Brazil’s vulnerability to COVID-19 quantified by a spatial metric

Highlights

• •An easy-to-apply index that can support decision-makers to identify the most vulnerable regions to COVID-19.
• •The impact of the lack of resources in the country-side to fight SARS-CoV-2 can be assuaged by shielding selected regions.
• •IndCo can support countries to prepare the reopening process while controlling COVID-19 in their regions.

Since the global outbreak of the novel coronavirus (COVID-19) pandemic, governments are dealing with an unprecedented challenge. Decision-makers are forced to decide urgently where to locate new resources (e.g., hospital beds, availability of health workers) to fight this rapidly evolving and propagating epidemic. Kandel and coworkers [1] overviewed the preparedness of existing health security capacities with regard to infectious disease outbreaks. Many countries in Latin America and Africa show a highly uneven distribution of their health infrastructure, and a weak operational readiness capacity, especially in the countryside [2]. It is therefore crucial, for these wide and impoverished regions, to identify where improvements of the health system are needed most urgently, and, if the health care system could not be improved on a short notice, to spot those areas that are to be protected in priority. For example, Brazil’s continental area, and its universal health system, is divided into 5570 municipalities, of which 36% lack beds in hospitals to counter this kind of pandemic. To address these issues, we developed a straightforward and robust spatial metric, IndCo, and applied it to Brazil to illustrate its pertinence (see Fig. 1), using governmental public data [3,4]. We combined basic data on the municipalities in a Geographical Information System, namely, the proportion of elderly population (b), the number of inhabitants per available hospital bed (c), and the number of inhabitants per qualified health worker (d). To deal with municipalities without hospitals in their territory and hence a potential population flow of Covid-19 patients to their vicinity, the lack of hospital beds in adjacent municipalities was integrated in the IndCo metric, through combining the aforementioned basic data on municipalities (b, c and d) into an indicator characterizing the neighbors of each municipality (e). We categorized the data of the panels (b) to (e) into five classes, labeled 1 to 5, representing increasing vulnerability. These four different types of data were then summed, all with the same weight, to compose IndCo. The red color in panel (a) indicates the regions in which the combination of the cited indicators is to be considered alarming, where health systems are expected to be overloaded rapidly, and where short term state intervention will be required in order to prevent health systems from collapsing shortly after the Covid-19 outbreak. IndCo could be improved in future works by integrating the epicenter(s) of contamination or by inclusion of spatial dispersal models [5].

Fig. 1.

(a) IndCo index applied to Brazil. The color scale enables to identify localities characterized by a critical combination of a large proportion of elderly aged people (b), a high number of inhabitants per hospital bed (c), the presence of a low number of health care workers (d), and by a lack of available beds in hospitals in the locality itself and in its neighboring municipalities (e). (For interpretation of the references to color in this figure legend, the reader is referred to the Web version of this article.)

Funding

None.

Ethical approval

Not required, once this work uses only public data from Brazilian agencies.

All the data used in this study will be available at <https://www.labgeotec.com.br/

Authors’ contributions

D.A. conceived the idea and designed the index, and together with L.P., worked on the data acquisition and interpretation. J.B. and L.M. provided analysis support; all the authors discussed data and contributed equally to the manuscript writing.

Declaration of competing interest

None declared.