Design and implementation of an evaluation framework for the Epidemic Intelligence from Open Sources (EIOS) system for international public health intelligence at the Robert Koch Institute, Germany, 2023

BACKGROUNDPublic Health Intelligence (PHI) aims to detect health threats early for a timely and effective response. The PHI team at the Robert Koch Institute (RKI) uses the Epidemic Intelligence from Open Sources (EIOS) system in combination with other sources for detecting signals of international public health threats relevant to Germany. However, while EIOS is increasingly used for PHI worldwide, it is rarely evaluated.AIMWe designed and conducted an attribute-based evaluation to assess EIOS's performance for international PHI in 2023 and to identify areas for improvement.METHODSWe adapted surveillance system attributes and designed attribute-specific data collection methods. We conducted a mixed-method evaluation combining prospective and retrospective operational data collection with feedback from PHI officers.RESULTSDuring 2 weeks in July 2023, the PHI team reported 20 signals: 16 detected using EIOS and four from other sources. Increasing the number of EIOS sources increased timeliness and sensitivity slightly but caused a 35-fold increase in articles to screen (35,546 vs 1,138). The team found EIOS flexible and simple for signal detection but identified challenges in simplicity of signal documenting and reporting and in completeness of EIOS sources screened by the team.CONCLUSIONThe current use of EIOS proved sensitive and timely. However, PHI must balance sensitivity, timeliness and resource requirements. To maintain this balance, we strongly recommend regular evaluations of the use of EIOS for PHI. Our evaluation offers practical guidance for other PHI teams. We recommend integrating EIOS with an event management system to facilitate signal documentation and reporting.