In theory,theory and practice are the same;in practice they are not

To the Editor,

I applaud and appreciate the tremendous amount of work that has been performed by the Classification Committee to establish the International Classification of Headache Disorders (ICHD) −3 beta classification. However, the ICHD has changed a couple of times since its first version, and the current version is unlikely to be a law of Medes and Persians that we should adhere to for the rest of our time.

Our medical knowledge has improved during the last centuries, because some doctors had doubts regarding what used to be fixed paradigms, or – in casu – fixed disease classifications. Such criticism is indeed “certainly important” (1). But, contrary to the statement of Schulte and May, in my opinion it is this very criticism that “gets to the core of the matter” (1). If not that every now and then even the most influential theories had been the subject of criticism, we would still be living on a flat earth, notwithstanding the fact that “scientific classifications are undoubtedly of vast importance in scientific research”, and that “the first step in any study, any scientific process … is a definition of the object prior to investigating it” (1).

Recently, Shevel and Shevel criticised the criteria to define migraine within the ICHD 3-beta (2). The remarks put forward are quite logical and justified (2). Therefore, the paper by the Shevels should initiate a discussion on how to classify migraine in an evidence-based manner. Without the possibility of debate, the process of making a consensus report seems too much a form of decision-making by GOBSAT (good old boys sat around a table) (2,3).

According to Schulte and May, “it does not matter how we call an object of scientific research as long as there is a consensus in the scientific community about the nature of this object, i.e. its basic characteristics” (1). Their first conclusion is that “there has to be consensus about the basic features” (1). I agree with these statements, and exactly herein lies the main point of the discussion: Given the manuscripts by the Shevels, and Lane and Davies, the desired consensus does not exist in medical practice, QED (2,3).

Criticism therefore can and should be used to redefine the defining criteria if necessary. That serves our patients. And if amendments to the framework have been made, the new definition should be used in the next studies regarding the topic.

Classifications define the key features of certain entities and thus demarcate them against others. However, what if not all patients can be classified with the current scientific key features? If we treat patients in clinical practice, it might happen that we are faced with a person who doesn’t fit the currently used defining criteria within a research classification, even though the clinician thinks the patient might have the disease that was meant to be described (Migraine is a disease that has been known for ages (3). So we all knew to some extent what migraine was long before the defining criteria in the ICHD were published some 30 years ago.)

This patient should probably be treated using the regular treatment for that disease. Therefore, a classification of “probable migraine” has been made already (3). Others suggest “cluster minus one”. I have seen a patient that would fulfil the criteria for “cluster minus one”, but who wasn’t treated by his doctors as a cluster headache patient because of, in my opinion, a too-narrow interpretation of research criteria in clinical practice. He was actually not treated at all, since he didn’t fit any diagnosis. This, even though the patient was rather suicidal from the attacks.

The nature of a definition is indeed to offer a short and concise explanation of the object. However, definitions and objects can change. For example, the big animal with a trunk and tusks, as described by Schulte and May, used to be a mammoth (1). Only after ameliorating the description with “living species” – or with imaging studies using photographs instead of cave drawings – do we know that we’re discussing elephants instead.

Schulte and May described three possibilities for changing a classification (1). One of these would be that “additional features of an object may come to light, which were not evident when the classification was made and might thus be added to the classification”.

However, I strongly concur with the Shevels that if the additional information or other data were available – see reference 2 – but not used for unclear reasons, that these constitute a very good fourth reason to change a classification as well.

So the theoretical option to reinvestigate the “headache population” – an immensely big population of objects with one general criterion, i.e. “headache” – either with new studies or by systematic review, and to try to identify clusters of characteristics commonly appearing together, is perhaps a way forward (1).

I agree with Lane and Davies that the primary headaches constitute more of a continuum, in which different entities exist (3). I think that headache triggers, depending on inter- and intrapersonal factors, might lead to characteristics fitting this form of headache in one person, and other symptoms fitting another headache in another person. And different intrapersonal factors might lead to the development of one type of headache at this moment and to another type at another moment. Therefore, I concur with Lane and Davies that to pretend that complex symptom-based syndromes can be constrained by the strict diagnostic criteria of the ICHD-3 beta, is unhelpful (3).

The ICHD-3 beta classification is a very nice framework, and I do agree that in studies we need to apply – indeed prior to investigation (4) – the current consensus criteria from this framework. Still, the framework itself can and should be the subject of scientific discussion; discourse is the basis for scientific progress (1). Claiming semantically and philosophically that we aren’t allowed to criticise the research criteria within the ICHD-3 beta, even whereas they hinder medical practice, is a bit absurd.

For research, Schulte and May may be right “the classification was not written for clinicians but for scientists” (4); for clinical practice “it has proven to be so exact that a diagnosis should not be done without it” (4) they are not.