(264) Contemporary Causes in Over- and Under-reading of scalp EEG

This abstract has been invited to present during the Better Patient Outcomes through Diversity Platform poster session

Rationale: The routine scalp EEG is the first objective study to evaluate epilepsy. However, the reading is susceptible to human errors and subjective judgement like false positive error (FPE) and false negative error (FNE). FPE could occur not only by overreading of benign variants, electrode artifacts, but also by findings which do not have consensus yet in clinical implications like scalp permeating HFO (spHFO) and Infra-slow delta (ISD).  FNE could be caused not only by limited knowledge of neurophysiology, unawareness of recording conditions, but also by readers subjections. Quite occasionally, under-readings are regarded as better than over-reading in some community practice. As digital EEG technology improves in resolution developing in high resolution EEG and scalp dense array recordings, feedbacking knowledge of spHFO and ISD may enrich even routine scalp EEG interpretations but they may not have community consensus. Those patients who are referred for critical evaluation of EEG to the contemporary seizure specialty clinics keep asking us if they are normal or not. This presentation is seeking efforts by updating the perspectives to minimize the distance between clinical reality and error.

Methods: In the level 3 Epilepsy center, all the EEGs of epilepsy clinic were read by two board-certified epileptologists.  Seventeen EEG recordings were identified in which the impressions were different in terms of “normal” vs “abnormal.” The cases were classified into two categories: (1) Over-reading and (2) Under-reading.  In each category, the potential causes of the discrepancy were studied.

Results: Among these 17 EEG records having disagreement in conclusion, under-reading was mostly attributed to technical issue like non-utilization of available montage, inattentive to recording parameters (i.e. HFF, LFF, sensitivity change), or readers subjective judgements of benign findings. Presence of “benign EEG variants” tends to shift the diagnosis to “normal EEG,” but the combination of other unusual findings may make it as abnormal finding according to Gibb’s traditional EEG text book. Presence of interictal sharp discharges are abnormal but it may be reported as normal for an EEG reader who knows that the patient had no clinical suspicion of seizure disorder but the EEG was ordered incidentally by residents. EEG-reading was subject to reader’s clinical judgement in that situation. Ictal EEG of FLE tend to be under-reading.Over-reading was attributed to difference of interpretation of uncommon EEG findings or yet agreed in clinical implications (i.e., ISD, spHFO, lateralized spikes and waves, frontal hypersynchrony, and single lead electrographic findings). Rhythmic ISD discharges remain controversial if it is epileptic or generated by non-epileptogenic mechanism.  Spike discharges and gamma frequency evolution pattern in single lead may be representing underlying abnormality but conservative reader may judge as incidental findings with unknown clinical significance or artifacts. Peri-spike gamma activities were frequently ignored by inexperienced eyes. 

Conclusions: Between the conflict of under-and over-reading, there are collisions of novel EEG perspective and old textbook examples of artifacts and benign variants which render us risks of ambiguous interpretation. We should adopt the new EEG perspective that may minimize the distance between under- and over-reading in the modern epilepsy clinic.

Funding: Please list any funding that was received in support of this abstract.: No