System approach to data analysis of pandemic development and forecasting

Abstract

The Covid-19 pandemic is one of the greatest challenges to humanity of last decades. Effective prevention of its development is the problem of decision-making with many criteria, high uncertainty of the results of a choice, insufficiently understood feedbacks. There has been collected statistics on the main indicators of the pandemic over the past year. However, according to most researchers, the initial data on the number of infected and fatal cases are significantly underestimated. This makes inaccurate other important indicators, in particular, those that characterize the nature of the dynamics, the rate of infection and its mortality. At the same time, a large number of research results provide additional sources of information to improve the quality of pandemic analysis and forecasts. Mathematical models infections spreading have been significantly developed, which make it possible to refine individual indicators.
According to model estimates, the maximum number of daily cases can be 50 times higher than the official data, and the dates of maxima can be up to 4-5 weeks earlier. For Ukraine, the highest values of "excess mortality" (as a percentage of the average level in 2015 - 2016) were observed in September (15.8%), October (20.3%), November (33.7%) and December (29.2%). In January 2021, according to incomplete preliminary data it is expected the significant reducing of the indicator to the value less than 2%. Per 100 thousand inhabitants, the "excess mortality" in 2020 was about 38.8 thousands (in the first 5 months there were negative values of the indicator), and for the period from June 2020 to January 2021 - about 53.5 thousands, or 14.1% of the base level for this period. An analysis of Google's search queries gives grounds to assume that in early January 2020, the pandemic spread in many countries on different continents, and precautionary measures were taken too late.
The obtained data show that the involvement of additional sources of information makes it possible to compensate for the imperfections of official operational data and to un-derstand better the patterns of occurrence and development of Covid-19 pandemic. At the same time, a system approach to decision-making on prevention the development of a pandemic should take into account the information on morbidity and mortality statistics as well as other information, in particular on virus mutations, re-infection, vaccination, testing, socio-economic consequences etc. This should be the subject of further research.