Use of an item bank to develop two short-form FAMCARE scales to measure family satisfaction with care in the setting of serious illness

Context Family satisfaction is an important and commonly used research measure. Yet current measures of family satisfaction are lengthy and may be unnecessarily burdensome - particularly in the setting of serious illness. Objectives To use an item bank to develop short forms of the Family Satisfaction with End-of-Life Care (FAMCARE) scale, which measures family satisfaction with care. Methods To shorten the existing 20-item FAMCARE measure, item response theory parameters from an item bank were used to select the most informative items. The psychometric properties of the new short-form scales were examined. The item bank was based on data from family members from an ethnically diverse sample of 1983 patients with advanced cancer. Results Evidence for the new short-form scales supported essential unidimensionality. Reliability estimates from several methods were relatively high, ranging from 0.84 for the five-item scale to 0.94 for the 10-item scale across different age, gender, education, ethnic, and relationship groups. Conclusion The FAMCARE-10 and FAMCARE-5 short-form scales evidenced high reliability across sociodemographic subgroups and are potentially less burdensome and time-consuming scales for monitoring family satisfaction among seriously ill patients.