SUNRISE CRITICAL INFRASTRUCTURE SERIES

The SUNRISE project aims to develop innovative solutions to monitor and predict the spread of infectious diseases and enhance the resilience of critical infrastructure in Europe. A cornerstone of this initiative is the development of advanced epidemiological models that incorporate a blend of deterministic and network-based approaches. This hybrid model leverages the strengths of both methodologies, enhancing the accuracy and reliability of disease forecasting. Our current focus is on analysing each NUTS1 region in Germany, integrating climate data to refine our predictions further.

Understanding the Epidemiological Model

Deterministic models are based on predefined equations that describe the dynamics of disease transmission. The most common example is the SIR (Susceptible-Infectious-Recovered) model, which divides the population into three compartments and uses differential equations to predict the number of individuals in each compartment over time. However, deterministic models often fall short in capturing the complexity of real-world interactions, such as variations in population density, mobility patterns, and localized outbreaks.

Network models, on the other hand, consider the population as a network of interconnected nodes (which can be people, households, or regions). These models are particularly useful for understanding how diseases spread through specific contacts, social interactions, and the spatial distribution of populations. In network models, regions (nodes) are connected by edges representing interactions or travel between regions. This approach is particularly useful for capturing the heterogeneity of disease spread across different areas.

The model used in the SUNRISE project combines the predictability of deterministic models with the realism of network models. By combining deterministic and network models, we can leverage the strengths of both approaches. The deterministic component provides a baseline understanding of the epidemic’s dynamics, while the network component adds the necessary detail to capture the complex interactions within and between regions. For instance, the deterministic model can predict the general trend of an epidemic, while the network model can show how an outbreak in one region might lead to subsequent outbreaks in connected regions. In this approach, each NUTS1 region in Germany is treated as a node in a network, with deterministic models applied to each node to predict local disease dynamics. The connections between nodes represent the movement of people facilitating the spread of the disease across regions.

Incorporating Climate Analysis

Climate factors significantly influence the spread of infectious diseases. Temperature, UTCI (Universal Thermal Climate Index), and precipitation can affect the survival and transmission of pathogens, as well as the behaviour of vectors and hosts. By integrating climate data into our epidemiological model, we aim to understand and predict seasonal patterns and potential climate-related changes in disease dynamics.

Current Analysis Results for NUTS1 Regions in Germany

Our current analysis focuses on the COVID-19 cases for NUTS1 regions of Germany, providing a detailed look at how this integrated modelling approach works in practice. We collected epidemiological data, including number of reported new cases, and population demographics, for each NUTS1 region. To understand the risk of infection, we fitted our model into the data. Climate data was sourced from meteorological databases, focusing on variables like temperature, UTCI, and precipitation. The integration of climate analysis into the epidemiological model revealed significant insights that some regions exhibited higher infection rates correlated with specific climate conditions, such as higher UTCI and precipitation.

General Observations:

• Temperature:

Often positively correlated with infection risk, indicating that higher temperature may be associated with higher infection rates.

• UTCI:

UTCI often show a significant relationship with infection risk, with higher UTCI generally associated with lower infection risks.

• Precipitation:

Often negatively correlated with infection risk, indicating that higher precipitation may be associated with lower infection rates.

Conclusion

The epidemiological model developed for the SUNRISE project represents a significant advancement in disease modelling, combining the strengths of deterministic and network approaches. By incorporating climate analysis, we can further refine our predictions and develop more effective intervention strategies. As we continue to analyse the NUTS1 regions in Germany, our goal is to provide actionable insights that can inform public health policy and ultimately reduce the impact of infectious diseases. By combining the strengths of different modelling approaches and factoring in the crucial role of climate, we are better equipped to predict, prepare for, and mitigate the impacts of infectious diseases in Europe.

Written by: Dr. Olga Hovardovska, Researcher, Team Clinical Epidemiology (Lead Dr. Berit Lange), Department for Epidemiology, Helmholtz Centre for Infection Research (HZI)