Spatial heterogeneity affects predictions from early-curve fitting of pandemic outbreaks: a case study using population data from Denmark

The modelling of pandemics has become a critical aspect inmodern society. Even though artificial intelligence can helpthe forecast, the implementation of ordinary differentialequations which estimate the time development in the numberof susceptible, (exposed), infected and recovered (SIR/SEIR) individuals is still important in order to understand thestage of the pandemic. These models are based on simplifiedassumptions which constitute approximations, but to whatextent this are erroneous is not understood since many factorscan affect the development. In this paper, we introducean agent-based model including spatial clustering andheterogeneities in connectivity and infection strength. Based onDanish population data, we estimate how this impacts the earlyprediction of a pandemic and compare this to the long-termdevelopment. Our results show that early phase SEIR modelpredictions overestimate the peak number of infected and theequilibrium level by at least a factor of two. These results arerobust to variations of parameters influencing connectiondistances and independent of the distribution of infection rates.