Background: The "Surprise Question" (SQ) is often used to identify patients who may benefit from a palliative care approach. The time frame of the typical question (a 12-month prognosis) may be unsuitable for identifying residents in nursing homes since it may not be able to differentiate between those who have a more imminent risk of death within a cohort of patients with high care needs.
Objective: To examine the accuracy and acceptability of 3 versions of the SQ with shortened prognostication time frames (3 months, 6 months, and "the next season") in the nursing home setting.
Design: A prospective mixed-methods study.
Setting/participants: Forty-seven health-care professionals completed the SQ for 313 residents from a nursing home in Ontario, Canada. A chart audit was performed to evaluate the accuracy of their responses. Focus groups and interviews were conducted to examine the participants' perspectives on the utility of the SQ.
Results: Of the 301 residents who were included in the analysis, 74 (24.6%) deaths were observed during our follow-up period. The probability of making an accurate prediction was highest when the seasonal SQ was used (66.7%), followed by the 6-month (58.9%) and 3-month (57.1%) versions. Despite its high accuracy, qualitative results suggest the staff felt the seasonal SQ was ambiguous and expressed discomfort with its use.
Conclusion: The SQ with shortened prognostication periods may be useful in nursing homes and provides a mechanism to facilitate discussions on palliative care. However, a better understanding of palliative care and increasing staff's comfort with prognostication is essential to a palliative care approach.
Keywords: advance care planning; clinical decision-making; end-of-life care; long-term care; nursing homes; palliative care.