Electronic cigarettes for smoking cessation
Abstract
Background
Electronic cigarettes (ECs) are handheld electronic vaping devices which produce an aerosol by heating an e‐liquid. Some people who smoke use ECs to stop or reduce smoking, although some organizations, advocacy groups and policymakers have discouraged this, citing lack of evidence of efficacy and safety. People who smoke, healthcare providers and regulators want to know if ECs can help people quit smoking, and if they are safe to use for this purpose. This is a review update conducted as part of a living systematic review.
Objectives
To examine the effectiveness, tolerability, and safety of using electronic cigarettes (ECs) to help people who smoke tobacco achieve long‐term smoking abstinence.
Search methods
We searched the Cochrane Tobacco Addiction Group's Specialized Register, the Cochrane Central Register of Controlled Trials (CENTRAL), MEDLINE, Embase, and PsycINFO to 1 July 2022, and reference‐checked and contacted study authors.
Selection criteria
We included randomized controlled trials (RCTs) and randomized cross‐over trials, in which people who smoke were randomized to an EC or control condition. We also included uncontrolled intervention studies in which all participants received an EC intervention. Studies had to report abstinence from cigarettes at six months or longer or data on safety markers at one week or longer, or both.
Data collection and analysis
We followed standard Cochrane methods for screening and data extraction. Our primary outcome measures were abstinence from smoking after at least six months follow‐up, adverse events (AEs), and serious adverse events (SAEs). Secondary outcomes included the proportion of people still using study product (EC or pharmacotherapy) at six or more months after randomization or starting EC use, changes in carbon monoxide (CO), blood pressure (BP), heart rate, arterial oxygen saturation, lung function, and levels of carcinogens or toxicants, or both. We used a fixed‐effect Mantel‐Haenszel model to calculate risk ratios (RRs) with a 95% confidence interval (CI) for dichotomous outcomes. For continuous outcomes, we calculated mean differences. Where appropriate, we pooled data in meta‐analyses.
Main results
We included 78 completed studies, representing 22,052 participants, of which 40 were RCTs. Seventeen of the 78 included studies were new to this review update. Of the included studies, we rated ten (all but one contributing to our main comparisons) at low risk of bias overall, 50 at high risk overall (including all non‐randomized studies), and the remainder at unclear risk.
There was high certainty that quit rates were higher in people randomized to nicotine EC than in those randomized to nicotine replacement therapy (NRT) (RR 1.63, 95% CI 1.30 to 2.04; I2 = 10%; 6 studies, 2378 participants). In absolute terms, this might translate to an additional four quitters per 100 (95% CI 2 to 6). There was moderate‐certainty evidence (limited by imprecision) that the rate of occurrence of AEs was similar between groups (RR 1.02, 95% CI 0.88 to 1.19; I2 = 0%; 4 studies, 1702 participants). SAEs were rare, but there was insufficient evidence to determine whether rates differed between groups due to very serious imprecision (RR 1.12, 95% CI 0.82 to 1.52; I2 = 34%; 5 studies, 2411 participants).
There was moderate‐certainty evidence, limited by imprecision, that quit rates were higher in people randomized to nicotine EC than to non‐nicotine EC (RR 1.94, 95% CI 1.21 to 3.13; I2 = 0%; 5 studies, 1447 participants). In absolute terms, this might lead to an additional seven quitters per 100 (95% CI 2 to 16). There was moderate‐certainty evidence of no difference in the rate of AEs between these groups (RR 1.01, 95% CI 0.91 to 1.11; I2 = 0%; 5 studies, 1840 participants). There was insufficient evidence to determine whether rates of SAEs differed between groups, due to very serious imprecision (RR 1.00, 95% CI 0.56 to 1.79; I2 = 0%; 8 studies, 1272 participants).
Compared to behavioural support only/no support, quit rates were higher for participants randomized to nicotine EC (RR 2.66, 95% CI 1.52 to 4.65; I2 = 0%; 7 studies, 3126 participants). In absolute terms, this represents an additional two quitters per 100 (95% CI 1 to 3). However, this finding was of very low certainty, due to issues with imprecision and risk of bias. There was some evidence that (non‐serious) AEs were more common in people randomized to nicotine EC (RR 1.22, 95% CI 1.12 to 1.32; I2 = 41%, low certainty; 4 studies, 765 participants) and, again, insufficient evidence to determine whether rates of SAEs differed between groups (RR 1.03, 95% CI 0.54 to 1.97; I2 = 38%; 9 studies, 1993 participants).
Data from non‐randomized studies were consistent with RCT data. The most commonly reported AEs were throat/mouth irritation, headache, cough, and nausea, which tended to dissipate with continued EC use. Very few studies reported data on other outcomes or comparisons, hence evidence for these is limited, with CIs often encompassing clinically significant harm and benefit.
Authors' conclusions
There is high‐certainty evidence that ECs with nicotine increase quit rates compared to NRT and moderate‐certainty evidence that they increase quit rates compared to ECs without nicotine. Evidence comparing nicotine EC with usual care/no treatment also suggests benefit, but is less certain. More studies are needed to confirm the effect size. Confidence intervals were for the most part wide for data on AEs, SAEs and other safety markers, with no difference in AEs between nicotine and non‐nicotine ECs nor between nicotine ECs and NRT. Overall incidence of SAEs was low across all study arms. We did not detect evidence of serious harm from nicotine EC, but longest follow‐up was two years and the number of studies was small.
The main limitation of the evidence base remains imprecision due to the small number of RCTs, often with low event rates, but further RCTs are underway. To ensure the review continues to provide up‐to‐date information to decision‐makers, this review is a living systematic review. We run searches monthly, with the review updated when relevant new evidence becomes available. Please refer to the Cochrane Database of Systematic Reviews for the review's current status.
