COVID-19 Recoveries Data: How to use Responsibly

In our explorations, we have come across very few datasets that contain information on the number of people in African countries who have recovered from COVID-19. As the Africa Data Hub, we believe that it is important to share and report on  recoveries data in an effort to facilitate a narrative of hope and perhaps even optimism around the COVID-19 pandemic; that people can and do recover from this virus. However, there are some important caveats to bear in mind when exploring recoveries data. 

COVID-19 recoveries data is measured and estimated, not collected

First of all, recoveries are estimates, at best. It would be impossible for any country to have an exact count of the number of people who have recovered from COVID-19. It is impossible for any country to know exactly how many people have even been infected with COVID-19, and even if this number is known, they would have to follow up with every single case, testing them again to see if that person had recovered. This means that recovery numbers are calculated, rather than measured. This calculation is also not as simple as “number of cases minus number of deaths” as it takes time to recover or succumb to the virus. Presumably every country calculates its recoveries in its own way, though it is likely that if someone hasn’t died due to COVID-19 within a set number of days after testing positive, it can be assumed that they have recovered. Each country is likely to have its own threshold for the number of days before a case can be considered recovered. The value for recoveries will also be affected by how well the country is detecting cases and tracking deaths. If a country is doing a lot of testing and can be confident that it has detected most of its cases, and that country is also diligent in following up on its COVID-19 deaths (i.e. matching specific cases to deaths), then it is possible that the recoveries data for that country is fairly accurate. Unfortunately, many countries do not test nearly enough to be confident that they have detected most of their cases, and similarly do not follow up with tracking deaths. All of this uncertainty makes it difficult to believe that the recoveries data is accurate and reliable, and also makes cross-country comparisons difficult. Inaccurate and inconsistent reporting exacerbates this problem.

Recoveries data must be read with case and death numbers in mind

Secondly, recovery numbers are not an indicator of how well or how poorly a country is handling the pandemic. Recoveries are entirely driven by case numbers. A country with a very high number of recoveries is not necessarily handling the pandemic better than a country with low recovery numbers. It is thus perhaps more useful to focus on a ratio of the number of recoveries to the number of cases. This, combined with a ratio of the number of deaths to the number of cases, can provide one with a sense of how well or how poorly a country is performing in response to the pandemic, as a higher deaths to cases ratio would indicate that more people who contract the virus end up dying. These ratios also provide a handy metric for the quality of reporting for a particular country, as one would expect these ratios to be the inverse of one another - a high recovery ratio should accompany a low death ratio and vice versa. 

In conclusion, recoveries data provides some unique and hopefully uplifting insight into the state of the COVID-19 pandemic in Africa, but one should keep in mind that case and death numbers tell the real story. Recoveries should be reported on with caution and should only be explored within the context of cases and deaths.