Analysis of the Surprise Question as a tool for predicting death in neonates

Abstract Purpose The Surprise Question (SQ) "Would you be surprised if the patient died in the next 12 months?" lacks pediatric research, particularly in neonatal patients. Our study aims to analyze the SQ’s predictive ability in neonates and explore pediatricians’ views on palliative care (PC) patient identification. Methods A prospective cross-sectional study was conducted from February 2021 to June 2023, including all newborns admitted to the Neonatal Intensive Care Unit of a pediatric tertiary hospital and its pediatricians. Patients with less than a year since admission were excluded from final analyses. Recorded variables included patient demographics and condition, pediatricians’ profile and opinions regarding the SQ and PC patient identification. The SQ was formulated to one or more pediatricians per neonate at admission, 7 days and 28 days of life, with patient status recorded after twelve months to elaborate a confusion matrix of prognostic test results. Results A total of 51 pediatricians participated. Most felt they had limited criteria for identifying PC patients (55%), believed the SQ could be useful (77%) and predict death (75%). The SQ was answered at least in one out of the three moments for 262 neonates (61% male and at least 35,5% preterm), with sufficient sample at each moment to study its predictive ability. High negative predictive values were consistent, with higher positive predictive value at 7 days (25,7%). Conclusions The SQ is a promising tool for predicting neonatal outcome and could guide professionals in initiating PC discussions. The 7-day mark appears more suitable for its application.