PICOs
Plain language summary
Can electronic cigarettes help people stop smoking, and do they have any unwanted effects when used for this purpose?
What are electronic cigarettes?
Electronic cigarettes (e‐cigarettes) are handheld devices that work by heating a liquid that usually contains nicotine and flavourings. E‐cigarettes allow you to inhale nicotine in a vapour rather than smoke. Because they do not burn tobacco, e‐cigarettes do not expose users to the same levels of chemicals that can cause diseases in people who smoke conventional cigarettes.
Using an e‐cigarette is commonly known as 'vaping'. Many people use e‐cigarettes to help them to stop smoking tobacco. In this review we focus primarily on e‐cigarettes containing nicotine.
Why we did this Cochrane Review
Stopping smoking lowers your risk of lung cancer, heart attacks and many other diseases. Many people find it difficult to stop smoking. We wanted to find out if using e‐cigarettes could help people to stop smoking, and if people using them for this purpose experience any unwanted effects.
What did we do?
We searched for studies that looked at the use of e‐cigarettes to help people stop smoking.
We looked for randomized controlled trials, in which the treatments people received were decided at random. This type of study usually gives the most reliable evidence about the effects of a treatment. We also looked for studies in which everyone received an e‐cigarette treatment.
We were interested in finding out:
· how many people stopped smoking for at least six months; and
· how many people had unwanted effects, reported on after at least one week of use.
Search date: We included evidence published up to 1st July 2022.
What we found
We found 78 studies which included 22,052 adults who smoked. The studies compared e‐cigarettes with:
· nicotine replacement therapy, such as patches or gum;
· varenicline (a medicine to help people stop smoking);
· e‐cigarettes without nicotine;
· other types of nicotine‐containing e‐cigarettes (e.g. pod devices, newer devices);
· behavioural support, such as advice or counselling; or
· no support for stopping smoking.
Most studies took place in the USA (34 studies), the UK (16), and Italy (8).
What are the results of our review?
People are more likely to stop smoking for at least six months using nicotine e‐cigarettes than using nicotine replacement therapy (6 studies, 2378 people), or e‐cigarettes without nicotine (5 studies, 1447 people).
Nicotine e‐cigarettes may help more people to stop smoking than no support or behavioural support only (7 studies, 3126 people).
For every 100 people using nicotine e‐cigarettes to stop smoking, 8 to 12 might successfully stop, compared with only 6 of 100 people using nicotine‐replacement therapy, 7 of 100 using e‐cigarettes without nicotine, or 4 of 100 people having no support or behavioural support only.
We are uncertain if there is a difference between how many unwanted effects occur using nicotine e‐cigarettes compared with nicotine replacement therapy, no support or behavioural support only. There was some evidence that non‐serious unwanted effects were more common in groups receiving nicotine e‐cigarettes compared to no support or behavioural support only. Low numbers of unwanted effects, including serious unwanted effects, were reported in studies comparing nicotine e‐cigarettes to nicotine replacement therapy. There is probably no difference in how many non‐serious unwanted effects occur in people using nicotine e‐cigarettes compared to e‐cigarettes without nicotine.
The unwanted effects reported most often with nicotine e‐cigarettes were throat or mouth irritation, headache, cough and feeling sick. These effects reduced over time as people continued using nicotine e‐cigarettes.
How reliable are these results?
Our results are based on a few studies for most outcomes, and for some outcomes, the data varied widely.
We found evidence that nicotine e‐cigarettes help more people to stop smoking than nicotine replacement therapy. Nicotine e‐cigarettes probably help more people to stop smoking than e‐cigarettes without nicotine but more studies are still needed to confirm this.
Studies comparing nicotine e‐cigarettes with behavioural or no support also showed higher quit rates in people using nicotine e‐cigarettes, but provide less certain data because of issues with study design.
Most of our results for the unwanted effects could change when more evidence becomes available.
Key messages
Nicotine e‐cigarettes can help people to stop smoking for at least six months. Evidence shows they work better than nicotine replacement therapy, and probably better than e‐cigarettes without nicotine.
They may work better than no support, or behavioural support alone, and they may not be associated with serious unwanted effects.
However, we still need more evidence, particularly about the effects of newer types of e‐cigarettes that have better nicotine delivery than older types of e‐cigarettes, as better nicotine delivery might help more people quit smoking.
