Brad Rodu responds to BMJ misinformation

A month after publication, the BMJ's deeply flawed article about smoking and COVID-19 has only three Rapid Responses below it, one of which is mine. This is puzzling because I know of two researchers who have submitted responses but not had them published. One of them is the main target of the hit piece, Dr Farsalinos. The other is Dr Brad Rodu. Brad has had enough of waiting so he has put his response on his blog.

 

Horel and Keyzer also failed to accurately cite and acknowledge studies that have shown a protective effect of current smoking on covid-19 (6,7,8).  They mentioned such a study from China but didn’t reference it, which is unfortunate because it was published in the New England Journal of Medicine (6).  They also mentioned and cited a study from France (7), but characterized it only as published quickly and as the origin of media hype. 

Horel and Keyzer completely omitted other studies, such as one involving veterans in the U.S., which found that smokers were much less likely to test positive for covid-19 than nonsmokers (AOR = 0.45, CI = 0.35 – 0.57) (8).  Another omission was a rapid evidence review of 28 observational studies finding that “Current data suggest that smokers in the community appear to be less likely to test positive for SARS-CoV-2 compared with never smokers,” “Across 405 studies, recorded current but not past smoking prevalence was generally lower than national prevalence estimates. Current smokers were at reduced risk of testing positive for SARS-CoV-2 and former smokers were at increased risk of hospitalisation, disease severity and mortality compared with never smokers.” (9)  However, one omission by Horel and Keyzer is especially inexplicable, because the study was published by this journal (BMJ) (10).  In that study smokers were significantly less likely than non-smokers to be diagnosed with COVID-19 and to be admitted to an intensive care unit (ICU), and the latter effect was dose-dependent.  The adjusted hazard ratios for ICU admission was 0.26 (CI = 0.19 – 0.37) for light smokers and 0.07 (CI = 0.01 – 0.47) for heavy smokers.   

In summary, there is substantial evidence that current smoking may be negatively associated with a covid-19 diagnosis and its subsequent course, including death. 

Read the whole thing.