Developing an electronic Surprise Question to predict end-of-life prognosis in a prospective cohort study of acute hospital admissions.

Abstract

Objective: Determining the accuracy of a method calculating the Gold Standards Framework Surprise Question (GSFSQ) equivalent end-of-life prognosis amongst hospital inpatients.

Design: A prospective cohort study with regression calculated 1-year mortality probability. Probability cut points triaged unknown prognosis into the GSFSQ equivalent "Yes" or "No" survival categories (> or < 1-year respectively), with subsidiary classification of "No". Prediction was tested against prospective mortality.

Setting: An acute NHS hospital.

Participants: 18,838 acute medical admissions.

Interventions: Allocation of mortality probability by binary logistic regression model (X²=6650.2, p<0.001, r² = 0.43) and stepwise algorithmic risk-stratification.

Main outcome measure: Prospective mortality at 1-year.

Results: End-of-life prognosis was unknown in 67.9%. The algorithm's prognosis allocation (100% vs baseline 32.1%) yielded cohorts of GSFSQ-Yes 15,264 (81%), GSFSQ-No Green 1,771 (9.4%) and GSFSQ-No Amber or Red 1,803 (9.6%). There were 5,043 (26.8%) deaths at 1-year. In Cox's survival, model allocated cohorts were discrete for mortality (GSFSQ-Yes 16.4% v GSFSQ-No 71.0% (p<0.001). For the GSFSQ-No classification, the mortality Odds Ratio was 12.4 (11.4 - 13.5) (p<0.001) vs GSFSQ-Yes (c-statistic 0.72 (0.70 - 0.73), p<0.001; accuracy, positive and negative predictive values 81.2%, 83.6%, 83.6%, respectively). Had the tool been utilised at the time of admission, the potential to reduce possibly avoidable subsequent hospital admissions, death-in-hospital, and bed days was significant (p<0.001).

Conclusion: This study has unique in methodology with prospectively evidenced outcomes. The model algorithm allocated GSFSQ equivalent EOL prognosis universally to a cohort of acutely admitted patients with statistical accuracy validated against prospective mortality outcomes.