back to top
Monday, 2 December, 2024
HomeResearch AfricaCritics say SA alcohol ban study 'lacks methodological detail'

Critics say SA alcohol ban study 'lacks methodological detail'

Is the acknowledged decline in hospital trauma admissions during the various lockdowns a result of the ban on alcohol, restrictions on people’s movement or a combination of both? Daily Maverick reports that an effort at research (conducted on behalf of the alcohol industry by a person with links to a lobby group that has been anti-lockdown) suggests the answer could lie somewhere in between.

Amended lockdown regulations over the Easter weekend saw the number of people who could attend religious and other gatherings more than double to 250 people indoors and 500 outdoors, and, while shebeens and restaurants remained open, the offsite sale of liquor was prohibited for four days.

DM reports that the apparent lack of a scientific basis for these decisions prompted the South African Liquor Brand Owners Association (SALBA) to release an investigative report it commissioned several months ago.

The investigation set out to interrogate the widely held view that there is a clear causal link between the three alcohol bans (excluding the recent Easter ban) and the visible and, at times, startling decline in hospital trauma admissions that occurred at the same time.

This decline has led to the widely held belief that the level of trauma admissions can be linked to the different levels of alcohol restrictions. The authors of the paper do not believe that this logic holds, however. They argue that this theory fails to address the fact that alcohol restrictions coincided with other restrictions that may also have had an impact on trauma admissions – not least of which was the level of lockdown.

In fact, the authors claim that the measured association between alcohol restrictions and trauma admissions in 2020 becomes statistically insignificant when an adjustment for the simultaneously imposed curfew is made.

DM reports that the data analysis was led by Ian McGorian of Silver Fox Consulting, in collaboration with Professor Mike Murray from the University of KwaZulu-Natal’s School of Mathematics, Statistics and Computer Science, with financial support from Distell.

McGorian is identified on the Pandemics Data & Analytics (Panda) website as a Panda team member. He does not disclose this link in his report.

DM says that it has reported in the past that at a time when reliable information on the COVID-19 pandemic, treatment and vaccines is considered so vital that many countries – including South Africa – have criminalised misinformation, a group of influential South African lobbyists led by Panda has been amplifying discredited and unscientific views and voices.

DM reports that Panda, a South African private sector research group that has repeatedly lobbied against lockdown denies being associated with a global coalition of pandemic denialists and sceptics – but online evidence suggests otherwise. Panda is one of the loudest anti-lockdown groups globally and openly states online that it hopes to “lead the world against lockdown”.

The authors of the anti-booze ban paper examined publicly available data sources, such as SA Police Service annual reports, data from the SA Medical Research Council, Google mobility data and the Sentinel Trauma Report which is available on the SAMRC website.

The analysis was reviewed by Professor Graham Barr from the department of statistical sciences at the University of Cape Town.

Key findings from their investigation included the fact that trauma cases in South Africa under lockdown dropped 60%. The paper states that other countries without an alcohol ban also experienced a similar phenomenon, including the UK (-57%), Ireland (-62%), Italy (-56.6%) and the US (-54%), raising questions about the efficacy of bans on alcohol.

The authors also noted that while alcohol is often cited as the leading factor in violent crime, the SAPS’ annual report of 2019-2020 shows that alcohol is officially confirmed in only a small percentage of cases: 5.4% of sexual offences, 5.3% of assaults, 6.7% of murders, 6.9% of attempted murders and 11.9% of assault with intent to cause grievous bodily harm.

SALBA chair Sibani Mngadi, who is also the corporate relations director for alcohol giant Diageo, said the industry was acutely aware of the realities and harms associated with the irresponsible consumption of alcohol by some consumers, as well as the pressure frontline healthcare personnel experienced during surges of COVID-19. “The analysis was not intended to dispute this, but rather open a constructive dialogue on facts, towards finding holistic solutions”, he said.

At the same time, he notes, the whole alcohol value chain, from retailers to glass manufacturers, had been damaged by the alcohol bans and the same would potentially be true of more stringent liquor regulations.

DM reports it is critical to have an objective analysis of the data being used to justify such actions, and any restrictions should be evidence-based.

The analysis showed the assumption that alcohol was a leading cause of trauma admissions was not supported by the data – using official, publicly available statistics for anyone to scrutinise.

DM reports that while SALBA plans to release further data and analysis in time, some experts contacted were sceptical.

“The link between alcohol and trauma is well established and we saw distinct drop-offs during the alcohol bans,” says the head of trauma at UKZN, Professor Timothy Hardcastle.

To compare the drop-off in South Africa to other countries without an alcohol ban is misleading, he adds. “These are countries where the policing is far stricter; where controls on the sale of alcohol are in place, or huge surcharges implemented on alcohol, making it less affordable.

Professor Charles Parry, director of the alcohol, tobacco and other drugs research unit at the SAMRC, questioned the methodology used. “This appears to be a report… not a published research paper. Its biggest flaw is it lacks the methodological detail needed to adequately assess the methodology applied.”

