Development of valid and reliable measures of patient and family experiences of hospice care for public reporting

Background: Increasingly, dying patients and their families have a choice of hospice providers. Care quality varies considerably across providers; informing consumers of these differences may help to improve their selection of hospices. Objective: To develop and evaluate standardized survey measures of hospice care experiences for the purpose of comparing and publicly reporting hospice performance. Design: We assessed item performance and constructed composite measures by factor analysis, evaluating item-scale correlations and estimating reliability. To assess key drivers of overall experiences, we regressed overall rating and willingness to recommend the hospice on each composite. Setting/Subjects: Data submitted by 2500 hospices participating in national implementation of the Consumer Assessment of Healthcare Providers and Systems (CAHPS®) Hospice Survey for April through September 2015. Measurements: Composite measures of Hospice Team Communication, Getting Timely Care, Treating Family Member with Respect, Getting Emotional and Religious Support, Getting Help for Symptoms, and Getting Hospice Care Training. Results: Cronbach's alpha estimates for the composite measures range from 0.61 to 0.85; hospice-level reliability for the measures range from 0.67 to 0.81 assuming 200 completed surveys per hospice. Together, the composites are responsible for 48% of the variance in caregivers' overall ratings of hospices. Hospice Team Communication is the strongest predictor of overall rating of care. Conclusion: Our analyses provide evidence of the reliability and validity of CAHPS Hospice Survey measure scores. Results also highlight important opportunities to improve the quality of hospice care, particularly with regard to addressing symptoms of anxiety and sadness, discussing side effects of pain medicine, and keeping family informed of the patient's condition.