Title: Causal models and phylo-spatio-temporal, multidimensional epidemiology: Dark figure estimation Authors:  Andrzej Jarynowski - FU Berlin (Germany) [presenting]
Vitaly Belik - FU Berlin (Germany)
Abstract: Triangulation of data and methods is the way to get into findings (tools previously developed by statisticians and econometrists) to offer valuable insights for ONE public health. A multidimensional analysis was conducted to understand disease transmission in Poland, integrating key branches. A) A/H5N1 Epizootic in Cats: i) In-depth analysis of the A/H5N1 Epizootic in cats, involving positive and negative cases, RNA sequences, and participatory epidemiology data. ii) Utilization of daily time series data, revealing phylodynamic spatiotemporal clusters, patterns of disease spread, and hotspots of transmission. iii) Identification of separate virus introductions in eastern and western Poland, with A/H5N1 likely circulating in cats a month before the first confirmed case. B) Impact of healthcare access on COVID-19 Burden: i) Causal modelling to analyze the relationship between healthcare access and COVID-19 incidence. ii) BIC to select the optimal structure of the model (paths), nonparametric bootstrap to assess the strength of links in the model. iii) Highlighting the significant role of healthcare access in shaping geographical variations in COVID-19 burden. iv) Providing a nuanced understanding of healthcare access' contribution to unreported or undiagnosed COVID-19 infections and the effect of vaccines on preventing deaths.