Response Bias in Hospice Evaluation

Abstract

As a relatively new means of health care in the United States, hospices are continually seeking feedback to disclose ways in which they may provide more effective palliative services for patients and their families. An important source of such information has been caregiver evaluations, since these surveys reflect the needs for services of respondents and satisfaction with care. Bass found significant differences between caregivers who responded in hospice evaluation efforts and those who did not; nonrespondents cared for patients who entered the hospice in a more incapacitated condition, had more nursing visits, and returned to a hospital or other facility to die. In an attempt to cross-validate Bass' findings, an analysis of response bias was carried out among thirty-four recipients of care in a home care based hospice in the southwest. Nonrespondents were found to have better bereavement prognoses and tended to care for patients who were younger, male, and who were in the program for a shorter length of time. Nonrespondents had also been in contact with the staff less than had respondents. While this study only in part supports Bass' findings, its data are consistent with the conclusion that there are significant differences between caregiver respondents and nonrespondents. Consequently, hospices may unknowingly underserve nonrespondents and their patients if they do not take measures to more aggressively investigate their needs.