Following on the recent more-than-doubling of 'official' Peruvian Covid-19 deaths (theguardian.com/world/live/202…) , perhaps a similar exercise is worth considering in South Africa? A thread ...
While officially reported Covid-19 deaths in South Africa are almost 57 000, the @MRCza and @UCT_news collaboration estimates that there have been more than 160 000 excess natural deaths in the country since May 2020. /2
The evolution of those excess deaths correlates closely with both the proportions testing positive, as well as the reported deaths (when the latter are analysed according to date of death). The full report on this is here: samrc.ac.za/sites/default/… /3
That report suggests that perhaps between 85 and 95% of the excess deaths are attributable to Covid-19. This would push the number of Covid-19 deaths in the country to around 145 000, a similar proportional increase to that recorded in Peru. /4
Yet government has been reluctant to accept the implications of the excess deaths monitoring. See, for example, this: news24.com/news24/southaf… /5
Problem is, we do not KNOW how many of the excess deaths in South Africa might truly be Covid. This is because the data received in near- real-time used to monitor the excess deaths does not contain information on cause of death. /6
All that is received is whether the death was due to natural or unnatural causes; the latter being accidents, homicide, suicide etc. The coding of causes of death is undertaken by StatsSA, and is a laborious task (even as automation is increasingly used). /7
Sub-thread: a longish thread on coding of causes of death here:
https://twitter.com/tomtom_m/status/1352522312533336064/8
The last StatsSA report on causes of death was released in March 2020, covering deaths up to 2017. On the present system it will take YEARS before we truly understand the burden of Covid-19. What are the potential solutions? /9
1.Short-term: funding for a rapid investigation into a (large) sample of death notification forms and to manually attribute a cause of death, which would allow a more accurate estimate of the number of Covid deaths to be derived /10
2.Long-term: overhaul the current notification system, to move it from a paper-based to an electronic system which would allow more, and better, information to be captured in near- real-time. /end
• • •
Missing some Tweet in this thread? You can try to
force a refresh
