Methodological aspects of prophylactic drug trials in migraine

The purpose of this paper is to analyse methodological aspects of prophylactic drug trials in migraine. A study of ferrum quartz in 33 patients provided the necessary data base. Migraine definitions, relation between interval headache and migraine, and a number of other clinical problems are discussed. A headache diary is presented which to a certain extent allows a separation of internal headache and migraine attacks. Virtually all patients with frequent migraine attacks have interval headaches. A statistical model which allows a separation of time effects and treatment effects is presented. The inter-patient variability was much greater than intra-patient variability. This indicates that it will be difficult to obtain sufficient power with a non-crossover design. The relation between duration of treatment periods and the necessary number of patients is shown to be inversely related.