WP leads: SSI (DK) and RKI (DE)
The SARS-CoV-2 pandemic has highlighted that timely sharing of laboratory test results is crucial for an informed response during an unfolding epidemic. Many countries experienced challenges and delays in sharing of test results in a timely fashion, for instance due to lack of capacity, fragmented systems, or lack of digitalization. Also, many countries do not yet use any methods for automatic outbreak detection.
Infectious disease outbreaks can have a considerable impact on the health and health systems of countries. Timely and accurate data sharing and detection of such events will help to contain further spread of disease and reduce harmful consequences. Automated tools can help manage and foster analysis across different datasets. Especially in context of limited human resources or high workload, this can contribute to a comprehensive surveillance and timely implementation of measures. Many countries do not use any methods for automatic outbreak detection yet or use methods that have been designed for other surveillance systems and datasets than their own. Also, there are no extensive evaluations of outbreak detection methods for a broad range of surveillance data.
Therefore, the objective of WP2 Outbreak detection is to support outbreak detection and pandemic preparedness by improving real-time surveillance for a coordinated response. By improving national surveillance systems, the goal is to strengthen overall surveillance in Europe.
This WP is structured in two main technical tasks:
1) Improving Laboratory Based-Reporting
2) Outbreak & Signal Detection.
Both tasks are subdivided in subtasks that include country-specific pilots that are aligned with national priorities. Lessons learnt from the pilot projects are shared with other participating countries in form of reports, training material and during site visits to piloting countries.
Improving laboratory-based reporting
Timely and reliable laboratory data is vital for infectious disease surveillance and an informed response. The importance of this has become particularly apparent during the unfolding of the SARS-CoV-2 pandemic, where many countries encountered challenges, such as insufficient timeliness in sharing laboratory test results, for instance due to lack of capacity, fragmented systems, or lack of digitalization. A specific challenge centered on the sharing of genetic data, due to its complexity. Task 1 on Improving Laboratory-Based Reporting therefore aims to make an inventory of current limitations and improving national laboratory surveillance systems through country-specific pilots.
Task 1 started off with a survey on mapping needs and gaps with regards to laboratory-based surveillance among countries of the UNITED4Surveillance consortium. The survey focused on general, legal, technical (data & IT), policy and organizational, as well as financial aspects and identified concrete gaps. Results were presented during a workshop and were summarized in a deliverable report. During the workshop, selected countries also further introduced their respective laboratory-based surveillance systems, followed by discussions on joint challenges. Another subtask of task 1 focused on data standards and exploring a logical data model for genotyping/subtyping data. This work started by a workshop during which selected countries presented the data model of their respective laboratory surveillance database, followed by a discussion on what an ideal data model should look like. Conclusions from this work were summarized in a milestone report.
Furthermore, four countries – Denmark, Finland, the Netherlands and Norway – are in the process of conducting pilots on either setting up a new open-source sequence database or upgrading existing laboratory surveillance systems, respectively.
Denmark’s pilot objective is to enhance reporting of microbial properties nationally through the implementation of a new data standard. Finland’s focus is on describing and integrating a data model for Shiga Toxin-producing E. coli (STEC) into the Finnish infectious diseases surveillance system. The Netherlands pilot an open-source reference database, as part of a newly developed infectious disease surveillance platform. Finally, Norway’s pilot focus in on developing a data management protocol for diagnostic test data on STEC reported to MSIS-laboratory database.
In 2024 all four piloting countries hosted a site visit to present the national surveillance of communicable diseases, the national laboratory infrastructure, and to share progress, experience and lessons learnt from the pilots.
Outbreak & Signal Detection
Task 2 on Outbreak & Signal Detection focuses on providing a tool that enables countries to use automated outbreak detection algorithms and improving algorithms for outbreak detection based on routine surveillance data for infectious diseases. Through early detection of infectious disease outbreaks, the further spread of diseases can potentially be contained and harmful consequences can be reduced. Automated signal detection tools can thereby help to identify specific events of concern and foster automated data analysis. The timelier and more reliable the signals are, the greater their contribution to comprehensive surveillance and effective implementation of outbreak management measures.
The work under task 2 started off with a survey and workshop aiming to get an overview of the types of surveillance systems with and without implemented outbreak detection methods of participating countries. Further, the countries’ functional and technical requirements for the tool were collected and different outbreak detection methods were systematically evaluated using real world surveillance data from participating countries. Based on these insights and needs, a Signal Detection Tool was jointly developed by data scientists from Austria, Germany, Finland and Denmark. The Signal Detection Tool is an R shiny app which can be used locally and is applied to surveillance data. The app uses provides a range of input parameters, including controls for filtering, stratification, algorithm selection and a signal detection period. Based on the selected settings, the tool visualizes signals and case numbers through maps, bar charts and time series plots. Additionally, the generated signals are summarized in a table, and a word or html report containing the results can be produced.
The Signal Detection Tool was piloted by ten countries from April to November 2024. In order to assess the tools’ ability to identify outbreaks in a timely and precise manner and its user-friendliness, users were invited take part in a monthly evaluation survey. The questionnaire was designed considering attributes used for evaluation of public health surveillance systems and were adapted accordingly. Overall, data on signals and whether these lead to further investigations, selected input parameters, ease of use, technical issues and desired features were collected. The continuous evaluation facilitated iterative and continuous improvements of the tool. In response to pilot feedback, four updated versions incorporating new features and bug fixes were released during the pilot phase.
The tool is publicly available on Github, including installation instructions and documentation on the usage of the tool. We also share course materials that we have compiled for a half-day workshop as part of EPIETs time-series analysis (TSA) module in December of 2024. The module itself is a yearly 5-day training course that introduces EPIET fellows of the current cohort (about 50 fellows) to different concepts of time-series analysis in the context of infectious disease epidemiology. Our half-day workshop comprised a 1-hour lecture on infectious disease outbreak detection and a 2-hour practice session. The practice session is divided into three parts and several tasks, leading the fellows from implementing very simple outbreak detection functions in R from scratch, to using functions from established R-libraries for the analysis of surveillance data such as the surveillance package, to finally using functions and the R-Shiny app within our Signal Detection Tool. We received positive feedback from the participants and hope to be able to repeat the course within the TSA module in the coming years. We are also going to use similar course material and agenda for an Interactive Learning Session (ILS) during the 12th TEPHINET conference (June 2 – 5, 2025).
UNITED4Surveillance is funded by the European Union under the EU4Health Programme (EU4H), project ID 101102070.
Views and opinions expressed are however those of the author(s) only and do not necessarily reflect those of the European Union or granting authority. Neither the European Union nor the granting authority can be held responsible for them.
Project Office UNITED4Surveillance
National Institute for Public Health and the Environment
P.O. Box 1 | 3720 BA Bilthoven | The Netherlands
E-Mail: JAsurveillance@rivm.nl
© 2023 • All Rights Reserved