ESCAIDE 2024 - A summary of day 2
The second day of ESCAIDE featured a plenary on advances in surveillance and another on preparedness for an unknown pathogen.
The day began with the plenary on the surveillance of infectious diseases, which explored the latest technological advances in surveillance.
Speaker Lieke van Alphen from Maastricht University provided a brief introduction to whole genome sequencing, describing the data needed for hospital surveillance, WGS workflows, and how it can help us better understanding disease transmission and the epidemiological links between cases. Using SARS-CoV-2 an as example, Lieke van Alphen described sequencing for surveillance and outbreak investigation. The importance of data sharing was also mentioned, including rapid regional data sharing for infection prevention.
André Charlett from the UK Health Security Agency spoke about surveillance, outbreak detection, and the impact of unexpected ‘shocks' in disease transmission that can pose challenges for surveillance and require adaptation of methods, such as the changes observed in human behaviour caused by the COVID-19 lockdowns. As post COVID performance of the Flexible Farron algorithm was not optimal after the 'shock', epidemiologists started to modify the algorithm, by selecting the period of time that defines the ‘shock’, allowing typical numbers reported and temporal trends to change and, if appropriate, to incorporate a non-linear temporal trend.
AI can be a black box – we don’t really know how it works. But there are techniques to address the black box issues. – Caterina Rizzo, Professor of Hygiene and Preventive Medicine at the University of Pisa, speaking at Plenary C
Lastly, Caterina Rizzo from the University of Pizza presented a project focused on hospital-acquired infections surveillance and prevention, which used Natural Language Processing (NLP) to analyse unstructured text of hospital discharge letters for surgical site infections identification and risk-stratification. NLP is a field at the intersection of artificial intelligence and linguistics, and has evolved from early rule-based systems to contemporary to deep learning models, significantly improving performance and capabilities. Further steps of the project include obtaining structured data, HDLs and federated learning and increasing privacy.
The Eurosurveillance lunchtime seminar focused on the importance and benefits of diversity and inclusion in research and scholarly publishing. Shirin Heidari from GENDRO highlighted the potential harms of gender bias in scientific research and development of AI models. Heidari also introduced the 2016 SAGER guidelines, which were developed to promote systematic reporting on sex and gender in all research involving human, animal and related materials.
“Without inclusion, diversity is tokenism. (…) Inclusion means fostering belonging, respect, and engagement—without it, diversity cannot thrive.“ – Gowri Gopalakrishna, Assistant Professor at the University of Maastricht, speaking at the Eurosurveillance lunchtime seminar
Gowri Gopalakrishna from the University of Maastricht emphasised the critical role of diversity, equity and inclusion in research. To address gaps in diversity and inclusion, Gopalakrishna called for decolonising research agendas, fostering equitable collaborations, prioritizing capacity-building in low- and middle-income countries (LMICs), involving local communities in every research stage, and combating "helicopter research", where external researchers fail to credit or engage with the communities they study.
Today’s final plenary focused on Disease X: the unknown pathogen that could cause the next pandemic, tackling pandemic preparedness and the role of the detection in containing an outbreak.
“1650 pathogens that could turn into the next pathogen X – are we going to have a development programme for each?“ – Ana Maria Henao-Restrepo, Technical Lead for the R&D Blueprint for Emergency Response at the World Health organization, speaking at Plenary D
Ana Maria Henao-Restrepo from the World Health Organization (WHO) explored the differences between priority pathogens, prototype pathogens and then an unknown pathogen X. She also emphasised the importance of collaboration and openness in research, as they can be critical in identifying the different pathogens that could cause the next pandemic, and the best protocols to deal with them.
Mika Salminen from Finnish Institute of Health and Welfare (THL) then explained how to approach preparedness, giving the example of the recent influenza A(H5N1) fur farm outbreak in Finland and the vital role that One Health plays in preventing zoonotic outbreaks.
Christian Drosten from Charité touched on different topics to strengthen detection of outbreaks, including genomic surveillance, the need to research continuously, the importance of promoting accurate information on health and the role of science in moderating conflicting opinions.
In a poll at the end of the plenary, influenza was voted by the participants as the disease most likely to cause the next pandemic.
“Defence forces are the most useful when they do not need to be used. We should think about pandemic preparedness in the same way. We need to have everything in place also in peacetime.” – Mika Salminen, Director of the Finnish Institute for Health and Welfare (THL)
The day was wrapped up with ESCAIDE’s 10th BarCamp session, where participants came together to discuss topics that were suggested and voted on by participants. These included the role of global and public health with the incoming US administration, AI skills needed for epidemiologists, and barriers to progress in the decolonisation of epidemiology. Participants agreed that skills and judgment are still needed when using AI, and some believed that economic interest could incentivise prevention in a challenging political climate.