Maayke Hunfeld

76 Chapter 3 Methods We conducted a cross-sectional anonymous electronic survey (using LimeSurvey, supplemental data 1) across European Society of Paediatric and Neonatal Intensive Care (ESPNIC) members and European Paediatric Neurology Society (EPNS) members. The survey was carefully designed by the authors in order to address all aspects of the research question and hypothesis (34). The questions in our survey were clear and straightforward, with an appropriate length of the questionnaire and check questions. The survey was reviewed by 4 ESPNIC section chairs (Neuro critical Care, Ethics, Outcome, Resuscitation) and an expert board member of EPNS. Subsequently, the survey was piloted on a paediatric neurologist, paediatric intensivist and fellow intensivist for clarity and face validity. The questionnaire was written in English; it consisted of 49 questions divided into 4 sections: general information, cardiac arrest, neuro-prognostication and follow-up (see appendix A) requiring 15 minutes on average to complete. The specific points of interest were practices concerning post-resuscitation care, neuro-prognostication and end-of-life decision making. The survey has been approved by the ErasmusMCMedical Review Ethics Committee (MEC 2019-0095). Paediatric (fellow) intensivists, paediatric neurologist and other healthcare professionals being a member of ESPNIC or EPNS received a link (by email or newsletter) to the online survey. The survey started in January 2019. After 3 weeks, a reminder to fill out the survey was sent. All valid responses received before the 8 th of March were included for analysis. For inclusion, a completed neuro-prognostication section was mandatory. Also surveys completed by physicians active in non-European PICUs were excluded (According to the geographic scheme of the United Nations). Descriptive statistics were used to analyse all answers. Most questions were analysed on individual level. However, questions concerning PICU characteristics were analysed on PICU-level. In case contradictory answers were given within one PICU, it was classified as intra PICU variability. Results The survey was sent to 1310 EPNS members representing 47 European countries and 611 ESPNIC members, representing 31 European and 28 non-European countries (exact number of PICUs unknown). Of the 108 respondents, 71 (66%) completed the “neuro-prognostication” section (Figure 1).