Abstract
Dear Sir we are grateful for the attention Bogduk (1) gives to our paper on radiofrequency treatment of cervicogenic headache (2), although the vehement reaction to our study is surprising, particularly since we seem to agree on several important issues. As we see it, the vehemence is nurtured by the fact that Bogduk argues against interpretations of our data which are not warranted and which we have not made. In particular, we have not taken our study ‘as an indictment of zygoapophyseal joint blocks or medial branch neurotomy’. On the contrary, we conclude that the procedure is not promising for patients fulfilling purely clinical criteria for cervicogenic headache, an opinion that seems to be shared by Bogduk. Many of the issues he raises are discussed in our article, but some additional comments are obviously needed.
One criticism is that we did not use complete blockade effect as a standard criterion for inclusion. Cervicogenic headache, according to the definitions cited above, is understood as a syndrome that can result from many possible sources in the neck, and conceivably, not all of these can be blocked. Therefore, the diagnosis of cervicogenic headache, at least from a theoretical point of view, may be appropriate even in cases where a clearcut blockade effect is lacking. This is apparently also acknowledged in the ICHD-2 criteria for cervicogenic headache from 2004 (3), in which blockade effect is not an obligatory criterion. Therefore, it was at that time, and it still is, of considerable interest to try to validate the criterion of blockade effects against treatment outcome. This could only be done by including patients on the basis of purely clinical criteria and then performing blockades with accurate measurements of effect.
Bogduk argues that only complete relief of pain after the blockade can be used as a basis for a valid diagnosis of cervicogenic headache. Although we agree that 100% blockade effect is a priori a better predictor of therapeutic success than a partial effect, it may be fruitful to discuss what is a complete blockade effect. In the International Spinal Injection Society practice guidelines, this is discussed in a note where it is admitted that something less than 100% effect may be a sufficient indication for complete relief. This is also acknowledged in the ICHD-2 classification on cervicogenic headache defining abolition of headache as ≥ 90% reduction in pain. We consider it a merit of our study, in contrast to many other studies on blockades, that we meticulously described the blockade effect in percentage. In the light of our results, we conclude the paper by recommending a consistent and close to 100% effect of blockades as a criterion for inclusion in later studies with the technique, which we understand is in accordance with Bogduk's view.
As to our surgical technique, Bogduk admits that it is adequate according to the description, but he suspects that it is nevertheless not done properly since the entire operation lasted only 90 min. In response to this, the surgeon (FK) who did all the procedures, can only assure that it was done exactly as described. The technique was learnt in another European centre where the operation had approximately the same duration as described here.
We feel that our study is both very relevant and important as radiofrequency treatment of facet joints in the neck is routinely performed in many centres on similar indications as in our study. A previous open study reported a positive outcome using the same type of criteria and similar radiofrequency technique (4). The response reported in this study is similar to that in our sham patients, indicating that this was a placebo response.
It is obviously common that fellow researchers may be critical of a study protocol, but the very negative tone in Bogduk's comments to our study is not conductive to constructive discussion, and one is left with the impression that he feels the study should not have been published at all. This raises some important and fundamental questions since it from a scientific point of view is as important to report studies with a negative outcome as to publish those with positive outcome. Because our results are somewhat different from earlier studies in this field, we have been careful to report the clinical features and the blockade and treatment outcome of all our patients in greater detail than is usual, to ensure full transparency about the data on which our conclusions are based. Bogduk has not been able to show that we have made unwarranted inferences from our data, although he fears that the results can be misinterpreted. As he is an important researcher in this field, it would be highly regrettable if he could use his influence to discourage publication of studies with a negative outcome, even if he is critical to some aspects of the study protocol. We are therefore grateful that the editor has accepted our study for publication in spite of Bogduk's harsh criticism.