Abstract
Dear Sir The paper by Jensen and Bendtsen (1) addresses one of the most important issues in headache therapy—medication overuse headache (MOH). I would like to make the following comments:
The authors state that ‘the use of simple analgesics and combination drugs has steadily increased in Denmark with 12% increase over the last 5 years’. This statement is not referenced and can thus not be validated—neither the mentioned percentage, nor the ‘steady increase’ that is supposedly based on data collected over the course of several years.
The authors nowhere mention the absolute numbers that the percentages refer to, although ‘this tendency may raise awareness and alarm’; these numbers would be significant. How many tablets of analgesics per capita of the population were actually taken per year—10, 50 or 1j.1468-2982.2009.01957.x A percentage value without its basis lacks all informative value. Are such figures available for Denmark, as just published by Diener et al. (2) for other countries over a period of 20 years? If so, they should be stated and referenced; if not, numbers games with percentages should be abandoned.
The statement that ‘it is generally assumed that 50% of all analgesics are used for treatment of headache’ may be true, but not very expressive without either knowledge of the actual underlying data or, in particular, links to the frequency of occurrence of painful conditions that elicit the use of analgesics.
The authors state in the Epidemiology section that ‘a recent very detailed epidemiological study in Norway reported the prevalence of MOH as 1.72%’, but they conceal that the investigated population-based sample was restricted to the age group between 30 and 44 years. It is thus unclear whether such prevalence applies to the entire population.
However, should this be the case, it is incomprehensible why ‘the prevalence of MOH in Scandinavia . . . probably represents only the tip of the iceberg’, since the authors themselves state that ‘total prevalence of chronic secondary headache according to the International Classification of Headache Disorders, 2nd edn criteria, was 2.14%’. To describe a possible difference between 2.14% and 1.72% as ‘only the tip of the iceberg’ is not an appropriate scientific evaluation.
The reference to ‘clinical settings’ (better ‘clinical case series’) is correct, but the authors fail to comment that case series, due to their inherent selection bias and other systematic distortions that cannot be quantified and therefore cannot be controlled, are an unsuitable method to answer the question of generalizable MOH prevalences and substance-specific MOH prevalences (3). Results of clinical case series are not suitable to fill gaps in the knowledge of epidemiological problems.
In my view, the authors would have better met the severity of the problem by presenting the available data and their limitations and by explaining the still existing gaps in essential knowledge, including the methodical approaches to fill the latter.