Redefining pandemic preparedness: Multidisciplinary insights from the CERP modelling workshop in infectious diseases, workshop report

Marta C. Nunes, Edward W. Thommes, Holger Fröhlich, Antoine Flahault, Julien Arino, Marc Baguelin, Matthew Biggerstaff, Gaston Bizel-Bizellot, Rebecca K. Borchering, Giacomo Cacciapaglia, Simon Cauchemez, Alex Barbier–Chebbah, Carsten Claussen, Christine Choirat, M Cojocaru, Catherine Commaille‐Chapus, Chitin Hon, Jude Dzevela Kong, Nicolas Lambert, Katharina B. Lauer, Thorsten Lehr, Cédric Mahé, Vincent Maréchal, Adel Mebarki, Seyed M. Moghadas, René Niehus, Lulla Opatowski, Francesco Parino, Gery Pruvost, Andreas Schuppert, Rodolphe Thiébaut, Andrea Thomas-Bachli, Cécile Viboud, Jian Wu, Pascal Crépey, Laurent CoudevilleInfectious 2024 · DOI

TL;DR

This paper summarizes a workshop on infectious disease modeling, highlighting the critical need for integrating diverse data sources (including novel ones like social media and wastewater) and fostering interdisciplinary collaboration for effective pandemic preparedness. Key takeaways emphasize the advancements in AI/ML for disease surveillance and the urgent need for international data-sharing frameworks.

What Problem Does This Paper Address?

The COVID-19 pandemic exposed significant gaps in global pandemic preparedness, particularly in our ability to rapidly understand, predict, and respond to infectious disease outbreaks. While mathematical modeling played a crucial role, its effectiveness was often hampered by data limitations, a lack of standardized frameworks, and siloed expertise. This paper addresses the need to learn from the COVID-19 experience to redefine and strengthen future pandemic preparedness by bringing together diverse experts to discuss the lessons learned in infectious disease modeling.

Methods

The paper doesn't present a single novel method but rather summarizes the diverse approaches discussed at the workshop. These include:

  • Multisource Data Integration: Combining traditional surveillance data with novel sources like mobility data, social media, and wastewater monitoring.
  • Machine Learning and Artificial Intelligence (AI/ML): Utilizing AI/ML for early detection, trend monitoring, and predictive modeling of disease outbreaks. Examples include unsupervised machine learning for variant tracking and NLP for social media analysis.
  • Hybrid Modeling: Integrating traditional knowledge (e.g., physics-based models) with machine learning for enhanced prediction accuracy, especially with limited data.
  • Ensemble Modeling: Combining predictions from multiple models to improve robustness and capture diverse perspectives, particularly for forecasting.
  • Agent-Based Modeling: Simulating individual behaviors and interactions to understand disease spread in complex environments.
  • Natural Language Processing (NLP): Analyzing text data from social media and surveys to extract epidemiological insights.

Datasets

The paper discusses the use of various datasets across different presentations:

  • Traditional Surveillance Data: Case counts, hospitalization data, death certificates, testing data.
  • Mobility Data: Data from telecommunication companies and other sources indicating population movement.
  • Social Media Data: Tweets, online discussions, and public opinion data.
  • Wastewater Monitoring Data: Viral concentrations in wastewater.
  • Pharmacy Consumption Data: Data on medication sales.
  • Genomic Data: Viral sequences for variant tracking.
  • Hospital Data Warehouses: Electronic Health Records (EHRs).
  • Publicly Available Datasets: Google Trends, GISAID, and national health databases.

No new benchmark datasets were explicitly introduced, but the paper highlights the challenges in accessing and standardizing existing data.

Key Insights and Ideas

  • Data is King, but Diversity is Queen: The COVID-19 pandemic underscored that no single data source is sufficient. Integrating diverse, often unconventional, data streams is crucial for a comprehensive understanding of disease dynamics.
  • AI/ML is a Game Changer, but Not a Silver Bullet: AI and ML have shown immense potential in early detection and prediction, but their effectiveness relies heavily on the quality and integration of data.
  • Collaboration is Non-Negotiable: Siloed expertise is a major impediment. The workshop emphasized the need for interdisciplinary collaboration between epidemiologists, data scientists, public health experts, and policymakers.
  • The "Human Element" Matters: Behavioral data and understanding public perception are critical for accurate modeling and effective intervention strategies.
  • International Cooperation is Lacking: A significant gap exists in a common international framework for data sharing, hindering global pandemic preparedness.
  • Transparency and Communication are Vital: Modellers must communicate their findings clearly, including uncertainties, to avoid sensationalization and politicization.

