Psychometric evaluation of a consumer-developed family-centred care assessment tool

Author(s): Nora Wells[sup.1] , Suzanne Bronheim[sup.2] , Stephen Zyzanski[sup.3] , Clarissa Hoover[sup.1] Author Affiliations: (1) Family Voices, Inc., Albuquerque, NM, USA (2) Georgetown University, Washington, DC, USA (3) Case Western [...]
The objective of this study was to create a psychometrically sound measure of family-centered care, the Family-Centered Care Assessment (FCCA), developed through a process led by families in collaboration with maternal and child health leaders. The items for the FCCA scale were initially developed by families of children and youth with special needs in partnership with pediatric providers and researchers. Using an Institutional Review Board-approved research protocol, the questions were revised based on input from focus groups of diverse parents in three states. Parental responses (N = 790) to the revised 59-item survey were collected online from families in 49 states. Item distributions uniformly showed excellent spread. A principal axes factor analysis confirmed the existence of a single factor. Rasch modeling item analyses identified a reduced subset of 24 items that demonstrated excellent psychometric properties. All items met the criteria for a linear Rasch scale. Empirical evidence in support of the construct validity of the 24-item measure was derived: all items had a positive and substantial item-total correlation; person alpha scale reliability was >0.80 and the item reliability was >0.90; both separation indices were >2.0; infit and outfit statistics were within 0.5-1.5; and item difficulties ranged between -2 and +2 logits. Strong rank-ordered associations and large effect sizes were observed for six indicators of quality of care. This study's family-led process produced a tool, the FCCA, to measure families' experience of care with excellent psychometric properties.