ABSTRACT
In the current pandemic context, it is necessary to remember the lessons learned from previous outbreaks in Africa, where the incidence of other diseases could rise if most resources are directed to tackle the emergency. Improving the access to water, sanitation and hygiene (WASH) could be a win-win strategy, because the lack of these services not only hampers the implementation of preventive measures against SARS-CoV-2 (e.g. proper handwashing), but it is also connected to high mortality diseases (for example, diarrhoea and lower respiratory infections (LRI)). This study aims to build on the evidence-based link between other LRI and WASH as a proxy for exploring the potential vulnerability of African countries to COVID-19, as well as the role of other socioeconomic variables such as financial sources or demographic factors. The selected methodology combines several machine learning techniques to single out the most representative variables for the analysis, classify the countries according to their capacity to tackle public health emergencies and identify behavioural patterns for each group. Besides, conditional dependences between variables are inferred through a Bayesian network. Results show a strong relationship between low access to WASH services and high LRI mortality rates, and that migrant remittances could significantly improve the access to healthcare and WASH services. However, the role of Official Development Assistance (ODA) in enhancing WASH facilities in the most vulnerable countries cannot be disregarded, but it is unevenly distributed: for each 50-100 US$ of ODA per capita, the probability of directing more than 3 US$ to WASH ranges between 48% (Western Africa) and 8% (Central Africa).