September 15, 2026

expert reaction to study comparing respiratory complications after surgery in smokers vs vapers

Scientists respond to study published in the British Journal of Anaesthesia comparing respiratory complications after surgery in smokers vs vapers

Dr Harry Tattan-Birch, Senior Research Fellow in the Department of Behavioural Science and Health, University College London, said:

“The study does not measure what it claims to, and thus the results are not what they claim to be. Basically, the supposed “electronic cigarette” group was actually selected using a diagnosis intended for serious lung injury linked to cannabis vaping, rather than simply people who used nicotine e-cigarettes. The smoking group, by contrast, was just people who smoked.

“Unsurprisingly, the first group then had more breathing problems after surgery. That tells us nothing about whether nicotine e-cigarettes are more or less harmful than smoking, and it would be irresponsible to claim otherwise.

“The authors present this as a comparison of an e-cigarette use group with smokers, but the “electronic cigarette” group was identified using U07.0, a diagnostic code introduced during the “EVALI” outbreak. That outbreak was largely driven by cannabis vaping products, and the code explicitly includes dabbing-related (a type of cannabis use) as well as vaping-related lung injury. So, this code does not specifically identify nicotine e-cigarette users.

“More fundamentally, it does not identify people who ‘just happen to vape’. It identifies those already diagnosed with a potentially vaping/dabbing-related lung damage or injury. In effect, the study compares general cigarette smokers with a highly selected group of patients who are likely to have suffered serious lung injury, and then attributes their much higher rate of subsequent respiratory complications to e-cigarette use in general. The fact that postoperative respiratory complications diverged so rapidly between the groups should itself have been a warning sign that the “electronic cigarette” group may already have entered surgery with substantially greater respiratory illness.

“This study tells us nothing about the effect of e-cigarette vaping versus smoking on post-operative outcomes. Given the vast literature showing that e-cigarette use typically results in substantially lower exposure to many of the toxic products of combustion than cigarette smoking, one would not expect general e-cigarette use to be as harmful as smoking.”

Dr Adam Jacobs, Senior Director, Biostatistical Science, Premier Research, said:

“This paper examined the healthcare records of both smokers and vapers undergoing various types of surgery to examine the rate of post-operative respiratory complications in each group. They came up with the surprising finding that respiratory complications in the 30 days after surgery were over twice as common in vapers than smokers (10.7% vs 5.1%). Given everything else we know about the dangers of smoking, this does not seem biologically plausible. So what is the explanation?

“To try to ensure that the smokers and vapers were comparable with each other, the authors used a matched pair analysis in which smokers and vapers were individually matched using a propensity score derived from various covariates, to try to ensure that the 2 groups were comparable apart from their choice of nicotine delivery method. This is in principle a sensible way to analyse the data, as for example smokers tend to be older than vapers, and failing to take account of such covariates could have led to important bias. However, the danger of matching is that patients may be over-matched, which can eliminate differences between the groups which are legitimately present. I believe that the groups have indeed been over-matched in this case.

“While matching on baseline variables such as age, sex, socioeconomic status, and potentially other variables is sensible, what is problematic here is that patients were also matched on coexisting diseases. Coexisting diseases are not baseline variables, but are a form of outcome. It is likely that they are a consequence of smoking or vaping rather than something that just happens to be true at baseline. As would be expected, the burden of coexisting disease was considerably higher in the smoking group than in the vaping group. For example, 20% of vapers had hypertensive disease, compared with 44% of smokers, and 3% of vapers had chronic obstructive pulmonary disease, compared with 14% of smokers.

“Only patients who could be matched were included in the analysis. Although the original database contained over 1.6 million smokers, only 3817 smokers were included in the analysis. These smokers were absolutely not representative of the full sample, and were considerably healthier.

“While smoking is, on average, highly harmful to health, a few lucky smokers seem to be able to smoke without succumbing to the various diseases that commonly afflict smokers. Many smokers will tell you about their great aunt Edna who smoked 60 cigarettes every day from when she was 16 and lived to be 110.

“The problem here is that the sample of smokers analysed in this study was a collection of great aunt Ednas. It is a small and biased sample of all the smokers, containing unusually healthy smokers. It is therefore unsurprising that this group should have a low rate of post-operative respiratory complications, as this is a sample who have somehow beaten the odds and got away with smoking in other respects.

“A further consideration here is that patients were only included in the study if they had actually had surgery. As smoking is a known risk factor for poor outcomes, it is possible that some smokers who might otherwise have been good candidates for surgery did not have surgery because their smoking led them to be considered too high risk, so again, that would make the sample of smokers included in the study more than usually healthy.

“Overall, this study does not seem to me to be adequate to overturn the large body of evidence showing that vaping, while not completely safe, is considerably safer than smoking.”

‘Association of preoperative electronic cigarette vaping versus conventional cigarette smoking with postoperative respiratory complications: a multicentre matched cohort study’ by Hsiang-Ling Wu et al. was published in the British Journal of Anaesthesia at 00.01 UK time Tuesday 15 September 2026. 

DOI: https://doi.org/10.1016/j.bja.2026.08.002

Declared interests

Adam Jacobs: “No conflicts of interest to declare.”

Harry Tattan-Birch: “I have no links with the tobacco or nicotine industries.”