Fractal dimension based geographical clustering of COVID-19 time series data.

Understanding the local dynamics of COVID-19 transmission calls for an approach that characterizes the incidence curve in a small geographical unit. Given that incidence curves exhibit considerable day-to-day variation, the fractal structure of the time series dynamics is investigated for the Flanders and Brussels Regions of Belgium. For each statistical sector, the smallest administrative geographical entity in Belgium, fractal dimensions of COVID-19 incidence rates, based on rolling time spans of 7, 14, and 21 days were estimated using four different estimators: box-count, Hall-Wood, variogram, and madogram. We found varying patterns of fractal dimensions across time and location. The fractal dimension is further summarized by its mean, variance, and autocorrelation over time. These summary statistics are then used to cluster regions with different incidence rate patterns using k-means clustering. Fractal dimension analysis of COVID-19 incidence thus offers important insight into the past, current, and arguably future evolution of an infectious disease outbreak.

The Effect of Transborder Mobility on COVID-19 Incidences in Belgium.

Belgium is a geographically small country bordered by The Netherlands, France, Germany, and Luxembourg, with intense transborder mobility, defined as mobility in the border regions with neighboring countries. It is therefore of interest to examine how the 14-day COVID-19 confirmed case incidence in the border regions is influenced by that of the adjacent regions in the neighboring countries and thus, whether and how it differs from that in the adjacent non-border regions within Belgium. To this end, the 14-day COVID-19 confirmed case incidence is studied at the level of Belgian provinces, well-defined border areas within Belgium, and adjacent regions in the neighboring countries. Auxiliary information encompasses work-related border traffic, travel rates, the proportion of people with a different nationality, the stringency index of the non-pharmaceutical interventions, and the degree of urbanization at the level of the municipality. Especially in transnational urbanized areas such as between the Belgian and Dutch provinces of Limburg and between the Belgian province of Antwerp and the Dutch province of North Brabant, the impact on incidence is visible, at least at some points in time, especially when the national incidences differ between neighboring countries. In contrast, the intra-Belgian language border regions show very little transborder impact on the incidence curves, except around the Brussels capital region, leading to various periods where the incidences are very different in the Dutch-speaking north and the French-speaking south of Belgium. Our findings suggest that while travel restrictions may be needed at some points during a pandemic, a more fine-grained approach than merely closing national borders may be considered. At the same time, in border regions with considerable transborder mobility, it is recommended to coordinate the non-pharmaceutical interventions between the authorities of the various countries overlapping with the border region. While this seems logical, there are clear counterexamples, e.g., where non-essential shops, restaurants, and bars are closed in one country but not in the neighboring country.

The COVID-19 wave in Belgium during the Fall of 2020 and its association with higher education.

Soon after SARS-CoV-2 emerged in late 2019, Belgium was confronted with a first COVID-19 wave in March-April 2020. SARS-CoV-2 circulation declined in the summer months (late May to early July 2020). Following a successfully trumped late July-August peak, COVID-19 incidence fell slightly, to then enter two successive phases of rapid incline: in the first half of September, and then again in October 2020. The first of these coincided with the peak period of returning summer travelers; the second one coincided with the start of higher education's academic year. The largest observed COVID-19 incidence occurred in the period 16-31 October, particularly in the Walloon Region, the southern, French-speaking part of Belgium. We examine the potential association of the higher education population with spatio-temporal spread of COVID-19, using Bayesian spatial Poisson models for confirmed test cases, accounting for socio-demographic heterogeneity in the population. We find a significant association between the number of COVID-19 cases in the age groups 18-29 years and 30-39 years and the size of the higher education student population at the municipality level. These results can be useful towards COVID-19 mitigation strategies, particularly in areas where virus transmission from higher education students into the broader community could exacerbate morbidity and mortality of COVID-19 among populations with prevalent underlying conditions associated with more severe outcomes following infection.