The occurrence of out-of-hospital cardiac arrest (OHCA) in children is rare, with an estimated annual incidence in the United States in 2016 of 8–38 per 100,000 child-years (1). Even so, it does have great consequence with high rate of mortality and morbidity, both short-term and long-term (1–4). For example, when a child is awake soon after return of circulation (ROC) is achieved, crude outcome is most likely favorable with a great relief for parents, siblings, and the medical team. At the other end of the spectrum, the resuscitated child may progress to death by neurologic criteria (DNC) (otherwise called brain death) (5). In those children who remain comatose (with preserved brainstem reflexes) after achieving ROC, predicting prognosis remains a challenge. In adult patients, there are international guidelines for neuroprognostication after cardiac arrest (CA) (6). However, at present, there is no such guidance for pediatric patients. Against this background, we know that in some instances life-sustaining therapies are withdrawn, and in other similar settings they are not. Together, these practices mean withdrawal of life-sustaining therapies (WLSTs) may create a “self-fulfilling” outcome, that is, the occurrence of death after WLST reinforces the presumed relationship between clinical neuroprognostic preconditions and the inevitable outcome of death (7). Therefore, understanding the process of neuroprognostication and WLST in children after CA across different countries is important.

Previous research in a U.S. single-center cohort of 191 consecutive pediatric cases (2005–2013) showed that mortality before hospital discharge after achieving ROC was 45% (7). DNC was the most common cause of death (47%) and 34% died after WLST for unfavorable neurologic prognosis; the timing of WLST ranged from 1 to 29 days. In our single-center Dutch cohort of 113 pediatric cases (2012–2017), PICU mortality of children with ROC after OHCA was 56% (8). In the 62 deaths, two-thirds died after WLST based on an expected unfavorable neurologic prognosis, and the other third died after the determination of DNC. In more than half of the children who died after WLST, the WLST decision was made within 24 hours after PICU admission.

In this issue of Pediatric Critical Care Medicine, Vassar et al (9) report their retrospective observations of WLST in a two-center U.S. cohort of 135 children with OHCA over a 5-year period (2016–2020). The authors also describe the number of children who survived with delayed awakening despite being comatose on day 3 post-CA. There were 63 of 135 (47%) who died within the hospital: of these, 34 died after WLST due to poor perceived neurologic prognosis, and 14 died after the determination of DNC. In those who died after WLST, 20 of 34 died within 3 days of PICU admission. Another important finding was that seven of 72 children who survived to hospital discharge were comatose on day 3 post-CA and, interestingly, they awakened during follow-up. Among these seven children, four children had a Pediatric Cerebral Performance Category (PCPC) score of 4 (with severe motor impairment), and three children had a PCPC score of 3 at 1 year after CA.

This finding underlines the importance of caution when deciding about neuroprognostication during the first 3 days after OHCA. This finding is also consistent with the 2019 scientific statement from the American Heart Association about pediatric post-CA care, which recommends waiting at least 72 hours before prognosticating in comatose children after OHCA (8). So, the question that arises from these data are, what approach should practitioners take in neuroprognostication? For example, is a multimodal neuromonitoring approach warranted in all comatose children after CA, which would include repeated neurologic examinations, early continuous electroencephalography (up to 24–48 hr post-CA), and brain MRI on days 3–5 post-CA (8,10). Of note, in this new cohort, brain MRI was performed in only 34% of children who died after WLST. Among patients who died less than or equal to 3 days post-CA, 76% underwent electroencephalography monitoring. Vasser et al (9) did not report the findings of these tests, which is crucial to better understand the grounds on which the WLST decision was made. Similar arguments can be made about the data used in those remaining comatose on day 3, but the opposite question: on what grounds did the clinicians decide to continue treatment? Were brainstem reflexes present, did the electroencephalography show normal background activity, or did the MRI reveal greater/lesser ischemic changes on diffusion-weighted imaging?

The PCPC outcome data in the seven survivors who were initially comatose in the current study by Vassar et al (9) should be treated as preliminary information because of the sample size. That said, no comatose survivor in the present study had a PCPC of 1 or 2, suggesting that when persistently comatose during the first 3 days after OHCA, the associated outcome in survivors is poor.

Taking all the above together, it is our view that International Guidelines are needed for neuroprognostication in children achieving ROC after OHCA. To achieve this, international collaboration is needed. This task will also require interpreting the value of various neuromonitoring modalities, clinical assessments, and investigations. As a starting point, we consider that the data necessary for such an approach should be standardized to include the following: 1) the uniform collection of individual patient premorbid information such as comorbidity; 2) the extent of post-CA care, such as temperature management, blood pressure, Paco2, and oxygen targeting; 3) the timing of intervals at which the neurologic examination should be performed, and how and when to interpret findings free of confounders such as sedation and hypothermia; 4) the minimum set of what and when investigations should be used post-CA (e.g., continuous electroencephalography, brain MRI); and 5) the outcome data in survivors, across domains and over at least 1-year post-CA. Until we have such guidance and information, there remains a risk of decision bias in use of WLST in the post-ROC in the OHCA pediatric patient.

1. Fink EL, Prince DK, Kaltman JR, et al.; Resuscitation Outcomes Consortium: Unchanged pediatric out-of-hospital cardiac arrest incidence and survival rates with regional variation in North America. Resuscitation. 2016; 107:121–128 2. Hunfeld M, Dulfer K, Rietman A, et al.: Longitudinal two years evaluation of neuropsychological outcome in children after out of hospital cardiac arrest. Resuscitation. 2021; 167:29–37 3. Hickson MR, Winters M, Thomas NH, et al.: Long-term function, quality of life and healthcare utilization among survivors of pediatric out-of-hospital cardiac arrest. Resuscitation. 2023; 187:109768 4. Ng ZHC, Ho SJ, Caleb T, et al.: Long-term outcomes after non-traumatic out-of-hospital cardiac arrest in pediatric patients: A systematic review. J Clin Med. 2022; 11:5003 5. Lewis A, Kirschen MP: Brain death/death by neurologic criteria determination. Continuum (Minneap Minn). 2021; 27:1444–1464 6. Nolan JP, Sandroni C, Bottiger BW, et al.: European Resuscitation Council and European Society of Intensive Care Medicine guidelines 2021: Post-resuscitation care. Intensive Care Med. 2021; 47:369–421 7. Hunfeld M, Ketharanathan N, Catsman C, et al.: A systematic review of neuromonitoring modalities in children beyond neonatal period after cardiac arrest. Pediatr Crit Care Med. 2020; 21:e927–e933 8. Topjian AA, de Caen A, Wainwright MS, et al.: Pediatric post-cardiac arrest care: A scientific statement from the American Heart Association. Circulation. 2019; 140:e194–e233 9. Vassar R, Mehta N, Epps L, et al.: Mortality and Timing of Withdrawal of Life-Sustaining Therapies After Out-of-Hospital Cardiac Arrest: Two-Center Retrospective Pediatric Cohort Study. Ped Crit Care Med. 2024; 25:241–249 10. Slovis JC, Bach A, Beaulieu F, et al.: Neuromonitoring after pediatric cardiac arrest: Cerebral physiology and injury stratification. Neurocrit Care. 2023 Apr 1. [online ahead of print]