We conducted the study according to the reporting standards of Diagnostic Accuracy (STARD) Guidelines.36 The indexing test was the independent interpretation of EEG by neurologists trained in clinical electroencephalography (`experts`). The standard of reference is the consensual interpretation of EEG by 8 independent experts who have all received at least one year of training in clinical neurophysiology. The study was conducted prospectively; methods of data collection and analysis were established prior to the index review and benchmark assessment. The Institutional Review Board of Massachusetts General Hospital in Boston approved the study and, as the study was not considered a risk to patients, waived the request for informed consent. In previous studies, expert estimates have been carefully selected, mostly short EEGs. The Scheuer et al.25 study, which included a comprehensive review of relatively long recordings, although by EEG technologists and clinical neurophysiologs not trained by scholarships, is a partial exception. However, a small number of patients were most involved in the study,41 (88%) who were assessed in an epilepsy monitoring unit. Therefore, previous studies have left the measurement of the reliability of experts in the general clinical environment open to questions of systematic and random errors. In this study, experts were selected on the basis of published criteria, taking into account a procedure guaranteeing the ability to carry out their assessments. The criteria were: the experience of judges in decision-making and decision-making; its academic and scientific reputation; their willingness and motivation to cooperate their objectivity Compliance with the requirements ; and their ability to implement the question classification techniques necessary to validate content . As a result of this procedure, experts were sought to comply with these characteristics in order to avoid the introduction of substantive distortions in the analysis of the data.
The results of this study indicate that experts are able to identify electroencephalograms that contain intercial epipilepiform discharges with significant reliability, and that differences of opinion on individual discharges of interctal epileptiforms can be widely explained by different experts who apply different thresholds to a common underlying statistical model. The reliability of IED recognition by specialists is not known. Several small studies indicate that the reliability of Interrater (IRR) may be poor for the detection of FDI in experts.16-31 These studies were based on a small number of patients, FDI and experts and focused on specially selected patient groups, so the reliability of EEG as a diagnostic test remained uncertain. Comfort and deliberate sampling were used. The participants were 14 doctors (doctoral) from the University of Granada (Spain), with an average professional experience (in research and teaching) of 14.92 years (SD: 10.37 years) and trained in teaching in the fields of pedagogical psychology, language and literary didactics, research and diagnostic methods in education, nursing, obstetric and gynecological care, and food and nutrition sciences (Table 1). In all cases, the relative number of IDUs marked by experts was strongly correlated (Figure 2D). The average pair correlation between FDI figures was 0.96 (range, 0.86-0.99).