Identification of spatial variations in COVID-19 epidemiological data using K-Means clustering algorithm: a global perspective

Abstract

ABSTRACTBackgroundDiscerning spatial variations of COVID-19 through quantitative analysis operating on the geographically designated datasets relating to socio-demographics and epidemiological data facilitate strategy planning in curtailing the transmission of the disease and focus on articulation of necessary interventions in an informed manner.MethodsK-means clustering was employed on the available country-specific COVID-19 epidemiological data and the influential background characteristics. Country-specific case fatality rates and the average number of people tested positive for COVID-19 per every 10,000 population in each country were derived from the WHO COVID-19 situation report 107, and were used for clustering along with the background characteristics of proportion of country’s population aged >65 years and percentage GDP spent as public health expenditure.ResultsThe algorithm grouped the 89 countries into cluster ‘1’ and Cluster ‘2’ of sizes 54 and 35, respectively. It is apparent that Americas, European countries, and Australia formed a major part of cluster ‘2’ with high COVID-19 case fatality rate, higher proportion of country’s population tested COVID-19 positive, higher percentage of GDP spent as public health expenditure, and greater percentage of population being more than 65 years of age.ConclusionIn spite of the positive correlation between high public health expenditure (%GDP) and COVID-19 incidence, case fatality rate, the immediate task ahead of most of the low and middle income countries is to strengthen their public health systems realizing that the correlation found in this study could be spurious in light of the underreported number of cases and poor death registration.