The number of “excess deaths” from COVID-19 are often angrily disputed. The SA Medical Research Council’s Prof Debbie Bradshaw, explains in an interview with TimesLIVE’s Tanya Farber, the systemic challenges in South Africa to compiling an accurate estimate.
Although South Africa’s COVID-19 infections have dropped to their lowest level since the first few weeks of the pandemic, an estimated 268,183 people have died from the virus (as at 31 October), three times more than the official death toll of 89,251 (as at 4 November).
The discrepancy is a repeat of what Bradshaw, chief specialist scientist at SAMRC, saw in the late 1990s with HIV, when she helped to design the Rapid Mortality Surveillance (RMS) system being used to estimate “excess deaths”. HIV were grossly underreported because of Aids denialism at the highest levels of government. But this time the challenges are systemic rather than political, she says.
RMS uses data from previous years to determine expected natural deaths in any week of the year based on seasonal variations. This number is compared with actual deaths to determine “excess deaths”, then analysed as to whether COVID can be primarily blamed for those excess deaths
One reason why official COVID-19 deaths are lower than the true figure is because only hospital deaths are counted. Even in hospitals, “some patients die from COVID before they have been tested while others are dead on arrival in an ambulance”, she said.
Another is the time lag caused by SA’s paper-based system for detailed cause of death data. Bradshaw said the forms doctors complete are submitted to Home Affairs before going to Stats SA. “Forms from 2020 are still being processed [at home affairs] and Stats SA is currently working on deaths from 2019,” she said.
Another issue is a 2014 regulation that keeps cause of death information entered by a doctor confidential. “The page with the cause of death detail self-sealing, with the specification that it may only be opened by an official from Stats SA,” said Bradshaw.
“Stats SA is required to treat all information confidentially and not share it with anybody. “There has to be a change so that details of the causes of death can be shared with the heath department — not only for COVID but also things like our cancer registry. This is a top priority if we want the country’s data to be used for public health,” said Bradshaw.
Additionally, provincial health departments rely on health facilities to report the number of people who have died of COVID, and Bradshaw thinks “there may be some dropout in that system”. “Not all of the deaths get through to the provincial head, and we also suspect some private sector deaths are not being recorded in the provincial system.”
This has led to uneven accuracy across the provinces: while the Western Cape’s reported COVID deaths are more than 70% of excess deaths, in Mpumalanga they are only 10%.
Several other countries also have triple the number of excess deaths versus reported deaths, and the World Health Organization is trying to standardise the process.
“They have asked countries to review their approach to estimating excess deaths, and in some countries there has been a deliberate downplaying of numbers,” said Bradshaw. “In South Africa, I don’t think this is the case. Our systems are just not functioning as effectively as we would like.”
TimesLIVE article – Real number of SA's Covid-19 deaths lost in all the paperwork (Open access)
See more from MedicalBrief archives:
UCT Centre for Actuarial Research: SA’s COVID death toll almost 3x official figure
SA and world COVID deaths 3x higher than official figures — The Economist
SAMRC/UCT analysis: Pandemic deaths in South Africa approach 300,000
Concern about other serious diseases falling through the cracks in SA
COVID1-9 pandemic is the deadliest event in US history