On the subject of causation, which requires that for A to cause B, a variety of criteria must be met, Parry notes that both instruments – the liquor sales bans and the restrictions on mobility – are causally related to trauma presentations.

“The fact that another intervention affects the outcome (trauma presentations) does not invalidate an intervention (the temporary liquor sales ban) from also causally impacting on that outcome,” he says.

While recognising that Mediclinic serves a private rather than a public market, its chief clinical officer, Dr Gerrit de Villiers, notes that following the withdrawal of the alcohol ban in February 2021, “we have seen a steady but not immediate increase in trauma-related emergency centre cases and admissions to our Mediclinic hospitals.

“The reintroduction of full trading hours for alcohol sales has not had any significant effect on trauma cases.”

Minimising transmission remains a key factor in managing the effects of the pandemic and everyone needs to play their part in preventing further spread, he adds.

DM reports that far from attempting to provoke an argument, Distell CEO Richard Rushton says the industry is simply asking that an objective view be taken of the data in order to facilitate more meaningful dialogue with decision-makers.

“We are all on the same side and we want to help find solutions. We are very clear that alcohol abuse is unacceptable and causes harm. Our view is that the focus must be on finding ways to deal with high-risk drinkers, rather than using blunt instruments that penalise all South Africans.
“Any proposed new regulations need to be evidence-based, rational and target problem areas.”

 

Chair of the SA Medical Association (SAMA), Dr Angelique Coetzee, said in a TimesLIVE report that the suggestion that the curfew rather than the alcohol ban reduced the number of trauma admissions “had merit”. “Restricting peoples’ movements means there is not a lot of time to go out and drink,” she said.

Coetzee noted that SAMA had asked for a curfew. “We didn't say ban alcohol,” she said. It was common sense that there is less alcohol-related trauma during the day than later at night, she said. “The more chance people have to drink the greater chance they may harm themselves,” she said.

But, TimesLIVE reports, Dr Suhayl Essa, a front-line doctor at a public hospital, supported the alcohol ban. While he agreed that the curfew was likely a significant factor in reducing trauma injuries, that, coupled with the alcohol ban, reduced the risk even further. “If you don't have alcohol and you're sitting at home, there's not much chance of hurting yourself,” he said.

Essa said his personal experience working in trauma units showed a predictable relationship between alcohol and admissions to trauma units. While Essa agreed that the evidence was anecdotal, “any doctor (would say) that alcohol has an affect on health care”, he said.

In June, as South Africa emerged from its grinding hard lockdown and trauma cases began rising again, Milpark Hospital trauma director Professor Ken Boffard noted that alcohol-related trauma cases accounted for 60% of all trauma-related admissions. During the hard lockdown, however, he recorded a 70% drop in trauma cases, he is quoted in TimesLIVE as saying.

The theory gained traction during the second lockdown when hospitals reported sharply reduced numbers of trauma patients.

On New Year’s Eve, the trauma unit at Baragwanath Hospital in Soweto had no patients in its trauma unit for the first time in its history, while in January, Professor Mervyn Mer, a principal specialist in critical care and pulmonology at the Charlotte Maxeke Academic Hospital in Johannesburg, is quoted in TimesLIVE as saying that alcohol restrictions had helped the hospital cope with the influx of patients, as trauma beds had been spared.

 

The research drew heavily on trauma admissions data from five Western Cape hospitals, but Parry said in a Sunday Times report, health officials told him they had not released the raw data. “Where did they get it from?” he said. “The logical conclusion is that they have imputed the data from screengrabs of graphs that have been released periodically.” That raised questions about the accuracy of the trauma admissions data.

Parry said the paper’s other flaws included having only one peer reviewer as well as a lack of clarity on whether the authors were paid for their work.

He also noted that much pre-existing research does document a causal link between alcohol restrictions and trauma admissions and other outcomes.
“If they want to claim that there is no causal link, they should also explain why this large body of international research is incorrect,” he said.

Professor Leslie London, chair of public health medicine at the University of Cape Town, said the research has no validity until it is properly peer reviewed. “The basic problem is there is no method in the paper,” he said. “They’ve released it into the public domain without any methods, and to me that’s unacceptable. It’s not science – it’s basically propaganda. And if you were serious about the science, you would release the reports, people could peer review the method.”

 

Full Daily Maverick report (Open access)

Full TimesLIVE report (Open access)

SAMRC Sentinel Trauma Report

Full Sunday Times report (Restricted access)

 

See also MedicalBrief archives:

Govt’s alcohol ban was not properly considered, argues SAB

Alcohol ban has been lifted but quest for ‘legal certainty’ continues

Constitutionality of alcohol ban challenged after government ‘fails to consider alternatives

Health versus economy – Trauma cases plummet under alcohol ban

Dlamini-Zuma promises regular re-evaluation of ’empirical evidence’ for alcohol ban

Getting the facts straight – The alcohol ban and COVID-19 beds

MRC modelling shows alcohol ban will cut trauma patients by 50,000

Lifting of alcohol ban strains provincial emergency departments

MedicalBrief — our free weekly e-newsletter

We'd appreciate as much information as possible, however only an email address is required.