- The government has disclosed the science on which it relied in deciding to ban tobacco products during South Africa’s coronavirus lockdown.
- In an affidavit, cooperative governance and traditional affairs minister Nkosazana Dlamini-Zuma appears to cite several studies showing a link between smoking and bad cases of Covid-19.
- She actually cites one set of statistics three different ways – plus some disagreement about what those numbers mean.
In papers filed with the high court in Pretoria this week, the South African government for the first time disclosed the scientific studies it says led to the decision to ban cigarette sales during the coronavirus lockdown.
In a 17-page affidavit in response to an action by the Fair Trade Independent Tobacco Association, cooperative governance and traditional affairs minister Nkosazana Dlamini-Zuma speaks at length about the harm cigarettes do generally, and the near-immediate benefits that quitting brings.
Thanks to the high prices and reduced access to illicit cigarettes, the temporary cigarette prohibition should see a lot of South Africans quit for good, especially young and poor people, said Dlamini-Zuma.
But there is also reason to suspect a direct link between severe Covid-19 and smoking she said – plus an increased risk of catching the disease in the first place for smokers.
In support of that, Dlamini-Zuma’s affidavit gives the appearance of citing four different studies that found a link between smoking and severe cases of Covid-19, plus another piece of research that proposes a mechanism by which smokers could be making themselves more likely to be infected.
In reality she effectively cites one study of some value three different ways. And that study did not deal with smoking, but only happened to include statistics others could use to reach a conclusion.
The additional piece of research on the mechanism (which Dlamini-Zuma does not present as a study) turns out to be a letter which, if anything, undermines her case.
This is what Dlamini-Zuma told a court about the science behind government’s cigarette ban – and what that research actually says.
Study 1: Smokers may be at more risk of bad cases of Covid-19 – but not everyone agrees.
On 28 February, well before South Africa’s lockdown, the New England Journal of Medicine published “Clinical characteristics of coronavirus disease 2019 in China”.
Dlamini-Zuma described it to the court like this:
“A large study of 1099 patients with Covid-19 found that, among the patients with severe symptoms, 16.9% were current smokers and 5.2% were former smokers, in contrast to patients with non-severe symptoms where 11.8% were current smokers and 1.3% were former smokers. In the group of patients that either needed mechanical ventilation, admission to an ICU or die, 25.5% were current smokers and 7.6% were former smokers.”
The study itself does not deal with smoking at all, but it does include a table which lists the outcomes for smokers, non-smokers, and former smokers with Covid-19. This was later used by others to reach conclusions about the link with smoking – but without clear consensus.
In correspondence that touches on the study – correspondence which Dlamini-Zuma also cites in her court papers (see “Not-quite-a-study” below) – another researcher points out that there were very few smokers in the study cohort compared to the prevalence of smoking in China.
The data shows “a trend towards association between smoking and severity of Covid-19”, said Hua Cai of the David Geffen School of Medicine at the University of California Los Angeles, in the letter, “but it was not significant.”
The link was taken more seriously in a review, however. (See “Study 3” below.)
Study 2: 5 smokers died of Covid-19
The second study Dlamini-Zuma cites, “Clinical course and risk factors for mortality of adult inpatients with COVID-19 in Wuhan, China: a retrospective cohort study” was published by The Lancet on 11 March.
She told the court that:
“Another study found that, among those who were infected with Covid-19 and died, 9% were current smokers.”
The study included data on 191 people, of which 11 were smokers. Five of those smokers died, while six survived.
The study itself does not deal with smoking or draw conclusions, but includes statistics on smokers in accompanying tables.
According to a review also cited by Dlamini-Zuma (See “Study 3, below), those numbers pointed to “no statistically significant difference between the smoking rates of survivors and non-survivors” of Covid-19.
Study 3 (and 4): Study 1 suggests smoking may be associated with bad Covid-19
Dlamini-Zuma’s phrasing makes it appear to be two different studies, one with original data and one review study.
She tells the court:
“Another study found that, among the patients with severe symptoms, 16.9% were current smokers and 5.2% were former smokers, while the history of smoking was a risk factor of disease progression.”
and then later says:
“A systematic review concluded that smokers suffer more severely from Covid-19, being 2.4 times more likely to be admitted to an ICU, need mechanical ventilation or die compared to non-smokers.”
Both those references point to the TID article, which reviews five previously published studies – including the two studies Dlamini-Zuma had cited already. Because the other four studies are of small groups, the review relies very heavily on Study 1. So heavily, in fact, that both the lines Dlamini-Zuma cites in her affidavit are based on Study 1 data.
The effect is that Dlamini-Zuma cites the same study three different times.
Study 3/4 is the only one that makes a case for a link between bad cases of Covid-19 and smoking, albeit guardedly.
“In conclusion, although further research is warranted as the weight of the evidence increases, with the limited available data, and although the above results are unadjusted for other factors that may impact disease progression, smoking is most likely associated with the negative progression and adverse outcomes of Covid-19,” the authors write.
Not-quite-a-study: Asian men’s smoking may make them more susceptible, but there isn’t enough data to be sure.
Dlamini-Zuma also cites in her affidavit a piece of correspondence published by The Lancet on 11 March, which appears to explain how smoking could directly make people more likely to catch SARS-CoV-2.
She tells the court:
“One possible reason [that smokers are more likely to end up in ICU] is that smokers express the receptor ACE2 (angiotensin II conversion enzyme-2 receptor). This is a receptor site for Covid-19 infections in the respiratory system. Because smokers have increased ACE2 receptors, there are potentially more ‘receptor sites’ for the virus to enter the body.”
But the writer, Hua Cai of the David Geffen School of Medicine at the University of California Los Angeles, actually makes a very different point.
New research, says the writer, suggests ACE2 expression (which could be linked to smoking) could explain why Asian men were apparently so much more likely to fall prey to Covid-19 than Asian women, and men in other ethnic groups.
But a study had found no difference in ACE2 expression between smokers and non-smokers who were white; only in smokers of Asian ethnicity did there appear to be a link.
“…the current literature does not support smoking as a predisposing factor in men or any subgroup for infection with SARS-CoV-2,” reads the letter.