Measuring the effects of pre-test probability on out-patient first EEG investigation in children – A guide to evidence-based EEG triage in a pandemic

• Published in: Seizure (London, England) Vol. 86; pp. 8 - 15
• Main Authors: McHugh, John C., Daly, Nicole, Colfer, Ailish
• Format: Journal Article
• Language: English
• Published: England Elsevier Ltd 01.03.2021
• Subjects: Electroencephalography; Epilepsy; Epileptiform abnormality; Paediatric; Triage; Paediatric; Electroencephalography; Triage; Epileptiform abnormality; Epilepsy
• ISSN: 1059-1311; 1532-2688; 1532-2688
• DOI: 10.1016/j.seizure.2021.01.007

•Algorithmic Triage of EEG referrals predicts epileptiform yield in children.•Epileptiform abnormality is 2–3 times more likely in high priority referrals.•Our methods permit evidence-based prioritization of EEG studies.•Mismatch between pre-test probability and EEG outcome merits special attention. The yield of epileptiform EEG abnormalities is lower in unselected Paediatric populations than in prospective studies of incident seizures or prevalent epilepsy studies. At a time of limited capacity, it is important to match available EEG resources to children who are most likely to benefit. In this study we evaluated a prospective triage tool for estimating the likelihood of epileptiform abnormality in children’s first out-patient EEG. We prospectively triaged 1865 out-patient referrals to the largest Paediatric EEG laboratory in Ireland. Based on a structured algorithm, we dichotomized first EEG referrals into priority and non-priority groups and assigned one of 5 sub-levels based on anticipated EEG yield. EEGs were reported by a single Consultant in Clinical Neurophysiology. Triage designated 757 (41 %) EEG referrals as non-priority. Priority exceeded non-priority referrals for all age groups except children between 18 months and 3.5 years. EEGs showed a two-fold higher incidence of interictal epileptiform abnormalities for priority referrals (36 % vs 18 %, p < 0.001). Rates of interictal epileptiform abnormality correlated with the 5 sub-levels of triage (p < 0.01). Epileptiform yield was highest (39 %) for children over 5 years vs 17 % for those under 5 years (p < 0.00001); these rates increased to 49 % and 20 % respectively for priority referrals. Structured pre-test triage of EEG referrals can identify children who have the greatest likelihood of epileptiform abnormality. In a mixed population of Paediatric referrals, the epileptiform yield of first time EEG is 49 % for children over 5 years who are referred with an appropriate EEG indication. This is subject to much variability with epileptiform yields as low as 13 % in younger children with non-priority referrals. The use of a structured triage algorithm can help to optimise utility of EEG in situations of limited laboratory capacity.