While explaining why lifting the ban on the sale of alcohol was “inappropriate”, Cele said 34,000 hospital beds were occupied by alcohol-related trauma patients “as we speak”.
Prof Andrew Nicol of the University of Cape Town, who the ministerial spokesperson said was the source of the figure, confirmed he had been involved in an as-yet-unpublished study modelling the effect of the lockdown on trauma admissions.
But the study did not predict close to 34,000 extra or current alcohol-related admissions and experts had no idea how this figure was obtained.
Restrictions are now slowly being eased. The sale of alcohol, which was banned during the first two months of lockdown, has been allowed since 1 June.
“34,000 beds are occupied in the Republic of South Africa as we speak,” he said. “That’s why some of us, we felt very strongly ... that it was inappropriate to open alcohol now.”
Does available data support the minister’s claim? We checked.
Hospital admission study under peer review
The minister’s spokesperson, Lirandzu Themba, told Africa Check that the minister was quoting information from Andrew Nicol. He is the director of the Groote Schuur Hospital trauma centre and professor of surgery at the University of Cape Town.
But Nicol was unfamiliar with the 34,000 number. He recently assisted with a modelling study on trauma admissions in South African hospitals. But it did not produce this figure, he said.
The study was conducted by Nicol together with Prof Charles Parry, director of the Alcohol, Tobacco and Other Drug Research Unit at the South African Medical Research Council, Richard Matzopoulos, epidemiologist and specialist scientist in the council’s Burden of Disease Research Unit, and Prof Pradeep Navsaria, specialist trauma surgeon at Groote Schuur Hospital. The study is still being peer-reviewed, the process where academic studies are evaluated and validated. It is yet to be published.
“We wanted to come up with an estimate of how much trauma was decreasing because of the lockdown and how much would come back if alcohol sales were allowed to resume,” Parry explained during a presentation of the model on 23 June.
The model used trauma admission statistics from 1999 and adjusted them to account for the increase in South Africa’s adult population.
Alcohol banned: 6,750 alcohol-related trauma admissions per week
The authors estimated that before the lockdown, there were approximately 42,700 trauma admissions per week across 350 secondary and tertiary hospitals. About 40% of those trauma admissions (17,000) were alcohol related.
During the lockdown, the model suggests trauma admissions dropped to 15,000 per week. Of these, 6,750 admissions were alcohol related.
But the drop cannot be solely attributed to the ban on alcohol. Parry noted that during this time, people were also confined to their homes and there were less people on the road.
Alcohol ban lifted: 11,750 alcohol-related trauma admissions per week
The model also looked at how these figures would change should the alcohol ban be lifted, as it was on 1 June.
Parry and his colleagues estimated that, in addition to the 6,750 alcohol-related trauma admissions during lockdown, hospitals would see a further 5,000 admissions per week. This would bring the total number of alcohol-related admissions to 11,750.
“I can’t see how the minister would get a number of 34,000 alcohol-related trauma admissions when alcohol sales resumed,” Parry told Africa Check.
Tests for alcohol costly, require patient consent
How do doctors decide which cases are alcohol related? And do all alcohol-related trauma cases require hospitalisation?
Formal testing for alcohol levels can take the form of blood, breath or urine analysis, Matzopoulos told Africa Check.
But hospitals cannot test the alcohol levels of all patients because of financial and ethical reasons, said Prof Elmin Steyn, head of surgery at Tygerberg Hospital in Cape Town and head of the academic surgical department at the University of Stellenbosch. Tests are costly and the consent of the patient is required.
“We only do tests if the answer will help us to make treatment decisions,” she said. “By the time the results of tests are available, the intoxication may have worn off.”
Although not all patients are tested, alcohol is “easily smelled by doctors on the breath of patients”, Matzopoulos said, which will influence how cases are managed.
Not all alcohol-related trauma cases will be admitted into a bed in the hospital, Steyn told Africa Check.
“However, if one defines a ‘bed’ as an emergency unit bed or trolley or stretcher, there could be a few thousand intoxicated patients all over the country, somewhere in an emergency room or trauma unit in South Africa.”
Alcohol does affect trauma load
The minister’s claim was baffling, Steyn said. But she added that alcohol does affect a hospital’s trauma case load.
“There is the obvious increase in all mechanisms of injury at times when alcohol is bought and consumed in great volumes.”
“In addition we see the obvious state of intoxication of the injured, drivers, perpetrators, passengers and victims. Many will not be admitted, but they cannot be discharged until sober, so they do indeed overload the system while they are in the emergency assessment areas.”
Those that do need to be admitted for surgery or intensive care then utilise resources that could be used for others, she said.
Conclusion: No data shows 34,000 beds occupied with alcohol-related cases
South Africa’s minister of police, Bheki Cele, claimed that 34,000 hospital beds were taken up by alcohol-related cases. His spokesperson said the information came from Prof Andrew Nicol. But Nicol was not aware of the number or where it came from.
A modeling study that Nicol participated in predicted South African hospitals would see approximately 11,750 alcohol related admissions per week when the lockdown alcohol ban was lifted. The study is still undergoing peer review and is yet to be published.
Experts say alcohol does affect the trauma load of hospitals. But the minister’s claim that 34,000 hospital beds were occupied by alcohol-related cases is unsubstantiated. We rate it unproven.