Authors' conclusions
Summary of findings
Nicotine EC compared to NRT for smoking cessation | ||||||
Patient or population: People who smoke | ||||||
Outcomes | Anticipated absolute effects* (95% CI) | Relative effect | № of participants | Certainty of the evidence | Comments | |
---|---|---|---|---|---|---|
Risk with NRT | Risk with Nicotine EC | |||||
Smoking cessation at 6 months to 1 year Assessed with biochemical validation | Study population | RR 1.63 | 2378 | ⊕⊕⊕⊕ | ‐ | |
6 per 100 | 10 per 100 | |||||
Adverse events at 4 weeks to 6‐9 months Assessed by self‐report | Study population | RR 1.02 | 1702 | ⊕⊕⊕⊝ | ‐ | |
27 per 100 | 27 per 100 | |||||
Serious adverse events at 4 weeks to 1 year Assessed via self‐report and medical records | Study population | RR 1.12 | 2411 | ⊕⊕⊝⊝ | 2 studies reported no events; effect estimate based on the three studies in which events were reported | |
6 per 100 | 7 per 100 | |||||
*The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). For cessation, the assumed risk in the control group is based on assumed quit rates for NRT assuming receipt of limited behavioural stop‐smoking support (as per Hartmann‐Boyce 2018a). The assumed risk for adverse events and serious adverse events is a weighted mean average of quit rates across control groups in contributing studies. | ||||||
GRADE Working Group grades of evidence | ||||||
aDowngraded one level due to imprecision; CIs consistent with benefit and harm |
Nicotine EC compared to non‐nicotine EC for smoking cessation | ||||||
Patient or population: People who smoke cigarettes | ||||||
Outcomes | Anticipated absolute effects* (95% CI) | Relative effect | № of participants | Certainty of the evidence | Comments | |
---|---|---|---|---|---|---|
Risk with non‐nicotine EC | Risk with Nicotine EC | |||||
Smoking cessation at 6‐12 months Assessed with biochemical validation | Study population | RR 1.94 | 1447 | ⊕⊕⊕⊝ | ‐ | |
7 per 100 | 14 per 100 | |||||
Adverse events at 1 week to 6 months Assessed via self‐report | Study population | RR 1.01 | 840 | ⊕⊕⊕⊝ | ‐ | |
9 per 100 | 9 per 100 | |||||
Serious adverse events at 1 week to 1 year Assessed via self‐report and medical records | Study population | RR 1.00 | 1272 | ⊕⊕⊝⊝ | 4 studies reported no events; effect estimate based on the 3 studies in which events were reported | |
3 per 100 | 3 per 100 | |||||
*The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). For cessation, the assumed risk in the control group is based on receipt of moderate‐intensity behavioural stop‐smoking support. The assumed risk for adverse events and serious adverse events is a weighted mean average of quit rates across control groups in contributing studies. | ||||||
GRADE Working Group grades of evidence | ||||||
aNot downgraded for risk of bias. One of four studies considered high risk of bias; removing this study increased the direction of the effect in favour of the intervention. |
Nicotine EC compared to behavioural support only/no support for smoking cessation | ||||||
Patient or population: People who smoke | ||||||
Outcomes | Anticipated absolute effects* (95% CI) | Relative effect | № of participants | Certainty of the evidence | Comments | |
---|---|---|---|---|---|---|
Risk with behavioural support only/no support | Risk with Nicotine EC | |||||
Smoking cessation at 6 to 12 months Assessed using biochemical validation | Study population | RR 2.66 | 3126 | ⊕⊝⊝⊝ | ‐ | |
1 per 100 | 3 per 100 | |||||
Adverse events at 12 weeks to 6 months Assessed via self‐report | Study population | RR 1.22 | 765 | ⊕⊕⊝⊝ | ‐ | |
66 per 100 | 80 per 100 | |||||
Serious adverse events at 4 weeks to 8 months Assessed via self‐report and medical records | Study population | RR 1.03 | 1993 | ⊕⊝⊝⊝ | 5 of the 9 studies reported no SAEs; MA is based on pooled results from 4 studies. | |
2 per 100 | 2 per 100 | |||||
*The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). For cessation, the assumed risk in the control group is based on receipt of limited stop‐smoking support. The assumed risk for adverse events and serious adverse events is a weighted mean average of quit rates across control groups in contributing studies. | ||||||
GRADE Working Group grades of evidence | ||||||
aDowngraded two levels due to risk of bias. Due to lack of blinding and differential support between arms, judged to be at high risk of bias. |
Background
Throughout this review, we discuss (1) conventional cigarettes and (2) electronic cigarettes, defined as hand held and producing for inhalation an aerosol formed by heating an e‐liquid using a battery‐powered heating coil. In this review, all mention of smoking, smoking cessation, cigarette use, smoke intake, etc. concerns combustible tobacco cigarettes. When the text concerns electronic cigarettes we use the abbreviation 'ECs'. EC users are sometimes described as 'vapers', and EC use as 'vaping'. We refer to ECs that do not contain nicotine as non‐nicotine ECs; these can also be conceptualized as placebo ECs, but we are using the term non‐nicotine EC, as they can be conceptualized as an intervention in themselves. This review does not address the use of vaping devices to inhale substances other than nicotine, such as cannabis.
Description of the condition
Stopping smoking is associated with large health benefits. Despite most people who smoke wanting to quit, many find it difficult to succeed in the long term. Almost half who try to quit without support will not manage to stop for even a week, and fewer than five per cent remain abstinent at one year after quitting (Hughes 2004).
Behavioural support and medications such as nicotine patches or gum increase the chances of quitting through providing nicotine to help alleviate withdrawal symptoms, but even with this additional support, long‐term quit rates remain low (Cahill 2016; Hartmann‐Boyce 2018b; Hartmann‐Boyce 2019). One of the limitations of current treatments is that, apart from providing nicotine more slowly and at lower levels than smoking, none adequately addresses the sensory, behavioural and/or social aspects of smoking that ex‐smokers miss when they stop smoking (e.g. holding a cigarette in their hands, taking a puff, enjoyment of smoking, feeling part of a group). ECs may offer a way to overcome these limitations (Notley 2018b).
