| Report cards documenting quality of care and satisfaction with services can inform consumer decision-making and encourage providers to improve quality. The purpose of the present study was to develop and test a questionnaire that could be used to measure the quality of hospice care and to provide performance measures for a hospice report card. The investigator worked with a committee of the American Hospice Foundation to define, and to generate questions for evaluating, the following categories of hospice care: pain/symptom management, communication/education, bereavement/spiritual support, family burden, and patient autonomy. Questions assessing overall satisfaction with services provided the criterion measure. The resulting instrument, called the Hospice Primary Caregiver Questionnaire (HPCQ), was completed by n = 234 primary caregivers who received at-home hospice care from January 1998 through March 1999 at three hospices in Illinois, Virginia, and the District of Columbia. The results indicate that most caregivers evaluate the hospices very highly, giving favorable responses on 90% or more of the questions in each category (except for pain/symptom management). They are likewise very satisfied with the services received: the mean rating on the global satisfaction measure was 4.53 (on a 5-point scale). Two different sets of analyses, so-called favorability analyses and regression by domain analyses, supported the hypothesis that favorable evaluations on the questions defining each category would be positively associated with global satisfaction. The rank order of the categories in predicting satisfaction was (1) bereavement/spiritual support, (2) family burden, (3) communication/education, (4) patient autonomy, and (5) pain/symptom management. In contrast to prediction, however, the responses to questions within categories were not more highly intercorrelated than those across categories: confirmatory factor analysis revealed that the items within categories did not correlate well with each other and were not internally consistent (Cronbach's alphas = .08–.66). Despite this, an overall measure of favorability based on the proportion of favorable responses made on the HPCQ did correlate reasonably well with global satisfaction, r = .71, p < .0001. Additional research is needed to refine the HPCQ with particular attention being given to eliminating ceiling effects and establishing test-retest reliability. |
