Abstracts

Rationale:
Electroencephalographic (EEG) monitoring is increasingly used to identify and manage electroencephalographic seizures, assess encephalopathy severity, and identify acute changes in critically ill neonates and children. Traditionally, EEG reports have been generated using free text. However, despite a standardized language and reporting guidelines from professional organizations, EEG reports are notoriously incomplete. In particular, up to 50% of free text EEG reports for critically ill children and neonates are known to lack documentation of key elements including presence of seizures and key background features. As these reports form the basis for clinical decision-making, initiatives to improve standardized EEG reporting represent a quality improvement goal. We aimed to determine whether a standardized EEG reporting system that was previously shown to have good acceptance by users improved completeness of EEG documentation.
Method:
In April 2017, we implemented an EEG reporting system using Epic SmartForms, which incorporated standardized terminology recommended by the American Clinical Neurophysiology Society (ACNS). This SmartForm has been used for all clinical EEG documentation since its inception. We performed a quality improvement study to determine the proportion of patients for which key EEG variables were documented using this novel documentation system. We defined key variables as continuity, voltage, presence or absence of seizures, and presence of push-button events. We focused on continuous video-EEG reports for critically ill children and neonates given prior studies assessing the completeness of EEG documentation in these cohorts.
Results:
Between April 2017 and October 2019, the novel documentation system was used to report 15,030 EEGs with 1,504 patients in the ICU setting, comprised of 564 EEGs in neonates and 940 in children. Overall, 93.8% of patients (1411) had EEGs describing continuity, 97.4% (1465) described voltage, 99.4% (1496) described presence or absence of seizures, and 99.2% (1492) described the presence of push button events. In prior studies of free text reports, none of these features were consistently documented in more than 70% of reports.
Conclusion:
Implementation of a standardized EEG reporting system directing the documentation of discrete key EEG variables yields high completeness of documentation. We find consistent implementation of recommended terminology in >93% of all EEG reports in critically ill neonates and children. This improved documentation meets the recommended requirements for standardized EEG documentation, which we believe may improve clinical care and lays the foundation for quality improvement studies to assess the impact of EEG guided management on patient-centered outcome measures.
Funding:
:None