‘Flawed’ model used to justify booze ban under fire
But the scientist behind the predictions stands by them, pointing to the large reduction in trauma cases reported at hospitals since the ban was reinstated
SA’s largest alcohol companies say the model used to justify SA’s booze ban fails to prove the causal effect between consumption and trauma injuries.
Charles Parry, the SA Medical Research Council director for research into alcohol, tobacco and other drugs, was asked by the government to provide a model that calculated how an alcohol ban would reduce the number of trauma patients in hospitals. The government’s concern was that every hospital bed and staff member would be needed to treat severely ill Covid-19 patients.
Parry suggested, from new and old studies, that half the trauma cases in the country were alcohol related, and he said prohibiting the sale of alcohol would reduce the number of all trauma cases by 10% in the first week, 20% in the second week and 40% after that.
The alcohol industry – including SAB, Distell, Heineken, Diageo (which makes Gordon’s gin), Pernod Ricard and VinPro (an association of more than 2,500 wineries in the Cape) – asked data and consulting company Kantar to review Parry’s model.
The Kantar review has pointed out various problems with the model. This includes the following:
- It uses data from 365 hospitals gathered in 1999, and does not adjust for the increase in the number of hospitals and the fact that more beds are now available to treat trauma patients;
- Critically, it doesn’t provide the evidence for the assumption that 50% of the trauma cases in SA are linked to alcohol; and
- It doesn’t separate what led to a reduction in trauma during the first ban in the lockdown – whether it was the 8pm curfew, fewer traffic accidents, fewer pedestrians on the road or fewer workplace injuries.
In a statement, the industry says: “Given the current data available, the extent of a causal relationship between alcohol sales and trauma-case presentations is difficult to isolate.”
But Parry, who admits to limitation in the model, thinks that he may actually have underestimated the impact of alcohol consumption on the number of car accident and injury patients.
For example, he says, data from Joburg’s Charlotte Maxeke academic hospital shows an 83% reduction in trauma patients in the week after the second alcohol ban, figures from Helen Joseph Hospital shows a 41% drop and the numbers from Port Elizabeth’s Dora Nzinga Hospital shows a 50% decline.
Still, the industry argues that Parry’s model fails to prove the causal link between excessive drinking and stabbings and alcohol accidents, in part because the alcohol levels of trauma patients aren’t tested.
But Parry argues he can prove this link. He points to a public health standard called the Bradford Hill criteria, which outlines what needs to be proved to suggest causation.
Briefly, it says that to prove cause and effect, scientists must look at the order in which things happen and how consistently they happen, the link must be able to be explained scientifically and the role that the quantity of something (like alcohol) has on the effect (like trauma) must be explained.
Parry argues this can be determined from a series of questions, including: Do people drink before trauma rather than afterwards? Is there a link between how much a person drinks and the likelihood of injury? And: Is there a biological mechanism that can explain what is happening?
On all three, he argues the answer is yes. For the last question, he says it’s well established that alcohol lowers a person’s inhibitions and perception, especially when driving.
‘Hardship and suffering’
Nonetheless, the industry believes the government should review the model and the ban, both for purely scientific reasons and because of the number of people who are losing their jobs as a result of the prohibition.
“We would like to believe that the government will consider the glaring limitations of the research by Prof Parry and the severe economic impact that the resultant decision to prohibit alcohol sales has on the country, including on our ability to contribute more than R51bn in tax revenue per year,” says Sibani Mngadi, the industry’s spokesperson.
Mngadi’s optimism may be misplaced: until this point, the industry’s attempts to lift the ban have come up against a brick wall.
The industry has sent at least four letters to the government, each of which details how the ban has decimated the livelihoods of hundreds of thousands of people, from restaurant owners to bottlers.
The most recent letter, sent on Monday to President Cyril Ramaphosa by 1,400 bottle store owners who belong to the Liquor Traders Association of SA, warns of “tremendous hardship and suffering”.
“If liquor stores are not reopened soon, a significant number of employees will face retrenchment. Many of [them] are unskilled or semi-skilled and will find it very difficult to find employment once retrenched,” the owners write.
The industry has been supported by the restaurant industry, which has also held at least two significant protests, using a campaign slogan #jobssavelives. None of this has worked.
What the industry hasn’t yet attacked is the assumption that every hospital bed will be needed by Covid-19 patients.
This assumption is far from certain, in part because predictions on bed use have been wildly inaccurate.
For example, a presentation on May 19 by modellers working under the guidance of the National Institute for Communicable Disease suggested between 4,000 and 5,300 ICU beds would be needed at the peak of the epidemic in the Western Cape. In the end, less than a tenth of that – 390 beds – were required at the peak.
Equally, that presentation estimated that 11,000 to 14,000 hospital beds would have to be available for Covid-19 patients in the province at the peak. Ultimately the maximum number of beds used was less than a fifth of that, or 1,900.
It’s a critical point, considering that the cornerstone for the government’s alcohol ban is the argument that there won’t be enough beds for Covid-19 patients. While it’s fair to be worried about the number of beds, the predications have been notoriously incorrect. Which raises questions about whether an outright ban, at the cost of thousands of jobs, was completely necessary.
Would you like to comment on this article or view other readers' comments?
Register (it’s quick and free) or sign in now.
Please read our Comment Policy before commenting.