Psychometric Evaluation of the Activity Impairment in Migraine-Diary (AIM-D) Using Data From a Phase 3 (PROGRESS) Clinical Trial in Patients With Chronic Migraine
Pranav Gandhi1, Richard Lipton2, David Dodick3, Diane Whalley4, Lauren Nelson4, Stuart Yarr4, Jonathan Stokes1
1AbbVie, 2Albert Einstein College of Medicine, 3Mayo Clinic, 4RTI Health Solutions
Objective:
To evaluate the psychometric properties and establish meaningful within-patient change (MWPC) thresholds of the Activity Impairment in Migraine-Diary (AIM-D) Performance of Daily Activities (PDA) and Physical Impairment (PI) domain scores, using data collected in a chronic migraine (CM) trial. 
Background:

The AIM-D is a novel, 11-item patient-reported outcome (PRO) daily diary measure comprising 2 domains: PDA (7 items) and PI (4 items) based on a 24-hour recall.

Design/Methods:
In a phase 3, 12-week, double-blind controlled trial (NCT03855137), participants with chronic migraine (CM) were randomized to receive atogepant 30 mg BID, 60 mg QD, or placebo. Based on pooled data, psychometric properties were evaluated using confirmatory factor analysis (CFA), assessments of reliability, validity (convergent and known-groups), responsiveness, and estimation of MWPC thresholds.
Results:

Of the 778 participants enrolled, the analysis included 682. CFA confirmed the 11-item, 2-domain factor structure (PDA and PI). Both domains demonstrated good internal consistency (Cronbach’s alpha: ≥ 0.90), and excellent test-retest reliability (intraclass correlation coefficients: >0.90). Convergent/divergent validity was supported by moderate/strong correlations between domain scores and migraine/headache days (0.30-0.48) and activity level/limitation (0.53-0.94). PDA and PI scores significantly differentiated based on Patient Global Impression of Severity (PGI-S) (P < .0001) and PDA and PI change scores differed significantly across levels of change on PGI-S, Patient Global Impression of Change (PGIC), and percent monthly migraine days. Anchor-based results demonstrated MWPC estimates in the range of -13.1 to -14.3 points for the PDA domain and −10 points for the PI domain as clinically meaningful.

Conclusions:

The AIM-D psychometric properties were confirmed and MWPC estimates were established using phase 3 CM trial data. The AIM-D is a robust and valid measure to assess treatment benefit related to improvement in performance of daily activities and reduction in physical impairment in clinical trials. 

10.1212/WNL.0000000000203615