Blog

In response to the varied impacts of COVID-19 on social dynamics, economic systems, and health systems, there is renewed need for evidence-based decision making. Several African countries have taken a ‘data-driven’ response position. For instance, South Africa, has used the national COVID-19 Epi Model (NCEM) [i] to predict transmission dynamics, severity and treatment pathways using data on COVID-19 incidence and incorporating a cost model to prioritise resources and forecast gaps. These models enabled policy actors to institute various mitigating measures for COVID-19. Similar data-driven models are used across the continent, including in Kenya [ii] and Nigeria [iii], to guide policy and social management of COVID-19 within health systems. But apart from reliance on data, mode of application and efficiency, modelling must also account for underlying social and cultural systems. Unfortunately, the use of social sciences and policy research data in modelling to address COVID-19 is limited in Africa.

In this blog, we explore: (a) the state of social science and policy research and (b) modelling the impact of COVID-19 on health systems. We draw on a high-level dialogue among African policy makers, researchers and scientists that took place in June 2021. The “tele-convening” was part of a broader project on science engagement to support evidence-informed policy responses to COVID-19 in Africa which considers targeted, relevant, and context-specific rapid review of existing evidence for priority areas to inform the COVID-19 policy responses of African governments.

Key messages

The discussion emphasised the agency of the community in data-driven decision making through:

• Enhanced care rather than control of populations. The politics of control and care are fundamental when using data to define decision making. Data-driven results can undermine the capacity of individuals and local communities to make decisions relevant to their own social and structural settings when results incorrectly assume collective agency of decision making. For instance, models of incidence rates may suggest closures or lockdowns to reduce spread. But these measures affect the capacity of households to pursue necessary activities of daily life, which may result in individuals feeling that they have no choice but to ignore quarantine policy. Therefore, research incorporating social science and policy research work along with disease data must be complementary and mutually informative across sectors to optimise outputs. By framing considerations in terms of care rather than control, the public is less likely to regard policy as being driven by narrow interests and more likely to invite collaboration and dialogue.
• Development of heterogenous COVID-19 mitigation measures. The solutions instituted in African countries so far have not sufficiently accounted for social protection in various contexts, societies, and structures – continentally, nationally, or locally. Most government and development partner approaches to mitigation assume that the effects of the pandemic and its impacts on states, communities, individuals, and households are homogenous, and therefore produce generalised mitigation measures. It has been assumed, for example, that the overriding need of communities under lockdown is food, ignoring or underestimating other societal needs. This false premise has had serious secondary effects, including increased Gender-based Violence (GBV). Mitigation measures, driven by data and other inputs, need to fully account for multiple vulnerabilities, and be integrated in existing social systems to ensure sustainability.
• Drawing lessons from past pandemics and health challenges. Pandemics are not new in Africa: Ebola, swine flu, HIV and tuberculosis are all examples. However, databases on response strategies and interventions have not been satisfactorily disaggregated, coordinated or collated. This frustrates the ability to draw on lessons of the past to manage the present and future epidemics. It is the responsibility of public institutions and private entities to share, collate and provide access to historical data to ensure that models are built accurately and contextually relevant. This calls for political and social responsibility of institutions to document learnings effectively.
• Strengthening relationships between decision makers and communities. Decision makers and communities obviously have different priorities and are subject to different power dynamics. Power dynamics are propagated by privilege and reinforced by geopolitical culture. For instance, while many decision makers, who tend to be privileged and in positions of power, have access to the necessary resources to allow them to work from home, most “regular” community members in Africa do not. Mitigation measures such as lockdowns, even when effective, can be a hindrance to the ability of people to earn a living. There is need to develop models that promote and encourage proactive (not reactive) engagement and collaboration to find ways to minimise risk without penalty.
• Enhancing investments in R&D to create and use data for managing COVID-19 and other health challenges. African countries still lack adequate data generation, management, and utility systems. Health data collection and related cascades usually derive from responses to emergencies or events. The consensus of discussants is that data collection, management and sharing systems could benefit from a coordinated continent-wide system rather than be built on isolated national systems. 

Data-driven policy making is fundamental to the development of mitigation measures to support effective and fair responses to pandemics in Africa (e.g., COVID-19). Just as important is that the data accounts for contextually relevant policy options, informed by African countries and communities. These contextual differences must consider the heterogeneity of the continent, draw on lessons from previous epidemics, more rigorously include social science and policy research, and model the context-specific impacts on health systems.

---

[i] South African COVID-19 Modelling Consortium (2020), National COVID Epi Model (24^th July 2020 Version). https://sacovid19mc.github.io/ [Accessed July 14^th 2021].

[ii] Mark Nanyingi (2020), Predicting COVID-19: what applying a model in Kenya would look like, The Conversation. https://theconversation.com/predicting-covid-19-what-applying-a-model-in-kenya-would-look-like-134675 [Accessed July 14^th 2021].

[iii] Roseline O.Ogundokun et al (2020), Predictive modelling of COVID-19 confirmed cases in Nigeria. https://doi.org/10.1016/j.idm.2020.08.003. [Accessed July 14^th 2021].