Impact of preventive pill-based treatment on migraine days: A secondary outcome study of the Childhood and Adolescent Migraine Prevention (CHAMP) trial and a comparison of self-report to nosology-derived assessments.

Abstract

To examine group differences in self-reported migraine days among youth who completed the Childhood and Adolescent Migraine Prevention (CHAMP) trial prior to its closure and explore the relationship between self-reported and "nosology-derived" (i.e., International Classification of Headache Disorders, 3rd edition [ICHD-3]) migraine days. The CHAMP trial compared amitriptyline and topiramate to placebo for migraine prevention in youth and proposed to analyze change in migraine days as a secondary outcome. There is considerable variability in the field regarding what constitutes a "migraine day," how this is determined and reported in trials, and how consistent these measures are with diagnostic nosology. CHAMP trial completers (N=175) were randomized to receive amitriptyline (n=77), topiramate (n=63), or placebo (n=35). Participants maintained daily headache diaries where they reported each day with headache and if they considered that headache to be a migraine. For each headache day, participants completed a symptom record and reported about symptoms such as pain location(s) and presence of nausea/vomiting or photophobia and phonophobia. We examined group differences in self-reported migraine days at trial completion (summed from trial weeks 20-24) compared to baseline. We also used an algorithm to determine whether participants' symptom reports met ICHD-3 criteria for migraine without aura, and examined the association between self-reported and "nosology-derived" migraine days. Results showed no significant differences between groups in self-reported migraine days over the course of the trial. Self-reported and "nosology-derived" migraine days during the baseline and treatment phases were strongly associated (r's=0.73 and 0.83, respectively; p's < 0.001). Regardless of treatment, CHAMP trial completers showed clinically important reductions in self-reported migraine days over the course of the trial (about 3.8 days less). The strong association between self-reported and "nosology-derived" migraine days suggests youth with migraine can recognize a day with migraine and reliably report their headache features and symptoms. Greater rigor and transparency in the calculation and reporting of migraine days in trials is needed.