Predictive model for the Migraine-ACT questionnaire in primary care

BACKGROUND: Migraine is frequently under-treated. The 4-item Migraine Assessment of Current Therapy (Migraine-ACT) questionnaire is a simple, patient-friendly tool to identify patients requiring change in acute migraine treatment. OBJECTIVE: To establish a predictive model for the Migraine-ACT questionnaire in primary care. RESEARCH DESIGN AND METHODS: This open-label, prospective, two-visit study at 1100 primary care sites enrolled patients reporting migraine for > 1 year and > or = 1 migraine attack/month. Predictive validity of the Migraine-ACT questionnaire was evaluated using a logistic regression model adjusted for baseline disability, diagnosis, sex, age, attack frequency, attack duration, baseline treatment change, treatment(s) for the study period and baseline Migraine-ACT score. RESULTS: A total of 3272 patients, 78% female, entered the study and 2877 patients (88%) returned at 3 months. Investigators changed baseline migraine treatment for 72% of returning patients. Percentages of patients who were completely or very satisfied with migraine therapy increased from 15% at baseline to 49% at 3 months. The adjusted logistics regression model yielded sensitivity of 86% and specificity of 85% to detect > or = 1-point change in Migraine-ACT score at 3 months (p < 0.001). The model yielded 77% sensitivity and 95% specificity to detect > or = 1-point improvement in patient satisfaction at 3 months among patients who improved > or = 1 point in the Migraine-ACT score (adjusted odds ratio, 2.63; 95% confidence interval, 1.97-3.51; p < 0.001). CONCLUSIONS: The Migraine-ACT questionnaire is a simple but sensitive and specific tool to predict improvements at 3 months resulting from changes in migraine treatment and can be used to detect patients with suboptimal migraine management and to monitor treatment effectiveness.