.

With the exception of focal lesions I find it hard to conceive of a neurological process that would impact on energy, movement, pain, cognition etc yet spare the brain regions associated with 'mood'.

Excluding patients who meet the key ME/CFS requirements (espec PEM) yet have a prior or 'co-morbid' psychiatric diagnosis makes no sense to me.