The need for revision of the criteria for 8.2 Medication-overuse headache and its subforms became apparent soon after publication of The International Classification of Headache Disorders, 2nd edition. Fred Sheftell first drew attention to potential problems with these criteria, and constructive criticism at the International Headache Research Seminar in Copenhagen in March, 2004, which discussed the classification in considerable detail, confirmed that some re-writing was necessary. This task was undertaken by Stephen Silberstein, chairman of the Working group on Headache Attributed to a Substance or its Withdrawal, together with the other members of the IHS Classification Subcommittee.
There are two major resultant changes. The first is the elimination of headache characteristics for the subforms of medication-overuse headache since it was agreed that none could be regarded as typical. The second is the introduction of a new subform, 8.2.6 Medication-overuse headache attributed to combination of acute medications, that recognises patients overusing medications of different classes whilst not meeting the criteria for overuse of any single class. This change required the re-numbering of subforms 8.2.7 (previously 8.2.6) Headache attributed to other medication overuse and 8.2.8 (previously 8.2.7) Probable medication-overuse headache.
Introduction to the Classification
This second edition of The International Classification of Headache Disorders, like the first, is intended equally for research and for clinical practice. No research studies are likely to be accepted in international journals without adhering to this classification, but the classification is equally important for clinicians. The great majority of evidence-based treatments for headache have been developed using the first edition of The International Classification of Headache Disorders. This second edition has not changed the major principles of the classification and diagnosis of primary headache disorders. Therefore, the existing body of evidence gained using the first edition remains valid for most diagnoses made using the second edition. When you look for patients who will respond to a triptan, you must diagnose your patient according to the diagnostic criteria for migraine with aura and migraine without aura of this classification.
The International Classification of Headache Disorders, 2nd edition, is perhaps the single most important document to read for doctors taking an interest in the diagnosis and management of headache patients. There is often a huge gap between researchers and clinicians. Many have suggested that there should be two classifications, one for research and one for clinical use. However, if there were two classifications, all new information would be gathered using the research classification and transfer of results from research studies to clinical practice would be difficult. Therefore, the universally-accepted view of experts in disease classification is that there must be one classification only, but constructed so that it can be used at different levels of specialisation.
The answer to the problem is hierarchical classification and this system was already adopted by the first edition and remains unchanged in The International Classification of Headache Disorders, 2nd edition. All headache disorders are classified into major groups and each group is then subdivided one, two or three times into headache types, subtypes and subforms. For example, 1. Migraine is a group consisting of one headache type (migraine) and the subtypes of migraine such as 1.2 Migraine with aura constitute the next level (second digit). Migraine with aura is again divided into subforms, for example 1.2.1 Typical aura with migraine headache. The practising family physician may only need to diagnose at the first level – migraine – in order to select acute treatment. However, when there is a problem of differential diagnosis, for example because headache is absent, it becomes necessary to distinguish between migraine with aura and other disorders that may mimic it, and thus to code at the second or third levels. Practising neurologists and headache specialists would normally diagnose the precise subform of migraine with aura at the third level. This system has proven utility at the different levels of healthcare systems throughout the world.
Classification means deciding on which kinds of diagnostic entities should be recognised and how to order them in a meaningful fashion. In doing so, one should draw upon all kinds of available evidence: clinical description, longitudinal studies of cohorts of patients, epidemiological studies, treatment results, genetics, neuroimaging and pathophysiology. This was done for the first edition and it has been repeated for the second edition of The International Classification of Headache Disorders. Fortunately, big changes have not been necessary, but a fairly large number of small but important changes have been made in the light of new evidence. Thus, we have introduced 1.5.1 Chronic migraine as a new diagnosis for those rare patients who fulfil the diagnostic criteria for migraine for 15 or more days a month without overusing medication. All secondary headaches are now described as “attributed to” another disorder while the first edition used the less-precise term “associated with”. The causal link between the underlying disorder and the headache is in most cases well-established, and we have therefore been able to strengthen the terminology.
With regard to psychiatric disorders, there is no reason to treat them differently from all other disorders that can cause secondary headaches. Therefore, we have included a new chapter 12. Headache attributed to psychiatric disorder. The problem is that research elucidating this field is extremely scarce so the chapter is very brief. The corresponding section in the appendix is more comprehensive and will hopefully greatly increase research into the relationship between psychiatric disorders and headache.
All headaches caused by infection are now placed in the same chapter 9. Headache attributed to infection whilst, previously, intracranial infections were placed in the chapter on intracranial disorders. A new chapter 10. Headache attributed to disorder of homoeostasis has been added. Some new entities such as 4.5 Hypnic headache, 4.6 Primary thunderclap headache and 4.7 Hemicrania continua have also been added while 13.17 Ophthalmoplegic “migraine” has been moved from chapter 1. Migraine to chapter 12. Cranial neuralgias and central causes of facial pain.
