In a retrospective analysis of the records of 500 patients with “primary headache” (i.e. unknown etiology) it was found thattreatments prescribed by the clinic physicians varied widely, seemed not to be closely correlated with the subtype of headache (e.g. migraine, muscle contraction), and seemed not to reflect the recommendations given in controlled studies in the scientific literature. A very large number of drugs were used, often in combinations, and some had scant or no rationale for use for the indication of headache. Efficacy varied, but over time only 30% seemed to maintain a degree of effectiveness.
Many drug studies in the literature are artefactually clear cut in terms of composition of patients with “pure” headache subsets, and include a relatively short follow-up. A more realistic appraisal of a drug's everyday efficacy could be obtained by utilizing patient populations which more accurately reflect reality, and by following them up for prolonged periods.