There is no doubt that people become dependent on tobacco, and find it difficult to stop smoking, primarily because of nicotine and its actions on the brain's reward system (Balfour 2004). However, developing dependence on tobacco smoking is a complex biopsychosocial process (Benowitz 2010; Rose 2006). Other tobacco chemicals, such as acetaldehyde and MAO inhibitors seem to potentiate effects of nicotine (Rose 2006). In addition, sensory and behavioural cues provide additional reinforcement of smoking behaviour (Rose 1993; Rose 2000) and may over time become almost as rewarding as nicotine. There are several lines of evidence to support this. Firstly, people who smoke appear to have a preference for cigarette smoke compared to other forms of nicotine delivery. This is partly related to the speed of nicotine delivery through smoke inhalation. However, even when nicotine is administered intravenously it does not provide the same level of satisfaction or reward as smoking (Rose 2000; Westman 1996). Secondly, the local sensory effects of smoking (e.g. the ‘scratch’ in the back of the throat) may be important for enjoyment and reward. Numbing the sensations of cigarette smoke by anaesthetizing the upper and lower respiratory tract leads to less enjoyment of smoking (Rose 1985). Conversely, products that mimic the sensory effects of smoking on the mouth and throat (such as citric acid, black pepper, and ascorbic acid) reduce craving and some withdrawal symptoms, at least in the short term (Levin 1993; Rose 1994; Westman 1995). Thirdly, very low nicotine content cigarettes (VLNCs), which have a very low content of nicotine (e.g. 0.08 mg instead of the normal 1 mg) and so have negligible or no central effects, have also been investigated for their role in aiding smoking cessation (Przulj 2013). Despite delivering low levels of nicotine, VLNCs are satisfying over the initial few days of abstinence from nicotine (Donny 2007; Donny 2015; Pickworth 1999; Rose 2000). They also reduce tobacco withdrawal symptoms, including urges to smoke and low mood (Barrett 2010; Donny 2009; McRobbie 2016; Perkins 2010; Rose 2000), and have been shown to improve long‐term continuous abstinence rates in one study (Walker 2012). Social aspects of smoking, such as feeling part of a like‐minded group, or including smoking behaviour as part of one's social identity are also elements of cigarette smoking that some people who smoke report to be drivers of cigarette use (Notley 2018a).
Considering the other factors that contribute to tobacco dependence, there is interest in developing smoking‐cessation products that would not only help relieve the unpleasant effects of nicotine withdrawal but would also act as an effective substitute for smoking behaviour and the rituals and sensations that accompany smoking, without the health risks associated with the inhalation of tobacco smoke. Until recently, the only pharmaceutical treatment available that had some of these characteristics was the nicotine inhalator. However, these do not have greater cessation efficacy than the other nicotine replacement therapy (NRT) products (Hajek 1999; Hartmann‐Boyce 2018a). This may in part be due to the considerable effort (e.g. 20 minutes of continuous puffing) needed to provide nicotine blood concentrations consistent with other NRTs (Schneider 2001). Adherence to correct use of the inhalator is low compared to other NRTs (Hajek 1999). It is therefore possible that any advantage of sensorimotor replacement is diminished by low nicotine delivery and limited similarities between inhalator use and sensations of smoking (Bullen 2010). A nicotine inhaler using pressurized air is approved as a smoking cessation aid in the UK. The nicotine delivery from this device is substantially lower than from cigarettes, and also lower than from the nicotine inhalator (Romeu 2020).
Description of the intervention
ECs are hand held and produce for inhalation an aerosol formed by heating an e‐liquid using a battery‐powered heating coil (E‐cigarette ontology 2021). The e‐liquid, usually comprising propylene glycol and glycerol, with or without nicotine and flavours, is stored in disposable or refillable cartridges or a reservoir or 'pod'. The commonly‐used term for this aerosol is vapour, which we use throughout the review. ECs are marketed as consumer products. Although routes are in place for licensing them as medicine or medical devices in some areas, no country yet has a licensed medicinal EC.
ECs provide sensations similar to smoking a cigarette. They provide taste and throat sensations that are closer to smoking than those provided by the nicotine inhalator (Barbeau 2013). The vapour that looks like tobacco smoke is only visible when the user exhales after drawing on the mouthpiece, not when the device is being held. In qualitative studies, users report a sense of shared identity with other users, similar to tobacco‐smoking identity, and also report pleasure and enjoyment of use, suggesting that ECs may be viewed less as medical cessation aids but rather as acceptable alternatives to tobacco smoking (Cox 2017; Notley 2018a).
There are many different brands and models of EC available. Variation exists both in the device ('product') and consumable (e‐liquid used). There is a wide variation in the composition of e‐liquids (nicotine content, flavours and other components) (Goniewicz 2012; Goniewicz 2014), with some users choosing to mix their own e‐liquids (Cox 2019b). Initial studies showed that early models of EC delivered very low amounts of nicotine to naïve users (Bullen 2010; Eissenberg 2010; Vansickel 2010). Later studies that have measured nicotine pharmacokinetics in both experienced and naïve EC users have found that some EC users can achieve blood nicotine levels similar to those achieved with smoking, albeit more slowly, and that their ability to do so often improves over time (Hajek 2015b; Vansickel 2012; Vansickel 2013; Yingst 2019a; Yingst 2019b).
Early on in their development, ECs looked like cigarettes and used disposable cartridges. These models were often called 'cig‐a‐likes'. The nicotine delivery from these products was low, and even the modern versions of EC devices that use pre‐filled cartridges, generally produced by the tobacco industry, for the most part have only low nicotine delivery (Hajek 2017). The later refillable, or 'tank', products have a larger battery and a transparent container that users fill with an e‐liquid of their choice, and usually provide faster and more efficient nicotine delivery, allow a wider choice of flavours and nicotine concentrations, and are typically used by experienced vapers who manage to switch to vaping completely (ASH 2019; Dawkins 2013b; Farsalinos 2014). Observational evidence suggests people who smoke are more likely to successfully quit using tank models than with cig‐a‐likes (Chen 2016; Hitchman 2015). Smaller 'pod' devices that use nicotine salt are also available (e.g. Juul). This nicotine formulation reduces irritant effects and allows the delivery of higher nicotine levels that closely mimic the pharmacokinetic profile of nicotine delivery from cigarettes, despite the low battery power of the devices (Hajek 2020). The EU Tobacco Products Directive (European Parliament 2014) does not allow sales of e‐liquids with nicotine content higher than 20 mg/mL, and so the US version of Juul (59 mg/nl nicotine) is not available within the EU (Huang 2019; Talih 2020). Most recently, there has been rapid growth in the use of small disposable devices (Tattan‐Birch 2022). These are available in a range of attractive flavours, generally have a high nicotine content, are low cost and have a closed system that is designed to be disposed of following use (approximately 200 puffs). According to ASH 2022, for adults in GB, tank style devices are the most popular. For youth, the ASH 2022 report disposables are now the most popular.
