#Data4COVID19 Africa Challenge Concludes: Major Takeaways

By Andrew J. Zahuranec and Uma Kalkar

This piece was also published on the #Data4COVID19 Africa Challenge blog site. Please visit to read updates throughout the challenge, blogs from project teams on their work, and other insights.

Copyright: World Bank / Ousmane Traore, CC BY-NC-ND 2.0

Almost one year ago, The GovLab, Agence française de développement (AFD), and Expertise France announced the #Data4COVID19 Africa Challenge. This initiative was an extensive undertaking, an effort to support data-driven action by Africa-based organizations toward the worst consequences of the COVID-19 pandemic. Dozens of teams across the continent submitted innovative proposals for how they might use data to respond to the pandemic. They suggested new and interesting collaborations. After a rigorous review process, the challenge sponsors selected seven teams to support through funding, access to expertise, and other resources.

Though there were substantial challenges along the way, the #Data4COVID19 Africa Challenge completed its work earlier this month. After six months working to launch their initiatives, which spanned from screening vulnerable populations to helping students in lockdown to addressing information, each team presented on the outcomes of their work to the challenge sponsors. While many of these initiatives will continue past the funding period, the end of the Challenge provided a moment for reflection and conversation on how far teams had come.

Below, we share the presentations that each team gave. We believe that these documents help capture what problems the teams targeted, what they did, and how that work will inform real-world action.

***

Institut National d’Hygiène Publique (INHP)

Led by Natalia Adler, the Institut National d’Hygiène Publique, Pebble Analytics, and Mitiga presented their work developing a participatory early warning system in Cote d’Ivoire: Afya App. The team discussed its use of search query data, self-reporting, and other data sources to detect cases of COVID-19 early and inform the public. The team spoke on how they would strengthen the app after the challenge by improving its UX, formatting its questionnaire as a chatbot and making the weekly survey as easy as possible for users. They also provided some insights on their work to spread word on the app and ensure people continue to use it.

Nigeria Centre for Disease Control

Led by Dr. Chinwe Ochu, the Nigeria Centre for Disease Control in partnership with the College of Medicine at the University of Lagos discussed its work to understand the factors affecting compliance with non-pharmaceutical COVID-19 preventative measures in Nigeria. Using six non-traditional datasets derived from online and phone-based surveys, the team found that most respondents followed stay-at-home orders throughout the crisis but that economic insecurity, hunger, and ignorance could spur non-compliance. More findings can be found in the team’s lay report here.

Population Council, Kenya

The Population Council, Kenya team (supported by the Busara Center for Behavioral Economics), meanwhile, presented on their work to understand “behavioural responses to the pandemic” to support Kenyan leaders in developing effective guidelines that could limit disease spread. The group discussed their work using machine learning to analyze Twitter to identify common topics of conversations. Among other findings, this work revealed that “a sharp increase in positive tweets [on vaccinations] along with a corresponding fall in neutral tweets after September 2020 which was when the second wave of case[s] was coming in Kenya.”

Code for Africa (CfA)

Code for Africa discussed their work on OUTBREAK.africa, a forthcoming effort to promote data-driven analysis on COVID-19 “to fight the infodemic and help decision makers identify how to geographically distribute vaccines within a country depending on it’s vaccination prioritisation rollout plan.” The group discussed how it had awarded 13 fellowships to journalists in Senegal, Cameroon, Nigeria, and Kenya to spur data journalism, and its fact-checking initiative, PesaCheck. The group also noted that it is collecting a “Rolodex of Scientific Experts” and developing an “OUTBREAK AFRICA” open data portal to support data-driven analysis in Cameroon, Kenya, Nigeria, Senegal, and South Africa.

IRD_ARCAD Santé Plus

Luis Sagaon Teyssier and his colleagues at ARCAD Sante Plus presented their work quantifying missed opportunities for COVID-19 screening in people with HIV/AIDS (and other vulnerable groups) across Mali. The researchers spoke about how they used information from the Nardis electronic medical records dataset of people living with HIV/AIDS in Mali to identify those who reported at least one potential COVID-19 symptom between 1 March 2020 and 31 August 2021. They then classified symptoms and devised indicators of “suspected cases.” Their work highlights a missed COVID-19 testing opportunity and a lack of equipment and trained professionals capable of addressing the needs of this group. It points to the need for a targeted COVID-19 testing strategy in Mali.

Université Alioune Diop de Bambey, Senegal

The team representing Université Alioune Diop de Bambey and their collaborators at the Université virtuelle du Sénégal and the Technical University of Munich presented on their study to understand the “social economic, and technical barriers” impeding effective online higher education in Senegal. Relying on survey methods, the team found that many Senegalese students were impeded by social conditions (e.g. living in a rural area), economic aspects (e.g. lack of financial support from parents and other figures), and technical difficulties (e.g. lack of stable internet connection). The research team noted that these preliminary conclusions could be used to support educational strategies by universities in Senegal and they noted ongoing conversations with several such institutions.

Data Science Institute

The Data Science Institute at the Institut National Polytechnique Félix Houphouët-Boigny and the consulting firm data354 unveiled ARSCO, a dynamic tool to monitor public perception of government-imposed health measures and overall COVID-19 information on social media networks. The researchers used machine learning models to analyze tweets posted between 28 July 2021 and 15 November 2021 to discern tweet tone and themes and investigate the relationship between tweet activity and COVID-19 statistics among French-speaking and local-language groups in Côte d’Ivoire. Their study underscores the importance of social media data analysis to inform and tailor communication and public policies for different parts of the population.

***

The research conducted by the #Data4COVID19 Africa Challenge teams will inform real-world action to counteract the pandemic. We encourage those interested in this work to check this blog regularly for final updates from the teams. We also encourage you to follow the teams at their respective websites.

Feel free to contact us if you have any questions or concerns about this work at datastewards [at] thegovlab.org.

Data Stewards Network

Responsible Data Leadership to Address the Challenges of…