A pandemia da COVID-19 no Brasil: uma aplicação do método de clusterização k-means

Abstract

A COVID-19 é uma infecção causada pelo coronavírus SARS-CoV-2, sendo que seus primeiros registros foram na cidade chinesa de Wuhan em dezembro de 2019, e foi considerada pela Organização Mundial da Saúde (OMS) uma pandemia mundial em março de 2020. No Brasil, a COVID-19 se espalhou atingindo as 27 unidades federativas (UFs). Com isso, as tomadas de decisões para diminuir a velocidade de transmissão foram baseadas nas recomendações da OMS, onde a principal é isolamento social. Entretanto, devido a heterogeneidade da população em cada uma das UFs, a pandemia se difundiu de forma distinta. Deste modo, é interessante fazer o agrupamento das UFs por similaridade devido algumas características, e assim, observar as medidas de combate a COVID-19 realizadas em cada um desse grupos. O objetivo deste estudo foi agrupar as UFs usando análise de cluster pelo método não-hierárquico k-means considerando os coeficientes epidemiológicos como incidência, prevalência e letalidade. Os dados foram obtidos do site do Ministério da Saúde do Brasil e foi constituído pelas variáveis número de casos e óbitos novos e acumulados nas UFs, além da população em risco. Para análise de cluster a base de dados foi dividida em três períodos cronológicos para os três coeficientes em estudo. Com a análise de cluster foi possível verificar a estratificação da UFs conforme suas similaridades em relação a COVID-19. Assim, a estratificação da incidência, prevalência e letalidade por UFs pode se apresentar como um recurso adicional para sinalizar quais locais e quais medidas deverão ser adotadas e onde essas medidas foram eficazes.

Use of grouping techniques applied to the numbers of infected by Covid-19 in the Brazilian States comparing with the HDI of each State

“…The non-hierarchical methods in general, the focus is to find the number "k" of clusters that can perform the division of observations in a satisfactory way, the method that can identify similarities and differences between the observations (Duarte, 2021). As it is a more dynamic and interactive grouping process, the non-hierarchical method, the number of groups is specified before the grouping process, the criterion, but used by this method is that of K-means, according to Alves (2020) this method has some conditions such as the previous information of the numbers of clusters k, where their observations are grouped in these k clusters using a function with objective and criterion. Being of simple application and fast processing according to Duarte (2021), the algorithm logic follows 4 steps:…”

“…In December 2019, the Chinese government announced in Wuhan (Hubei, China) a discovery of a new coronavirus, named SARS- , With this announcement, the World Health Organization (WHO) was alerted, declaring that the infection caused by this virus contaminated humans with high transmission potential (Alves, 2020).…”

“…WHO recommendations to slow down transmission, its main decision was social isolation, Covid-19 has spread throughout countries, in Brazil it reached 27 federative units, this fact occurred due to the challenges regarding the conditions of social vulnerability, housing and precarious sanitation, in addition to household overpopulation, however, due to the heterogeneity of the population in each of the states, the pandemic spread differently (Alves, 2020).…”

Procurement management of strategic inputs in coping with COVID-19: scoping review protocol

“…Due to the spread of the new Coronavirus in several countries, the World Health Organization (WHO) declared a public health emergency of international importance on January 30 th , 2020; and, subsequently, on March 11 th , 2020, declared a pandemic situation in which refers to infection by the etiologic agent SARS-CoV-2 (Alves et al, 2020). In Brazil, the first case of COVID-19 was identified on February 26 th , 2020, in São Paulo and, less than a month after its confirmation, all Brazilian states had already reported cases of the disease, with deaths recorded in eight of them (Silva et al, 2020).…”

Local Moran Index: an application in epidemiological coefficients of the COVID-19 pandemic in Brazil

“…In Brazil, this pandemic spread rapidly, and this occurred due to the challenges regarding the conditions of social vulnerability, housing, poor sanitation, and overcrowding in the home (Alves et al, 2020;Fernandes et al, 2020;Werneck & Carvalho, 2020). These factors make it difficult for millions of Brazilians to follow the WHO recommendations for partial and even total social isolation in some situations and hygiene with alcohol gel and respiratory masks, and these measures were effective in other countries (Werneck & Carvalho, 2020).…”