In South Africa at least 5 000 people are killed on the roads every year. Or is it closer to 14 000? Or more than those dying of HIV/Aids and tuberculosis, as has been claimed?
Roadside memorials marking the locations of fatal crashes have become a common sight.
Suppose a roadside memorial were to be erected for all South African crash victims in a given year.
Would the number exceed the deaths from tuberculosis and HIV/Aids?
Yes, according to a press release issued after a recent meeting of Parliament’s portfolio committee on transport. It stated: “Carnage on the roads causes more deaths in South Africa than HIV and TB.”
How true is the claim?
The Department of Home Affairs registers all deaths, which Statistics South Africa (Stats SA) then analyses to produce an annual report.
According to Stats SA’s 2012 Mortality and Causes of Death report, released last month, TB was the leading killer, accounting for around 9.9 percent of all deaths in that year.
Year Deaths
2012 47 472
2011 54 827
2010 63 375
2009 69 849
2008 75 249
Obtaining accurate data on HIV/Aids deaths is trickier.
The Mortality and Causes of Death report suggests that HIV/Aids accounted for just 3.9 percent of deaths (18 663) in 2012.
But HIV/Aids deaths are said to be 10 times higher than those recorded on death certificates and do, for instance, account for many deaths ascribed to diseases like TB.
Ria Laubscher, a senior scientist at the Medical Research Council’s Burden of Disease Research Unit, said: “It is important to note that a large number of HIV deaths have been misattributed to the immediate causes of death such as tuberculosis, diarrhoeal diseases and lower respiratory infections.”
Stats SA therefore calculated an estimate which took into account data from other sources, Stats SA’s manager for demographic analysis, Diego Iturralde, explained.
“In particular we use the HIV prevalence rates coming from the antenatal clinics and then we also use the information coming from Department of Health related to antiretroviral treatment for adults and children separately,” Iturralde said.
“This is compared to other causes of death (as reported in the Mortality and Causes of Death report) that are HIV-related but coded as something else in order to convince ourselves that the numbers are correct.”
By following this process, Stats SA estimated that in reality 203 293 deaths could be attributed to HIV/Aids in 2012.
Were more people dying on our roads? Not nearly, it would seem.
The corresponding figure in the 2012 Mortality and Causes of Death report is almost 40 times lower than the Aids death estimate:
Year Deaths
2012 5 284
2011 5 164
2010 5 698
2009 5 783
2008 5 818
Confusingly though, the number of deaths reported by the Road Traffic Management Corporation differs.
The corporation is a government agency responsible for compiling and analysing accident and traffic data and gathers its information from the SAP’s accident reports.
A senior manager at the traffic agency, Magadi Gainewe, provided figures for up to 2011, but said the data for 2012 had not yet been approved and could not be released.
The traffic agency’s total for 2011 is almost three times that of Stats SA:
Year Deaths
2011 13 947
2010 13 967
2009 13 768
2008 13 875
How do we then know what the road death toll really is?
While deaths were fairly accurately recorded they were often poorly categorised and this affected the Mortality and Causes of Death report, said Richard Matzopoulos, a specialist scientist at the Medical Research Centre’s Burden of Disease Research Unit.
“Even though the reporting of total injury deaths is fairly complete, the (death certificates) have poor coding of injury deaths, that is, a poor distinction between road accidents, other unintentional causes, homicide, etcetera,” he said.
On the other hand, the accuracy of the traffic agency’s statistics is affected by underreporting.
“This forces them to estimate by applying an adjustment factor to take into account the unreported deaths, which are disproportionately pedestrian deaths,” Matzopoulos said.
Previously, a death was recorded as a road fatality when a person died within seven days of the accident.
This changed last year when the Road Traffic Management Corporation adopted the international standard of counting all deaths that occurred within 30 days, Gainewe said.
Still, it is dependent on the police to follow-up with hospitals on whether a crash victim has died, a practice that has been described as “problematic”.
Given the way the traffic agency’s data was adjusted to provide for underreporting, Matzopoulos estimated it could be 20 percent higher – therefore in the region of 16 700 road deaths for 2011.
The agency’s chief executive, Makhosini Msibi, detailed a grand plan for improving road fatality statistics at the portfolio committee on transport’s meeting.
He attributed inaccuracies to different reporting systems used in provinces and said that the agency planned to create a standardised system as well as appoint accident information officers at key police stations.
He said it would also work with the departments of Health and Home Affairs, the Council for Scientific and Industrial Research, the Medical Research Centre and the SAPS to improve data accuracy.
But the personnel we spoke to at the Medical Research Centre were not yet aware of this. We made various attempts over the course of 10 days to contact Msibi. He did not return calls or text messages.
The Western Cape province had improved its road fatalities data since following the recommendations made by the Medical Research Council in 2009, said a spokesman for the Transport MEC, Siphesihle Dube. Raw data was collected daily and was available online.
Statistics showed that on average 6.2 people were killed a day over pay day weekends – ascribed to people having money to buy alcohol and travel – up from 3.6 people, allowing his department to warn motorists to be vigilant.
Road accident deaths in South Africa are far lower than those that can be attributed to HIV/Aids and tuberculosis.
But we cannot say for sure how much lower.
The country has adopted a UN resolution to reduce traffic fatalities by half by the end of the decade. In order to achieve this, the Road Traffic Management Corporation must implement improvements to traffic fatality statistics as a matter of urgency. Accurate data forms an integral part of infrastructure and enforcement planning and policy implementation.
Marius Luyt, a spokesman for the Automobile Association, summed it up: “If you know 500 pedestrians were killed on a stretch of road, you could put up a bridge there.”
* This article first appeared on Africa Check (www.africacheck.org), a non-profit organisation run from the Journalism Department, University of the Witwatersrand, which promotes accuracy in public debate, testing claims made by public figures around the continent.