Reliability of the migraine disability assessment score in a population-based sample of headache sufferers

Background. The Migraine Disability Assessment (MIDAS) score is used to qua ntify headache-related disability. In a previous study, we showed that the MIDAS score was highly reliable in population-based samples of migraine hea dache sufferers in two countries. Objectives. To examine the test-retest re liability and internal consistency of the five items comprising the MIDAS s core and the overall MIDAS score in a population-based sample of both migra ine and nonmigraine headache sufferers. Methods. Using a clinically validat ed telephone interview, a population-based sample of migraine and nonmigrai ne headache sufferers was identified in Baltimore, Maryland, USA. A total o f 97 migraine cases and 80 nonmigraine subjects completed the MIDAS questio nnaire on two occasions an average of 3 weeks apart. The MIDAS score is der ived from five questions about missed time from work (or school) and househ old work (one question each about missed days and days with at least 50% re duced productivity) and missed days of nonwork activities. Results. Among a ll headache sufferers the test-retest Spearman's correlations of individual MIDAS questions ranged from 0.67 to 0.73. The Spearman's correlation for t he MIDAS score (i.e., sum of lost days and reduced effectiveness days in ea ch domain) was 0.84. Cronbach's alpha, a measure of internal consistency, w as 0.83. Mean and median item values and the overall MIDAS scores differed between migraine and nonmigraine cases. Even after adjusting for difference s in headache frequency, the mean MIDAS scores differed substantially (i.e. , 10.3 points) between migraine cases and nonmigraine eases. Conclusions. T he reliability and internal consistency of the MIDAS score are high, as tes ted in a population-based sample of headache sufferers. MIDAS scores are su bstantially higher in migraine cases than in nonmigraine cases, supporting the validity of the measure.