Key Results

The paper focuses on summarizing discussions and insights rather than presenting specific quantitative results from a single study. However, key qualitative findings and demonstrated capabilities include:

  • Early Warning Systems: Several presentations showcased the development and application of early warning systems using AI/ML and diverse data sources, demonstrating the ability to detect trends and potential outbreaks days to weeks in advance.
  • Predictive Modeling Advancements: Significant progress has been made in developing predictive models for disease spread, hospitalizations, and outcomes, with examples from various countries.
  • Wastewater Epidemiology Success: Wastewater monitoring has proven to be a valuable early indicator of disease prevalence, often preceding clinical testing.
  • Social Media Insights: NLP techniques can extract valuable information about self-reported symptoms and public sentiment from social media, aiding in early detection and understanding.
  • Ensemble Model Performance: Ensemble models generally showed more consistent performance than individual models in forecasting, highlighting the benefit of diverse perspectives.

Limitations and Future Work

  • Data Quality and Accessibility: A persistent challenge is the lack of standardized, high-quality, and easily accessible data, especially across different countries and for real-time analysis.
  • Model Transparency and Interpretability: Ensuring complex models are understandable to policymakers and the public remains a challenge.
  • Behavioral Dynamics: Accurately incorporating human behavior, which is influenced by societal norms and misinformation, into models is difficult but crucial.
  • Global Collaboration: Fostering stronger global collaboration and knowledge exchange, particularly through public-private partnerships, is essential.
  • Projection Horizon: The effective projection horizon for infectious disease models is often limited, especially with the emergence of new variants.
  • Data Sharing Frameworks: The absence of a common international framework for data sharing is a major impediment to global preparedness.

Future work should focus on developing more robust and adaptable modeling frameworks, enhancing data integration strategies, strengthening interdisciplinary collaboration, and establishing international data-sharing mechanisms. The paper also suggests exploring agent-based modeling and additional data sources for future developments.

Summary updated

Abstract

In July 2023, the Center of Excellence in Respiratory Pathogens organized a two-day workshop on infectious diseases modelling and the lessons learnt from the Covid-19 pandemic. This report summarizes the rich discussions that occurred during the workshop. The workshop participants discussed multisource data integration and highlighted the benefits of combining traditional surveillance with more novel data sources like mobility data, social media, and wastewater monitoring. Significant advancements were noted in the development of predictive models, with examples from various countries showcasing the use of machine learning and artificial intelligence in detecting and monitoring disease trends. The role of open collaboration between various stakeholders in modelling was stressed, advocating for the continuation of such partnerships beyond the pandemic. A major gap identified was the absence of a common international framework for data sharing, which is crucial for global pandemic preparedness. Overall, the workshop underscored the need for robust, adaptable modelling frameworks and the integration of different data sources and collaboration across sectors, as key elements in enhancing future pandemic response and preparedness.
BibTeX
@inproceedings{nunes2024redefining,
  author = {Marta C. Nunes and Edward W. Thommes and Holger Fröhlich and Antoine Flahault and Julien Arino and Marc Baguelin and Matthew Biggerstaff and Gaston Bizel-Bizellot and Rebecca K. Borchering and Giacomo Cacciapaglia and Simon Cauchemez and Alex Barbier–Chebbah and Carsten Claussen and Christine Choirat and M Cojocaru and Catherine Commaille‐Chapus and Chitin Hon and Jude Dzevela Kong and Nicolas Lambert and Katharina B. Lauer and Thorsten Lehr and Cédric Mahé and Vincent Maréchal and Adel Mebarki and Seyed M. Moghadas and René Niehus and Lulla Opatowski and Francesco Parino and Gery Pruvost and Andreas Schuppert and Rodolphe Thiébaut and Andrea Thomas-Bachli and Cécile Viboud and Jian Wu and Pascal Crépey and Laurent Coudeville},
  title = {{Redefining pandemic preparedness: Multidisciplinary insights from the CERP modelling workshop in infectious diseases, workshop report}},
  booktitle = {Infectious},
  year = {2024},
  doi = {10.1016/j.idm.2024.02.008},
}
Figure