Using intuition or a formal palliative care needs assessment screening process in general practice to predict death within 12 months: A randomised controlled trial.

BACKGROUND: Population ageing will lead to more deaths with an uncertain trajectory. Identifying patients at risk of dying could facilitate more effective care planning.

AIM: To determine whether screening for likely death within 12 months is more effective using screening tools or intuition.

DESIGN: Randomised controlled trial of screening tools (Surprise Question plus the Supportive and Palliative Care Indicators Tool for Surprise Question positive patients) to predict those at risk of death at 12 months compared with unguided intuition (clinical trials registry: ACTRN12613000266763).

SETTING/PARTICIPANTS: Australian general practice. A total of 30 general practitioners (screening tool = 12, intuition = 18) screened all patients ( n = 4365) aged ≥70 years seen at least once in the last 2 years.

RESULTS: There were 142 deaths (screening tool = 3.1%, intuition = 3.3%; p = 0.79). General practitioners identified more at risk of dying using Surprise Question (11.8%) than intuition (5.4%; p = 0.01), but no difference with Surprise Question positive then Supportive and Palliative Care Indicators Tool (5.1%; p = 0.87). Surprise Question positive predicted more deaths (53.2%, intuition = 33.7%; p = 0.001), but Surprise Question positive/Supportive and Palliative Care Indicators Tool predictions were similar (5.1%; p = 0.87 vs intuition). There was no difference in proportions correctly predicted to die (Surprise Question = 1.6%, intuition = 1.1%; p = 0.156 and Surprise Question positive/Supportive and Palliative Care Indicators Tool = 1.1%; p = 0.86 vs intuition). Screening tool had higher sensitivity and lower specificity than intuition, but no difference in positive or negative predictive value.

CONCLUSION: Screening tool was better at predicting actual death than intuition, but with a higher false positive rate. Both were similarly effective at screening the whole cohort for death. Screening for possible death is not the best option for initiating end-of-life planning: recognising increased burden of illness might be a better trigger.