Geographical monitoring and prediction of infectious diseases: Developing a geostatistical analysis approach for assessing risk and clustering of infectious diseases at district level in Germany

**Introduction:** Routine geographic monitoring of notifiable infectious diseases in Germany predominantly relies on descriptive methods and basic epidemiological measures. While sufficient for a situational overview, current approaches often lack rigorous modelling techniques. Consequently, important latent spatial and temporal factors, as well as societal disparities are often underrepresented in routine assessments. Bayesian statistics have gained increasing attention in infectious disease modelling due to their capacity to incorporate prior knowledge, account for uncertainties, and model complex spatiotemporal dynamics. **Objectives:** This cumulative doctoral thesis aims to develop a framework for an analytical approach to geographic monitoring and prediction of infectious diseases at the district level in Germany, integrating latent spatiotemporal factors, key contextual determinants of health, and strategies for prevention and control. **Methods:** This doctoral thesis proposes a structured framework for a geostatistical analysis approach that bridges concepts from spatial and social epidemiology with empirical evidence from the doctoral candidate’s published work. The framework is grounded in Gaussian Markov Random Field theory, and concepts on spatial inequalities in social determinants of health. Drawing on five published articles using geographic surveillance data for measles and Covid-19 as example diseases, this thesis provides a conceptually and statistically detailed modelling approach to routinely i) assess the complex interplay between latent spatiotemporal effects and contextual population-level attributes, and ii) monitor potential societal drivers, and predict spatial patterns and clustering of infectious disease risk. **Results:** The proposed framework includes i) an epidemiological study design for geographic monitoring, ii) epidemiological measures for descriptive and analytical purposes, iii) appropriate likelihood functions for count data, iv) modelling techniques for structured and unstructured spatial and temporal random effects, v) incorporation of contextual covariates as fixed effects, and vi) a comprehensive flowchart outlining the complete analytical procedure. **Conclusion:** This empirically, conceptually, and statistically grounded framework offers practical, methodological guidance for researchers and relevant public health actors seeking to develop or enhance routine geographic monitoring systems for infectious diseases in Germany.