Abstract
Sir,
We were pleased to read the editorial by Lipton et al. (1), which exemplifies the steps that are necessary for the development of health-related quality of life assessment tools using the Comprehensive Headache-Specific Quality of Life Questionnaire (CHQQ) questionnaire (2). Thanking Professor Lipton and coworkers for showing interest in our work, we would like to reflect on some points raised by the editorial.
First of all, although CHQQ performed well in the validation study, we still have to assess its performance in various clinical settings. Similarly to most of the existing headache-specific HRQoL instruments, CHQQ was validated in a specialized headache centre, so whether the results of our study can be generalized to population-based samples is still to be determined. We are currently preparing a nationwide headache epidemiology survey and will take this opportunity of field-testing CHQQ on a large representative sample.
The next task in examining the generalizability of the CHQQ is to see how it performs in other languages. As a first step, we prepared an English version of the questionnaire, based on the ISPOR recommendations (3). The English version (which has a 4-week recall period) was pilot-tested on a group of 34 medical students from the English programme of Semmelweis University, Budapest, Hungary. The sample included four native English students and 30 students who had English as their best-learned foreign language (time learning: average 15.5 years, SD 4.8 years). Completing the questionnaire took 5 minutes on average (SD: 2.4), and generally was easy (30 rated it as not at all difficult, three as slightly and one as moderately difficult). Most of the students felt the questions were moderately (13 students) or very (14 students) relevant. Interestingly, of the seven students who felt that the questions were only slightly relevant or not relevant at all, four came from Far Eastern countries (there were six Far Eastern students in the whole group), suggesting that cultural differences may be relevant in perceiving the usefulness of the questionnaire. Of those 28 students who actually had headaches during the previous year, 12 considered the questionnaire as very relevant, another 12 as moderately relevant, three slightly relevant and one not relevant. (Of the latter four, two were Far Eastern.) Headache diagnoses were episodic tension-type headache (17 students), migraine without aura (three students), migraine with aura (one student) and probable migraine (two students). Diagnosis was not available for five students. The 4-week English version is now ready for evaluation.
In order to facilitate the evaluation process and contribute to the many collaborative efforts that the field of headache research is fortunate to have, we would like to offer the CHQQ to the headache specialists’ community for further evaluation and use. In order to do so, we have launched a website, http://headache-questionnaire.com, where both online and downloadable versions of the CHQQ are ready for trial.
Lipton et al. point out that although our choice of using a 2-week recall period may provide more accurate recall, it is probably prone to temporal sampling error. We fully agree that a 2-week period may not be sufficient to capture the patients’ headache experience, especially if the headache frequency is low. Moreover, as many instruments use a 4-week recall period, using the CHQQ for research purposes would also benefit for setting a similar recall period. Therefore we recommend using a 4-week questionnaire in further studies – this version is now being tested in chronic migraine patients, and also in patients with medication overuse headache.
The editorial also makes reference to our poster about using IRT methods during the development of CHQQ (4). Although we did use IRT methods during the preparation phase of the CHQQ, the poster, based on data from 117 migraineurs, served mostly to illustrate the usefulness of IRT in test development. Besides reporting exploratory factor analysis data, the poster illustrated the use of item characteristic curves, reliability assessment and model fit calculations. A pdf version of the poster is available at the questionnaire website.
We hope that, by making the CHQQ available for further study, we can encourage collaborations that might benefit headache sufferers and the headache research community as well.