Modeling the incidence rate of tuberculosis in the regions of the Russian Federation based on a family of spatial autoregression models

This paper continues the research in the field of modeling the incidence rate of tuberculosis in the regions of the Russian Federation. A distinctive feature of the work is both the expansion of the range of variables under study and the corresponding observation period, as well as the application of a model that allows modeling spatial effects, taking into account, among other things, the influence of spatial error lags (Spatial error model -SEM). Like the SAR model, it allows, to a certain extent, to solve the problem of missing variables, for example, those with the property of clustering in space, and in some cases, external shocks. The results obtained in the study allow us to correct some previous conclusions regarding the influence of certain factors, such as gender, crime rate, unemployment and real income, on the corresponding incidence rates. The conclusions drawn can be useful both in terms of further improvement of approaches to modeling incidence rates, as well as in the context of making specific management decisions aimed at reducing them, determining the possibility of assessing the effectiveness of the regional healthcare system in a specific area of the disease by comparing incidence rates in different regions after purification from the influence of various external factors (ceteris paribus).