Article

Understanding spatiotemporal patterns of COVID-19 incidence in Portugal: A functional data analysis from August 2020 to March 2022

Understanding spatiotemporal patterns of COVID-19 incidence in Portugal

PLoS ONE

Manuel Ribeiro; Leonardo Azevedo; Maria Pereira — 2024 — PLoS

Key information

Authors:

Manuel Ribeiro (Manuel Ribeiro); Leonardo Azevedo (Leonardo Azevedo Guerra Raposo Pereira); André Peralta-Santos; Pedro Pinto Leite; Maria Pereira (Maria João Correia Colunas Pereira)

Published in

February 1, 2024

Abstract

During the SARS-CoV-2 pandemic, governments and public health authorities collected massive amounts of data on daily confirmed positive cases and incidence rates. These data sets provide relevant information to develop a scientific understanding of the pandemic’s spatiotemporal dynamics. At the same time, there is a lack of comprehensive approaches to describe and classify patterns underlying the dynamics of COVID-19 incidence across regions over time. This seriously constrains the potential benefits for public health authorities to understand spatiotemporal patterns of disease incidence that would allow for better risk communication strategies and improved assessment of mitigation policies efficacy. Within this context, we propose an exploratory statistical tool that combines functional data analysis with unsupervised learning algorithms to extract meaningful information about the main spatiotemporal patterns underlying COVID-19 incidence on mainland Portugal. We focus on the timeframe spanning from August 2020 to March 2022, considering data at the municipality level. First, we describe the temporal evolution of confirmed daily COVID-19 cases by municipality as a function of time, and outline the main temporal patterns of variability using a functional principal component analysis. Then, municipalities are classified according to their spatiotemporal similarities through hierarchical clustering adapted to spatially correlated functional data. Our findings reveal disparities in disease dynamics between northern and coastal municipalities versus those in the southern and hinterland. We also distinguish effects occurring during the 2020–2021 period from those in the 2021–2022 autumn-winter seasons. The results provide proof-of-concept that the proposed approach can be used to detect the main spatiotemporal patterns of disease incidence. The novel approach expands and enhances existing exploratory tools for spatiotemporal analysis of public health data.

Publication details

Publication version

AM - Accepted manuscript

Publisher

PLoS

Title of the publication container

PLoS ONE

First page or article number

e0297772

Volume

19

Issue

2

Fields of Science and Technology (FOS)

earth-and-related-environmental-sciences - Earth and related environmental sciences

Keywords

  • covid-19
  • geostatistics
  • functional data
  • classification
  • public health

Publication language (ISO code)

eng - English

Rights type:

Open access