The different device types may differ significantly in their efficacy in helping people who smoke to quit, as they differ in delivery of nicotine. Nicotine itself, when delivered through mechanisms and doses similar to that delivered in traditional NRT, is not considered harmful (Hartmann‐Boyce 2018a). The safety profile of the different types of nicotine EC may be similar as they use the same constituents, although within the generic range of EC types, there is some evidence to suggest EC providing less nicotine may pose higher risks. This is because low‐nicotine delivery devices need to be puffed with higher intensity to provide users with the nicotine levels that they seek, and more intensive puffing is accompanied by increased inhalation of potential toxicants (Dawkins 2016; Dawkins 2018; Smets 2019). Throughout this review, we refer to a nicotine‐containing EC as ‘nicotine EC’ and to nicotine‐free EC as 'non‐nicotine EC', which can also be considered 'placebo EC'. The 'placebo' comparison is a test just of the nicotine effect and not of the potential sensorimotor or behavioural and social replacement that the EC may provide.
There is no one agreed classification system for EC devices, and product development has moved so quickly that the definitions used within trials of the devices tested may no longer necessarily be fit for purpose. In this review, the definitions used are based on those drawn from the included trials. We currently label three different types of EC as 'cartridges' for devices with disposable cartridges and ‐ typically, but not always ‐ low nicotine delivery (e.g. cig‐a‐likes); refillable ECs for devices that vapers fill with their own choice of e‐liquids; and pods for the small devices that commonly use nicotine salts. To date, there are no trials of disposable devices, so we do not include this category in the current review. We may review this categorization system in future versions of the review as new trials and devices emerge.
Why it is important to do this review
Since ECs appeared on the market in 2006, there has been a steady increase in their use. In the UK, the ASH 2022 surveys found 19.4% of the adult population had ever tried vaping, but only 8.3% were current vapers. EC use is most prevalent in current (22%) and former (14%) smokers (ASH 2022). Only 1.3% of never‐smokers report currently using ECs. Prevalence data from the USA in 2019 showed that 4.4% of adults were current EC users (Du 2020). Data from lower‐income countries suggest similar levels of EC use and awareness (Besaratinia 2019; Jiang 2016; Palipudi 2016).
Regulatory approaches being used for ECs currently vary widely, from no regulation to partial and complete bans (McNeill 2022). Within the USA, for example, the Food and Drug Administration (FDA) has classified EC as tobacco products and laws include prohibition of EC use indoors, requirement for retailers to have a license to sell, and prohibition of sales to minors. Laws prohibiting sales to minors apply nationwide, but other laws vary by state (Du 2020). The European Union includes ECs in their Tobacco Products Directive, except where therapeutic claims are made or in instances where they contain over 20 mg/nl of nicotine, when they will require medicines authorization (European Parliament 2014).
Categorical statements about the toxicity of ECs are not possible because of the large number of devices and liquids available and the frequent addition of new products to the market. In 2019, cases of severe lung injury associated with EC use were reported in the USA and, by February 2020, there were around 2800 hospitalized cases or deaths (CDC 2020). This illness was termed E‐cigarette or Vaping‐Associated Lung Injury (EVALI) and caused concern throughout the world (Hall 2020), and a negative change in people's perception of the risks of EC use compared to smoking (Tattan‐Birch 2020). These cases were somewhat at odds with data from trials and cohort studies, and it was later found that these injuries were related to use of tetrahydrocannabinol (THC)‐containing products adulterated with vitamin E acetate (Blount 2020; Hartnett 2020). Amongst those brands of nicotine EC that have been tested, levels of toxins have been found to be substantially lower than in cigarettes (Hajek 2014; McNeill 2022). Long‐term effects beyond 12 months are unclear, although based on what is known about liquid and vapour constituents and patterns of use, a report from the UK's Royal College of Physicians has concluded that using an EC is likely to be considerably safer than smoking (RCP 2016). The US National Academies of Sciences, Engineering, and Medicine (NASEM) concluded that ECs are likely to be far less harmful than continuing to smoke cigarettes, with the caveat that the long‐term health effects of e‐cigarette use are not yet known (NASEM 2018).
Despite general acknowledgement that EC use exposes the user to fewer toxicants and at lower levels than smoking cigarettes (McNeill 2021; McNeill 2022; NASEM 2018; RCP 2016), there remains some hesitancy in making these products available to people who smoke as a harm‐reduction tool or smoking‐cessation aid (e.g. McDonald 2020). Lack of quality control measures, possible harms of second‐hand EC vapour inhalation, concerns that the products may be a gateway to smoking initiation or nicotine dependence among nicotine‐naïve users or may prolong continued dual use of tobacco amongst cigarette smokers, concerns that ECs may undermine smoke‐free legislation if used in smoke‐free spaces, concerns about the involvement of the tobacco industry, and concerns that the long‐term effects of EC use on health are not yet known are often cited (McNeill 2022). A report from the US Preventive Services Taskforce concluded "that the current evidence is insufficient to assess the balance of benefits and harms of electronic cigarettes (e‐cigarettes) for tobacco cessation in adults" (USPFTS 2021). However, others suggest that potential benefits outweigh potential disadvantages (Farsalinos 2014; Hajek 2014; McNeill 2021; McNeill 2022; NASEM 2018; RCP 2016).
