Hopelessness at the end of life: the utility of the hopelessness scale with terminally ill cancer patients.

OBJECTIVES: We sought to investigate the utility of the Beck Hopelessness Scale (HS) in a sample of terminally ill cancer patients by examining the scale properties. Moreover, we sought to identify and remove potentially problematic items in order to ascertain a "purer" index of hopelessness for this population.
DESIGN: A cross-sectional study of 200 hospice inpatients with a life expectancy of less than 6 months. The HS, as well as several other distress measures, were administered to patients at bedside by trained clinicians.
METHODS: An item analysis of the HS was conducted, looking specifically at item endorsement and item-total correlations. Three abbreviated versions (3-item 7-item, 13-item) were developed based on certain denoted item-total correlation cut-offs. Reliability and validity of the original 20-item HS was then compared to that of the newly developed abbreviated version.
RESULTS: All scales were found to be reliable and valid measures of hopelessness. The three abbreviated versions were more highly correlated with the distress measures than the original version, and the 7-item and 13-item subscales outperformed the original HS in the prediction of suicidal ideation and desire for hastened death.
CONCLUSION: The data suggest that the HS may be improved, when applied to a terminally ill sample, by the elimination of problematic items. The development of a shorter, purer measure of hopelessness for this population is crucial given the need to reduce the burden placed on those who participate in end-of-life studies, and the important role of hopelessness in the prediction of suicide and desire for hastened death.