Measuring hemophilia caregiver burden: validation of the Hemophilia Caregiver Impact measure

Carolyn E. Schwartz, Victoria E. Powell, Adi Eldar-Lissai

The purpose of this article is to describe the psychometric development of the Hemophilia Caregiver Impact measure. Qualitative interviews (n = 22) and a cross-sectional web-based study (n = 458) were implemented with caregivers of people with hemophilia. Classical test theory and item response theory analyses were implemented to evaluate the psychometric characteristics of the measure. The study sample had a mean age of 39 and a median level of college education. It was predominantly female (88%), and had an average of two children. 85% of this study sample had at least one child with hemophilia. The final 36-item Hemophilia Caregiver Impact measure is composed of seven subscales assessing relevant negative aspects of caregiver impact (Burden Summary) as well as one subscale reflecting a positive aspect of caregiver impact (Positive Emotions). These two summary scores are orthogonal and can be used together in analyses examining negative and positive aspects of caregiver impact. The items included within each subscale reflect a unidimensional construct, demonstrate good item information and trace lines, and lack of local dependence. The resulting subscales demonstrate high reliability, and good construct validity. They show moderate incremental and discriminant validity. The Hemophilia Caregiver Impact measure is a useful new tool for clinical research on hemophilia. In addition to having eight relevant subscales, the measure can also be summarized with two scores. This versatility can be useful in analyzing studies with very small samples, which is to be expected when dealing with a rare condition like hemophilia.