People who smoke, healthcare providers and regulators are interested to know if ECs can help smokers quit and if it is safe to use them to do so. In particular, healthcare providers have an urgent need to know what they should recommend to people to help them to stop smoking. The largest health gains are achieved from stopping smoking completely, as opposed to reducing cigarette consumption and, as such, this review focuses on the effectiveness of ECs in aiding complete smoking cessation.
This review was first published in 2014, and updated in 2016, 2020, 2021 and 2022.
Following the publication of the 2020 update of this review, we are maintaining it as a living systematic review (Brooker 2019). This means we are continually running searches and incorporating new evidence into the review. For more information about the living systematic review methods being used, see Appendix 1. A living systematic review approach is appropriate for this review, for three reasons. First, the review addresses an important public health issue: the role of ECs in enabling people who smoke to stop smoking, with potential for substantial ongoing individual and societal benefits, if effective. Secondly, there remains uncertainty in the existing evidence; more studies are needed to confirm the degree of benefit for different comparisons and product types, and there is considerable uncertainty about adverse events and other markers of safety. Thirdly, we are aware of multiple ongoing trials on this topic that are likely to have an important impact on the conclusions of the review.
Objectives
To examine the safety, tolerability and effectiveness of using electronic cigarettes (ECs) to help people who smoke tobacco achieve long‐term smoking abstinence.
Methods
Criteria for considering studies for this review
Types of studies
We include randomized controlled trials (RCTs) and randomized cross‐over trials in which people who smoke are randomized to ECs or to a control condition. RCTs are the best available primary evidence, but the continued paucity of RCTs in this area requires that we also include uncontrolled intervention studies in which all participants are given an EC intervention.
We include studies regardless of their publication status or language of publication.
Types of participants
People defined as currently smoking cigarettes at enrolment into the studies. Participants could be motivated or unmotivated to quit.
Types of interventions
Any type of EC or intervention intended to promote EC use for smoking cessation, including studies which did not measure smoking cessation but provided ECs with the instruction they be used as a complete substitute for cigarette use. ECs may or may not contain nicotine.
Types of comparators
We compare nicotine ECs with non‐nicotine ECs, ECs versus alternative smoking cessation aids, including NRT or no intervention, and ECs added to standard smoking cessation treatment (behavioural or pharmacological or both) with standard treatment alone.
Types of outcome measures
Primary outcomes
-
Cessation at the longest follow‐up point, at least six months from the start of the intervention, measured on an intention‐to‐treat basis using the strictest definition of abstinence, preferring biochemically‐validated results where reported
-
Number of participants reporting adverse events or serious adverse events at one week or longer (as defined by study authors)
Secondary outcomes
Number of people still using study product (EC or pharmacotherapy) at longest follow‐up (at least six months). Product could be that provided by the study, or could be the same product type but bought independently by the participant.
Changes in the following measures at longest follow‐up (one week or longer):
-
Carbon monoxide (CO), as measured through breath or blood
-
Blood pressure
-
Heart rate
-
Blood oxygen saturation
-
Lung function measures
-
Known toxins/carcinogens, as measured through blood or urine (toxicant names and abbreviations are listed in Appendix 2)
Studies had to report one of the primary or secondary outcomes above to be eligible for inclusion.
Search methods for identification of studies
Electronic searches
Searches are conducted monthly. This update includes results from searches conducted up to 1st July 2022:
-
Cochrane Tobacco Addiction Group Specialized Register (CRS‐Web)
-
Cochrane Central Register of Controlled Trials (CENTRAL 2022; Issue 6) via CRS‐Web
-
MEDLINE (OVID SP; 1st January 2004 to 1st July 2022)
-
Embase (OVID SP; 1st January 2004 to 1st July 2022)
-
PsycINFO (OVID SP; 1st January 2004 to 1st July 2022)
-
ClinicalTrials.gov (via CENTRAL 2022; Issue 6)
-
WHO International Clinical Trials Registry Platform (ICTRP: www.who.int/ictrp/en/, via CENTRAL 2022; Issue 6)
At the time of the search, the Register included the results of searches of MEDLINE (via OVID) to update 20220614; Embase (via OVID) to week 202224; PsycINFO (via OVID) to update 20220613. See the Tobacco Addiction Group website for full search strategies and a list of other resources searched.
For the first version of the review, we also searched CINAHL (EBSCO Host) (2004 to July 2014). We did not search this database from 2016 onwards, as it did not contribute additional search results to the first version of the review. The search terms were broad and included 'e‐cig$' OR 'elect$ cigar$' OR 'electronic nicotine'. The search for the 2016 update added the terms 'vape' or 'vaper' or 'vapers' or 'vaping'. The 2020 searches added further terms, including the MESH heading 'Electronic Nicotine Delivery Systems' and terms to limit by study design. All current search strategies are listed in Appendix 3. The previously‐used search strategy is shown in Appendix 4. The search date parameters of the original searches were limited to 2004 to the present, due to the fact that ECs were not available before 2004.
Searching other resources
We searched the reference lists of eligible studies found in the literature search and contacted authors of known trials and other published EC studies. We also searched abstracts from the Society for Research on Nicotine and Tobacco (SRNT) Annual Meetings.
Data collection and analysis
Selection of studies
Two review authors (for this update from: ARB, JHB, NL, AT) independently prescreened all titles and abstracts obtained from the search, using a screening checklist, and then independently screened full‐text versions of the potentially relevant papers for inclusion. We resolved any disagreements by discussion or with a third review author.
