Sources of interrater variability and prognostic value of standardized EEG features in post-anoxic coma after resuscitated cardiac arrest

Abstract

Highlights We identified a new approach to improve the prognostic value of EEG patterns. Interrater agreement was evaluated and reported for each different EEG pattern. Causes for discrepancy were elucidated to improve interrater concordance.

Prediction of poor neurological outcome in comatose survivors of cardiac arrest: a systematic review

“…Sensitivity and FPR for index tests based on clinical examination, biomarkers, electrophysiology, and imaging are reported in Tables 3,4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31,32,33,34,35,36,37,38,39,40,41,42,43,44,45,46,47,48,49,50,51,52,53,…”

“…22,23,24,25,26,27,28,29,30,31,32,33,34,35,36) In one study [69], an isoelectric EEG, defined as all activity below 2 µV, predicted poor outcome at 12 and 72 h from ROSC with 0% FPR and high precision (95% CIs 0-4%) ( Table 22). In six studies, a suppressed EEG background defined according to the ACNS terminology (i.e., all activity of the record < 10 µV; see ESM Table E4) almost invariably predicted a poor neurological outcome [52,53,61,68,69,72], especially after 16-24 h from ROSC. In studies using continuous EEG recording, sensitivity of suppressed EEG background progressively decreased over the first 48 h after ROSC [61,68], and between ≤ 24 h and 36-72 h after ROSC (Tables 23, 24).…”

“…A low-voltage EEG, defined according to ACNS [116] as all activity below 20 µV, was assessed in four studies and predicted poor outcome with 0-9% FPR from 6 to 72 h from ROSC in one study [68]. However, in another study [53], FPR for this sign was 78%. In two other studies [52,72], a low-voltage EEG at a median of 76-77 h from ROSC predicted poor outcome with 11-12% FPR (Table 26).…”

“…Four studies [53,61,69,110] assessed the accuracy of sporadic, non-rhythmic/periodic discharges on EEG from 0-8 up to 72 h from ROSC. Their specificity was lower than that of rhythmic/periodic discharges (Table 38).…”

“…In four studies [53,61,68,72], the presence of seizures defined according to ACNS (unequivocal seizures,see ESM Table E4) from 0-8 h to 96 h after ROSC was invariably associated with poor neurological outcome, except in one [68], where FPR was 2.6 [0.1-13.5]% at 96 h from ROSC (Table 39). In two other studies, not adopting ACNS definitions [51,109], the presence of seizures predicted poor outcome with 0% FPR.…”