SA’s disproportionate number of deaths by natural causes likely linked to COVID-19

Between May 2020 and October 2021, South Africa recorded more than a quarter of a million excess deaths, all of which are recorded as being related to “natural causes”. But according to Professor Tom Moultrie, the director of the Centre for Actuarial Research (CARe) at the University of Cape Town (UCT), researchers suspect that roughly 90% of these cases are directly linked to SARS-CoV-2.

What this means, Professor Moultrie explained, is that the nationwide COVID-19 death toll has probably far surpassed the current number of 88 754 cases (as of Thursday, 21 October). These worrying statistics have emerged, week on week, in the South African Medical Research Council’s (SAMRC) Rapid Mortality Monitoring System.

The tracking system, developed in partnership with a team of academics at UCT, was initially devised to track the number of HIV/AIDS-related deaths in the country. But in the wake of the COVID-19 pandemic, Moultrie said, the system was redesigned and repurposed to track the emerging deaths from COVID-19. The tracking system provides information on deaths registered in real time and logged with the Department of Home Affairs.

UCT News spoke to Moultrie and asked him what researchers know about this disproportionate number of deaths by natural causes, and the links to COVID-19.

Niémah Davids (ND): Please unpack the term “excess deaths”.

“Excess deaths are the deaths observed in a specific period above those that may have been expected under normal conditions and patterns of mortality.”

Tom Moultrie (TM): Excess deaths are the deaths observed in a specific period above those that may have been expected under normal conditions and patterns of mortality. Using historical data, we are able to build an accurate picture of the number of deaths (by age, sex and province) for any given week.

ND: Tell us about the Rapid Mortality System and how it works.

TM: In an attempt to get a handle on HIV/AIDS mortality in the early 2000s, the system was built by researchers at the SAMRC in conjunction with UCT’s Emeritus Professor Rob Dorrington. At the time, there was widespread denialism of the disease, and the system allowed us to paint an accurate picture of the emerging impact of HIV/AIDS mortality in the country.

We used the data on deaths notified to the Department of Home Affairs, and the monthly number of reported deaths were compared with those anticipated for the same time period. Thanks to this SAMRC–UCT collaboration, we were able to provide the first strong empirical demonstration of the emerging burden of mortality as a result of HIV/AIDS.

In January 2020, aware of the rapid spread of SARS-CoV-2 in China, the team began repurposing the system to provide estimates of the impact of COVID-19 on South African mortality. By the end of March 2020 we had rebuilt the system to use weekly data from Home Affairs to track excess mortality in near-to-real time. The redevelopment was completed and our first report was released the same week that South Africa recorded the very first COVID-19 death – that was about five weeks before the effects of COVID-19 truly became visible in the data. Since then, we have released a report every week with the exception of the week over Christmas 2020.

Unlike data on infections (which depend on the volume of testing, and the testing protocols in place at any point in time), or hospitalisations (which require detailed administrative data from hospitals), the data on excess deaths provides an incontrovertible marker of the scale of emerging mortality in the country.

ND: Examining your figures, what does the data reveal about those who are dying?

TM: We know that in South Africa, between May 2020 and October 2021, a quarter of a million more people have died of natural causes than we would have expected over this time – the majority of whom are over the age of 60.

“These results show that the most affected provinces are the Northern Cape, the Eastern Cape and the Free State.”

The data we get contains age, sex, and the province of the Home Affairs office at which the death was registered, which is not necessarily the area in which the deceased lived. Our public releases of deaths focus on province and the major metropolitan areas. We also provide a measure of mortality that accounts for the differential population size and age distributions of different provinces’ populations. These results show that the most affected provinces are the Northern Cape, the Eastern Cape and the Free State provinces; the Western Cape, Gauteng, and North West provinces are relatively less badly affected.

We are working to refine the spatial detail of our analysis, although the effort is some way away from getting to a point [at which we can] release these results.

ND: What is the leading cause of death among South Africans?

TM: Unfortunately, we don’t get information on the exact cause of death. All we get is an indicator as to whether the death was attributed to natural causes (illness), or unnatural causes (accidents or homicides).

What we have picked up is an exceedingly close correlation between COVID-19 waves in the country and the high number of excess deaths that are said to have occurred during these periods. The number of excess deaths per week follows the various peaks of reported COVID-19-related deaths. 

ND: What are your conclusions about the excess death rate and its parallels with COVID-19?

TM: Our work with colleagues at the SAMRC suggests that between 85% and 95% of excess deaths in South Africa might be attributed to COVID-19. But we can’t be 100% sure just yet. We will only be able to empirically validate this estimate once we receive coded causes of death, which may take several years before it’s reported by Statistics South Africa.

What we know is that the number of deaths reported by the national Department of Health each evening reflects those who have died of COVID-19. Sadly, those reports miss many deaths, especially those who didn’t die in a healthcare facility.

ND: Other than measuring the number of excess deaths, how else has the system been of value during the pandemic?

TM: Very early in the pandemic’s course in South Africa, there was a general realisation among health researchers that it was the trajectory of mortality that would provide the best indicator of the spread of the virus in a developing country. We are pretty much the only developing country in the world that’s effectively and accurately built a system of this kind. While it remains a lagging indicator, it has allowed us to debunk the myth that somehow Africa has escaped the impact of the virus.

“These findings point to an urgent need for more conversations about our collective national relationship with alcohol.”

In addition, our work has had further benefits. The sequence of lockdowns and alcohol bans during the COVID-19 peaks allowed us to investigate the impact of alcohol on unnatural deaths in the country, using a quasi-natural experimental design. These findings point to an urgent need for more conversations about our collective national relationship with alcohol.