Инновационная медицина Кубани (Mar 2023)

Cluster Analysis Results for Assessment of COVID-19-Related Mortality Differences Between Russian Regions

  • V. T. Korkhmazov,
  • V. I. Perkhov

DOI
https://doi.org/10.35401/2541-9897-2023-26-1-65-71
Journal volume & issue
Vol. 0, no. 1
pp. 65 – 71

Abstract

Read online

Background: Russian state policy for health protection, rehabilitation, and health improvement requires studying regional mortality rates, including those related to COVID-19.Objective: To assess differences in COVID-19-related mortality between the regions of the same federal district.Materials and methods: Mortality data are sourced from death records in the Unified State Register of Civil Status Acts. The data were analyzed using unweighted arithmetic means, specific indicators, and standard deviation (the mean ± standard deviation). To eliminate the distortion by an age factor in mortality rates comparison, we replaced the age structure of the individual region’s population with that of Russia through indirect standardization of mortality rates. We used K-means clustering to group the regions by COVID-19-related mortality rates.Results: In 2021 Russia had 2,446,922 deaths, i.e. 648,615 (36.1%) and 163,645 (7.2%) cases more compared to 2019 and 2020, respectively. Of the total number, 17.3% of cases (424,252) had COVID-19 as a primary cause of death: nearly three times more than in 2020 (144,691 COVID-19-related deaths). Based on the average sizes of all individual age groups related to the respective region, nonstandardized and standardized COVID-19-related mortality rates were 265.30 ± 103.16 and 279.28 ± 91.07 per 100,000 persons in 2021, respectively. The cluster analysis showed that the largest number of regions (28 regions in 8 federal districts) comprised the third cluster with an average mortality rate of 276.26 ± 15.16 per 100,000 persons. The first cluster with an average mortality rate of 406.43 ± 29.26 per 100,000 persons included 12 regions in 7 federal districts. The second сluster included 21 regions (341.49 ± 18.16 per 100,000 persons) in 6 federal districts, the fourth cluster – 17 regions (196.73 ± 25.05 per 100,000 persons) in 6 districts, and the fifth cluster – 7 regions (87.22 ± 12.42 per 100,000 persons) in 5 districts.Conclusions: There is no common explanation for the COVID-19-related mortality differences not only between the regions of the same country but also between countries. This lack of understanding gets worse because one should also separate the pandemic’s health factors from social, psychological, and economic ones. The government should play a more important role in healthcare management, reform payment systems, and eliminate private financial intermediaries used to pay for medical services.Restrictions: The Unified State Register of Civil Status Acts data, which consisted of preliminary death certificates, may differ from the data of the Federal State Statistics Service, which became available to researchers later in 2022.

Keywords