Maayke Hunfeld

236 Chapter 8 Timing and cause of death The study presented in chapter 4 explored the timing and cause of death in 113 children admitted to our PICU following ROC after OHCA between 2012 and 2017. The causes of OHCA were diverse; the most common causes were drowning (21%) and arrhythmia (17%). We found that 56% of the children who achieved ROC had died prior to hospital discharge. These children’s most common cause of death was WLST based on poor neurological prognosis (WLST-Neuro; 67%), whether or not combined with WLST due to refractory circulatory and/or respiratory failure or recurrent CA. In in most cases, the decision had been made within 72 hours after the CPR event and, in half of the cases even within 24 hours. Other causes of death were brain death (BD, 29%), recurrent CA (2%) or refractory circulatory and/or respiratory failure (2%). Only one other study, from the USA, described the causes and timing of death of children admitted to the PICU after OHCA (33). In this cohort, BD was the most common cause of death (47%), while WLST based on poor neurological prognosis was less common (34% vs 67% in our study). The following reasons might explain this discrepancy: First, the definition of BD differed between studies, because the BD criteria differed between the Netherlands and the USA. For that matter, up to now, there are no worldwide consensus criteria for BD. Second, the children in our study may initially have been in a better condition after ROC. If so, this could be due to the setting in the Netherlands: a higher occurrence of automatic external defibrillator (AED) use and bystander CPR, the availability of helicopter emergency medical service, and thus short transfer time from the incident scene to the hospital (34-37). Nevertheless, this assumption is purely speculative, because documentation of these variables is lacking in both studies. Although in our study the survivors were discharged from the PICU with relatively good outcome (median PCPC = 2; IQR, 1–3), we did not focus on long-term outcomes in domains such as neuropsychological assessments and quality of life (QoL). The question remains whether in view of the lower percentage of WLST in the USA study, the survivors will have developed more severe long-term neurological deficits. In our study, we defined BD as clinically BD; i.e., a GCS score of 3 without brainstem reflexes for more than 24 hours after CPR, no sedation for at least 24 hours, possible effects of neuromuscular blockade administration at the time of neurological exam, and a temperature of at least 32°C. For various reasons, our study might have given an underestimation of the total number of BD patients. First, in 2016, the Dutch guideline for BD diagnosis was changed to the effect that BD was not to be determined until 12–24 hours after the CPR event, which was a new requirement (38). This implies that children admitted before 2016 may have been declared BD without fulfilling the BD criteria of the present study (wait for at least