As a major change in code-numbering, the tabulation below now includes WHO ICD-10NA codes (in parentheses) because these are the codes used in daily practice. In many places The International Classification of Headache Disorders, 2nd edition, is more detailed than the WHO Classification. This means that some headache subtypes are not uniquely coded under the ICD-10NA system but the most appropriate ICD-10NA code has in each case been attached to the ICHD-II code.
The basic construct of each chapter in the second edition of The International Classification of Headache Disorders is the same as for the first edition. For each chapter the classification for that chapter is shown. Then there is an introduction and the different headaches are presented one by one in the order of the classification. For each major disease we give previously used terms and specify disorders that are related but coded elsewhere and we present short descriptions that in words try to define the disease. After that we present explicit diagnostic criteria. Finally, we provide written comments and a selected bibliography at the end of each chapter.
The explicit diagnostic criteria need a comment. Previously they were called operational diagnostic criteria, but the meaning of “operational” is not generally known. “Explicit” means “unambiguous, precise and with as little room for interpretation as possible”. In other words, the aim is to write criteria so clearly and have such clear requirements that different doctors in different parts of the world are able to use them in the same way. Terms that are open to interpretation such as “sometimes”, “often” or “usual” are largely avoided. Patients must fulfil all criteria listed as A, B, C, D, etc. For each criterion there are specific requirements such as “two of the following four characteristics”, etc. The same system was used in the first edition where it proved to be reliable and reproducible. It has also been shown that the first edition was applicable in all settings spanning from epidemiological studies in the general population to tertiary headache referral centres. A validation of the first edition and its explicit diagnostic criteria was provided by the triptan studies, in which the rates of success were equal in different countries indicating that case ascertainment had been the same. Furthermore, the high success rate of injectable sumatriptan demonstrated that, at least from pathophysiological and pharmacological points of view, the diagnostic criteria for migraine with and without aura had delineated a fairly homogeneous entity. For this and many other reasons we have only made small changes in the diagnostic criteria for migraine.
Classification and diagnostic criteria can be aetiological or descriptive and the latter can be syndromic or symptom-based. Both first and second editions of The International Classification of Headache Disorders are aetiological for the secondary headaches and symptom-based for the primary headaches. If the course or the evolution of headache syndromes should be taken into account, there would have to be so much information available that the diagnosis of migraine, for example, would enable the prediction of a particular course for that patient. The fact is that the evolution of primary headache syndromes cannot be predicted. Some patients will worsen and their symptoms become chronic, others will be relieved of their primary headache and yet others will stay the same for decades.
It is an important task for the future to provide prognostic factors and other characteristics that may make it possible to classify subtypes of migraine and tension-type headache. For some time it looked as if 3.1.2 Chronic cluster headache could be subdivided into chronic from onset and evolving from episodic, but then it was shown that a number of patient with chronic cluster headache reverted to 3.1.1 Episodic cluster headache. Thus, many different evolutionary patterns seem to be crossing each other. The same is true for migraine according to the longitudinal studies of Bille and others. For these reasons the evolutionary history cannot be classified until much bigger and better studies of the evolution of migraine patients become available.
Like the first, this second edition of The International Classification of Headache Disorders classifies patients according to the phenomenology of their headache(s). For clinical use, drug trials and pathophysiological studies this would normally mean that the patient must have had that type of headache within the last year and is likely to have further attacks. For other uses, particularly genetic studies, we are more concerned with the lifetime history of the patient. Thus, if the patient has had migraine attacks twenty years ago, but no attacks after that time, the patient still has the phenotype of migraine in a genetic study. These principles make it possible for a patient to have one diagnosis at one time and another diagnosis a few years later. It also makes it possible and necessary to give some patients more than one headache diagnosis and even two or more migraine diagnoses.
So far, only two migraine genes have been identified. They account for only half of patients with the rare disorder 1.2.4 Familial hemiplegic migraine. Thus, genetics have not had significant impact on The International Classification of Headache Disorders, 2nd edition. However, it is expected that within the next ten years migraine genetics will be elucidated. This will undoubtedly lead to major alterations in the way we classify headaches, but it is not possible at present to say exactly how such changes will be. Some monogenic entities will probably be identified and it will become obvious that our clinically-defined phenotypes are heterogeneous. On the other hand, mutations in the same gene may cause quite different phenotypes as recently demonstrated by studies of familial hemiplegic migraine. Thus, the genetics of migraine may simply prove to be so complex that, in daily practice and perhaps to some extent in research, we shall continue with clinically-defined diagnoses.
A classification and its diagnostic criteria should be reliable, valid and exhaustive. Fortunately, as partly discussed above, the first edition of The International Classification of Headache Disorders has been shown to have fairly high degrees of reliability and validity. It has also proven to be exhaustive in several studies spanning from population-based studies to studies in headache clinics. We believe that the second edition is even more reliable, valid and exhaustive, but only future research can prove or disprove this belief.