Data extraction and management
Two reviewers (for this update from: ARB, AT, CN, PB) extracted data from the included studies using a pre‐piloted data extraction form, and checked them against each other. We resolved any disagreements by discussion or with a third review author. We extracted data on:
-
Author
-
Date and place of publication
-
Study dates
-
Study design
-
Inclusion and exclusion criteria
-
Setting
-
Summary of study participant characteristics
-
Summary of intervention and control conditions
-
Number of participants in each arm
-
Smoking cessation outcomes
-
Type of biochemical validation (if any)
-
Adverse events (AEs), serious adverse events (SAEs), number of people still using study product, and relevant biomarkers
-
Continued EC use or pharmaceutical intervention (PI) use at longest follow‐up
-
Assessment time points
-
Study funding source
-
Author declarations of interest
-
Risk of bias in the domains specified below
-
Additional comments
We adopted a broad focus to detect a variety of adverse events.
One review author (JHB) then entered the data into Review Manager 2020 software for analyses, and another checked them (NL for this update).
Assessment of risk of bias in included studies
Two review authors (for this update from: ARB, AT, CN, PB) independently assessed the risks of bias for each included study, using the Cochrane risk of bias tool v1 (Higgins 2011). This approach uses a domain‐based evaluation that addresses seven different areas: random sequence generation; allocation concealment; blinding of participants and providers; blinding of outcome assessment; incomplete outcome data; selective outcome reporting; and other potential sources of bias. We assigned a grade (low, high, or unclear) for risk of bias for each domain. We resolved disagreements by discussion or by consulting a third review author.
Specific considerations about judgements for individual domains in this review are outlined below:
-
Random sequence generation/allocation concealment: We rated all non‐randomized studies at high risk in these domains;
-
Blinding of participants and personnel: We did not evaluate this domain for non‐randomized studies, as we considered it not to be applicable. For randomized studies which did not use blinding, we considered studies to be at low risk in this domain if the intervention was compared to an active control of similar intensity, as we judged performance bias to be unlikely in this circumstance. If studies were unblinded and the comparator group was a minimal‐intervention control or of lower intensity than the intervention group, we considered the study to be at high risk of bias in this domain;
-
Following standard methods of the Cochrane Tobacco Addiction Review Group, we considered studies to be at low risk of detection bias (blinding of outcome assessment) if our primary outcome was objectively measured or if the intensity of the intervention was similar between groups, or both. For studies where cessation was measured, our judgement was based on whether cessation was biochemically verified. Where cessation was not measured, we judged this domain based on adverse or serious adverse events;
-
Again following standard methods of the Cochrane Tobacco Addiction Group, we rated studies at high risk of attrition bias if loss to follow‐up was greater than 50% overall or if there was a difference in follow‐up rates of more than 20% between study arms.
We judged studies to be at high risk of bias overall if they were rated at high risk in at least one domain, and at low risk of bias overall if they were judged to be at low risk across all domains evaluated. We judged the remaining studies to be at unclear risk of bias overall.
Measures of treatment effect
We analyzed dichotomous data by calculating the risk ratio (RR). For cessation, we calculated the RR as ((number of events in intervention condition/intervention denominator)/(number of events in control condition/control denominator)) with a 95% confidence interval (CI), using data at the longest follow‐up period reported.
We analyzed continuous data (other measures of tobacco exposure) by comparing the difference between the mean change from baseline to follow‐up in the intervention and comparator groups, or by comparing absolute data at follow‐up where insufficient data were available on mean change. For outcomes other than cessation where data were reported at multiple time points, we used data at the longest follow‐up point at which ECs were still being provided or their use was encouraged.
Unit of analysis issues
In the case of trials with multiple arms, we do not combine data between arms unless this is the way it has been presented by study authors, or there is no evidence of difference between similar trial arms for the outcome of interest. We note in our analyses where this is the case.
For all but one study, the unit of assignment was the individual. Dawkins 2020 assigned condition based on homeless support service; this was a small pilot study with very few events and hence we judged clustering to have very little impact on our overall result. If larger cluster‐randomized trials are eligible in the future, we will assess whether study authors have adjusted for this clustering, and whether this had an impact on the overall result. When clustering appears to have had little impact on the results, we will use unadjusted quit‐rate data; however when clustering does appear to have an impact on results, we will adjust for this using the intraclass correlation (ICC).
For randomized cross‐over trials, we report results at the end of the first assignment period where available and where sufficiently long to meet our inclusion criteria for outcomes. All other outcomes from randomized cross‐over trials are reported narratively. We offer a narrative synthesis of data from non‐randomized studies and outcomes from comparative trials which aren't reported in sufficient data for meta‐analysis, using effect direction plots as described in the Cochrane Handbook where possible (Higgins 2021).
Dealing with missing data
For smoking cessation, we used a conservative approach, as is standard for the Cochrane Tobacco Addiction Group, treating participants with missing data as still smoking. We based the proportion of people affected by adverse events on the number of people available for follow‐up, and not the number randomized. For other outcomes, we use complete‐case data and do not attempt to impute missing values.
Assessment of heterogeneity
We assessed the clinical and methodological diversity between studies to guide our decision whether data should be pooled. We were also guided by the degree of statistical heterogeneity, assessed by calculating the I2 statistic (Higgins 2003), and considering a value greater than 50% as evidence of substantial heterogeneity. We did not present pooled results where I2 values exceeded 75%.
Assessment of reporting biases
Reporting bias can be assessed using funnel plots, where 10 or more RCTs contribute to an outcome. However, there was only one analysis with sufficient studies to support this approach.
