Psychometric properties of the Zarit Caregiver Burden Interview administered to caregivers to patients with Duchenne muscular dystrophy: a Rasch analysis

Purpose: To explore the psychometric properties of the full 22-item English (UK and US) version of the Zarit Caregiver Burden Interview administered to caregivers to patients with Duchenne muscular dystrophy. Materials and methods: Caregivers to patients with Duchenne muscular dystrophy from the United Kingdom and the United States, recruited through the TREAT-NMD network, completed the Zarit Caregiver Burden Interview online. The psychometric properties of the Zarit Caregiver Burden Interview were examined using Rasch analysis. Results: A total of 475 caregivers completed the Zarit Caregiver Burden Interview. Model misfit was identified for 9 of 22 items (mean item fit residual 0.061, SD: 2.736) and 13 of 22 items displayed disordered thresholds. The overall item-trait interaction chi-square value was 499 (198 degrees of freedom, p < 0.001). The mean person fit residual was estimated at -0.213 (SD: 1.235). The Person Separation Index and Cronbach's alpha were estimated at 0.902 and 0.914, respectively. Item dependency was low and we found no significant differential item functioning by country or sex. Conclusion: Our Rasch analysis shows that the Zarit Caregiver Burden Interview fails to fully operationalize a quantitative conceptualization of caregiver burden among caregivers to patients with Duchenne muscular dystrophy from the United Kingdom and the United States. Further research is needed to understand the psychometric properties of the Zarit Caregiver Burden Interview in other populations and settings.