Reliability of an illness severity measure for headache in a population sample of migraine sufferers

To examine the reliability and related measurement properties of an illness severity measure for headache derived from responses to a 16-item self-administered questionnaire, the Headache Impact Questionnaire (HImQ), and to determine if there is support for combining measures of pain and disability into a single scaled measure of severity.

A population-based sample of migraine headache sufferers completed the HImQ twice, an average of 38 days apart. The HImQ included questions about number of headaches in the last 3 months, headache duration, last headache, pain intensity (two questions), need for bedrest (two questions), disability in specific domains of activity (seven questions about interference with ability to work, do household chores, and engage in non-work activity), and symptoms (two questions).

Test-retest correlations of individual questions ranged from 0.65 to 0.93. In principal components analysis, a single factor with significant loading emerged. One measure of pain intensity (average pain score from 0 to 10) and items related to disability (i.e., missed days due to headache, and reduced effectiveness because of headache) in defined activity domains (work for pay, housework, nonwork activities) had the greatest weights on this single factor, supporting prior work on combining measures of pain and disability into a single scale. The HImQ score was derived as the sum of average pain intensity and total lost time in each of the three domains of activity, expressed as lost days. The latter was derived as the sum of actual missed days in each activity domain and reduced effectiveness day equivalents in each activity with a headache. The test-retest correlation was 0.77 for all eligible subjects and 0.85 when one influential outlier was removed. Cronbach's alpha was 0.83.

The HImQ score, based on eight items, is a highly reliable and internally consistent measure of headache severity.