Africa faces difficult choices in responding to COVID-19.

Coronavirus disease 2019 (COVID-19) is now established in Africa, with more than 63 000 cases and 2200 deaths in 53 countries, as of May 11, 2020. Fragile health systems leave African countries vulnerable to the anticipated surge in severely ill patients with COVID-19, despite much younger populations.

To flatten the curve, some African governments have imposed stringent public health measures (lockdown) based on physical distancing to reduce transmission. However, the safety of this approach in poor communities has not been evaluated, and it is plausible that lives lost to lockdown could exceed those saved from COVID-19. Potentially fatal unintended consequences include widespread economic disruption and hunger, worsening food insecurity if harvesting is disrupted, and increased domestic and state actor violence. Large numbers of African patients with HIV and tuberculosis depend on functional health services, with substantial individual and public health consequences if treatment access is disrupted. Although anticipated by national programmes, some treatment interruptions are inevitable during prolonged lockdown.

With clear understanding of risk, governments can make informed decisions about harms and benefits. We used Spiegelhalter's approach to compare age-group specific infection fatality ratios from COVID-19 to background (non-COVID-19) mortality risk in Malawi, South Africa, the UK, and India.3, 4, 5 This assumes COVID-19 infection fatality ratios similar to China, but true age-specific case-fatality rates might be higher with fragile health systems. For context, Malawi has not yet triggered lockdown, whereas the UK, South Africa, and India have. We estimate that in the UK, having COVID-19 confers risk of death equivalent to approximately 12 months of background mortality risk, averaged across all age groups. By contrast, in Malawi this risk is equivalent to 4 months of background mortality (appendix). This reflects higher background mortality rates in Malawi, underscoring the fragility of health under normal circumstances.

Malawi (median age 17 years) also has relatively few older citizens, with 6·6% of the population older than 60 years. This makes alternative strategies potentially safer and more feasible than lockdown—eg, community-led approaches to support older people to self-isolate with provision of food, medicine, and wellbeing support.

Although we fully agree that macroeconomic arguments against lockdown cannot justify widespread loss of life in Europe and Asia, the considerations are very different in Africa, where lockdown could cost many lives. We urge African governments to carefully contextualise safe physical distancing policies that maximise likely benefits. Without a context-specific, ethical approach to physical distancing, unintended harms from stringent lockdown could pose more harm than the direct effects of COVID-19 itself.