AIMS: To evaluate psychometric properties of the MPFID, a patient-reported outcomes (PRO) instrument assessing the impact of migraine on physical functioning.
METHODS: Psychometric properties of the MPFID were assessed using data from an observational study among subjects with a history of C4 migraine days/month (for 3 months before screening) receiving standard migraine care. Migraine headaches, symptoms, and impacts (including MPFID items) were collected daily for 1 month using an electronic diary. Data were also collected using Patient Global Impression of Severity (PGIS), Patient Global Impression of Change (PGIC), PROMIS Physical Function Short Form (PROMIS-PF), Headache Impact TestTM (HIT-6TM), and Migraine-Specific Quality of Life Questionnaire (MSQ). Item analysis and item reduction were conducted, along with exploratory factor analysis (EFA), followed by tests of reliability (internal consistency, test–retest) and validity (convergent, known-groups).
RESULTS: The sample included 342 subjects. Mean age was 40.1 years (SD = 11.7), and 87.2 % were female. Item-level analyses informed item reduction, and 4 of the 17 items were dropped. EFA indicated two domains: Impact on Everyday Activities (EA) and Physical Impairment (PI). Daily domain scores range from 0 to 100 (higher scores indicate greater impact) and are averaged across a month. Both domains exhibited high internal consistency reliability (EA: a = 0.96 and PI: a = 0.92). Test–retest reliability (among stable subjects based on PGIS and PGIC) was acceptable (ranges of intraclass correlation coefficient—EA: 0.75–0.82; PI: 0.77–0.86). Convergent validity was demonstrated by moderate correlations (all p\0.0001) between MPFID domain scores and number of headache days (r = 0.57–0.59) and migraine days (r = 0.64–0.65), PROMIS-PF (r = -0.61 to -0.64), HIT-6 (r = 0.50–0.55), and MSQ domains (Role-Restrictive: r = -0.63 to -0.68; Role-Preventive: r = -0.65 to -0.68; Emotional Function: r = -0.51 to -0.58). Analysis of covariance models exploring known-groups validity showed that both MPFID domains differentiated between groups of subjects who varied with respect to migraine days, migraine interference levels, and migraine pain intensity, and median split groups for PROMIS-PF, HIT-6, and MSQ domains (all p is less than or equal to 0.05).
CONCLUSIONS: This study yielded good evidence of the psychometric properties of the MPFID. Findings support two distinct domains, both with evidence of excellent reliability and construct validity in assessing impacts of migraine on physical functioning.
Share on: