Abstract
Dear Sir Anna Ferrari and co-authors at the Headache Centre in Modena (Italy) have made very commendable, scientifically sound efforts towards an urgently needed, more differentiated consideration and classification of medication overuse headache (MOH) (1). The more are we bewildered, therefore, about a methodically extremely doubtful approach upon which two papers in Headache by this working group are based (2,3). In both cases, the authors apply the Leeds Dependence Questionnaire (LDQ), developed to measure alcohol and opiate dependence, to the use of analgesics. The authors assume that this self-completing questionnaire with 10 items that is validated for alcohol and opiate consumers in addiction and psychiatric settings can easily and without problems be applied to consumers of analgesics, although they rightly advocate elsewhere (1) not to mix psychotropic and non-psychotropic substances. The developers of the LDQ (D. Raistrick et al.) correctly emphasize that ‘there needs to be further validity testing particular with other drug groups and the relationship to intake’ (4).
Some examples can demonstrate that the approach by Ferrari et al. is not appropriate. The second item, for example, reads in the original: ‘Is drinking or taking drugs more important than anything else you might do during the day?’; the third item reads: ‘Do you feel your need for drink or drugs is too strong to control?’, where item 2 is to measure ‘salience of substance use’ and item 3, ‘the compulsion to start’. In the questionnaire used in their studies, Ferrari et al. merely replaced the terms ‘drink or drug/s’ by ‘analgesic/s’. Item 3, for example, reads as follows with Ferrari: ‘Do you feel your need for analgesics is too strong to control?’. They obviously presume that these questions are as meaningful and the answers as significant for individuals taking analgesics for chronic pain or even for episodic headaches as for alcohol and opiate addicts. Such an approach is not permissible.
In order to determine whether the LDQ is indeed a suitable instrument for consumers of analgesics, specific validation steps would be necessary. Without verifying the altered items for their validity and reliability, it cannot be used arbitrarily for any other substances or substance groups. For the LDQ, also in the view of its developers, there is only ‘evidence for the validity and reliability as a measure of substance dependence for opiate and alcohol users’, i.e. it is not a suitable instrument for use in consumers of analgesics without any additional validation procedures.