Comparing symptom ratings by staff and family carers in residents dying in long-term care facilities in three European countries, results from a PACE-survey.

CONTEXT: Symptom management is essential in the end of life care of long-term care facility residents.

OBJECTIVES: To study discrepancies and possible associated factors in staff and family carers' symptom assessment scores for residents in the last week of life.

METHODS: A post mortem survey in Belgium, the Netherlands and Finland: staff and family carers completed the "End-Of-Life in Dementia - Comfort Assessment in Dying" scale (EOLD-CAD), rating 14 symptoms on a 1 to 3-point scale. Higher scores reflect better comfort. We calculated mean paired differences in symptom, subscale and total scores at a group level and interrater agreement and percentage of perfect agreement at a resident level.

RESULTS: Mean staff scores significantly reflected better comfort than those of family carers for the total EOLD-CAD (31.61 versus 29.81; P-value < 0.001) and for the physical distress ( 8.64 vs 7.62; P-value < 0.001) and dying symptoms (8.95 vs 8.25; P-value < 0.001) subscales. No significant differences were found for emotional distress and well-being The largest discrepancies were found for "gurgling", "discomfort", "restlessness" and "choking" for which staff answered "not at all" whereas the family carer answered "a lot", in respectively 9.5, 7.3, 6.7 and 6.1% of cases. Interrater agreement κ ranged from 0.106 to 0.204, the extent of perfect agreement from 40.8 for lack of serenity to 68.7 % for crying.

CONCLUSION: There is a need for improved communication between staff andfamily and discussion about symptom burden in the dying phase in long-term care facilities.