Data synthesis
We provide a narrative summary of the included studies. Where appropriate, we have pooled data from these studies in meta‐analyses. For dichotomous data, we used a fixed‐effect Mantel‐Haenszel model to calculate the RR with a 95% confidence interval, in accord with the standard methods of the Cochrane Tobacco Addiction Group for cessation studies.
For continuous outcomes, we pooled mean differences (or standardized mean differences for studies using different measures for the same construct), using the inverse variance approach (also with a 95% CI).
Subgroup analysis and investigation of heterogeneity
We had planned to undertake subgroup analyses to investigate differences between studies, such as:
-
Intensity of behavioural support used;
-
Type of EC (cartridge; refillable; pod);
-
Instructions for EC use (e.g. study provision, length of provision, whether participants had a role in product choice);
-
Type of participants (e.g. experience of EC use).
However, there were too few studies to conduct such analyses. Should further studies become available in future, we will follow this approach. For continuous outcomes, we subgroup data based on whether absolute values or change scores were available. For adverse events, we subgroup data by length of follow‐up for descriptive purposes.
In the absence of sufficient data for subgroup analyses on EC type, in the text we specify the type of nicotine EC when reporting pooled results for cessation.
Sensitivity analysis
We conducted sensitivity analyses to detect whether pooled results were sensitive to the removal of studies judged to be at high risk of bias.
Summary of findings and assessment of the certainty of the evidence
Following standard Cochrane methodology, we created summary of findings tables for our three main comparisons using GRADEpro GDT: nicotine EC versus non‐nicotine EC; nicotine EC versus NRT; and nicotine EC versus behavioural support only/no support. We selected these comparisons a priori as being the most clinically relevant. In the summary of findings tables, we present data on our primary outcomes (cessation, adverse events, serious adverse events) for these main comparisons. Also following standard Cochrane methodology, we used the five GRADE considerations (study limitations, consistency of effect, imprecision, indirectness and publication bias) to assess the certainty of the body of evidence for each outcome, and to draw conclusions about the certainty of evidence within the text of the review.
Results
Description of studies
Results of the search
For this update, our bibliographic database searches identified 2534 non‐duplicate records (Figure 1). We screened all records and retrieved the full‐text papers of 220 potentially relevant articles. After screening and checking the full‐text of 220 papers, we included 88 records, representing 17 new studies for this update (Bonafont Reyes 2022; Caponnetto 2021; Edmiston 2022; Hajek 2022; Kerr 2020; Kimber 2021; Morphett 2022a; Morphett 2022b; Morris 2022; Myers‐Smith 2022; NCT03492463; Okuyemi 2022; Pratt 2022; Skelton 2022; Tattan‐Birch 2022; Vickerman 2022; White 2021), 41 new articles linked to studies already identified, and 30 new references to ongoing studies (see Characteristics of ongoing studies). Secondary study reports, commentaries, and correspondence relating to included studies are linked to studies in the reference section. Figure 2, Figure 3, Figure 4, Figure 5 and Figure 6 present PRISMA flow charts for previous versions of this review.
PRISMA diagram for 2022 update
PRISMA diagram for 2021 update (Autumn update)
2021 update flow diagram (Spring update)
2020 update flow diagram
2016 update flow diagram
2014 flow diagram
Included studies
In total, we include 78 studies, with 17 new included studies and 61 eligible included studies included in previous versions of the review. Key features of the included studies are summarized below. Further details on each included study can be found in the Characteristics of included studies tables.
Participants
The 78 included studies represented 22,052 participants. Thirty‐four studies were conducted in the USA, 16 were conducted in the UK, eight in Italy, five in Australia, four in Greece, two each in New Zealand and Canada, and one each in Belgium, Ireland, Poland, the Republic of Korea, South Africa, Switzerland, and Turkey. All studies were conducted in adults who smoke. Twenty‐two studies exclusively recruited participants who were not motivated to quit smoking, and 39 studies exclusively recruited participants motivated to quit; motivation was not specified for the other studies. Twenty‐nine studies were recruited from specific population groups; these included nine studies which recruited participants based on physical health condition (heart attack, cancer, HIV, periodontitis, awaiting surgery, smoking‐related chronic disease), five studies which recruited participants with serious mental illness, four studies which recruited participants in treatment or having recently completed treatment for alcohol or other drug use, and three studies in dual users of EC and conventional cigarettes. Two studies recruited people accessing homeless centres or using supported temporary accommodation. One study each recruited: people aged 55 or older, young adults, people who self‐identified as African‐American, pregnant women, and black and Latino participants.
Interventions and comparators
Three studies recruited dual users of combustible cigarettes and EC at baseline, and instructed them to continue using their own EC devices (Czoli 2019; Martinez 2021; Vickerman 2022); the remaining studies all provided some form of nicotine EC.
In two studies where nicotine ECs were provided on their own, nicotine levels were judged to be so low as to be clinically comparable to non‐nicotine EC (Lee 2019; Van Staden 2013); we include these studies in non‐nicotine EC comparisons. Ten studies compared nicotine EC with non‐nicotine EC, 22 studies compared nicotine EC to behavioural support only or to no support, and 17 studies compared nicotine EC to NRT. Five studies compared high‐ versus low‐nicotine EC devices (Caponnetto 2013a; Cobb 2021; Kimber 2021; Morris 2022; White 2021), three studies included comparisons based on flavours (Edmiston 2022; Morris 2022; White 2021), two studies directly compared device types (Kimber 2021; Yingst 2020), and two studies directly compared a freebase nicotine to a salt‐based nicotine device (Morris 2022; Russell 2021). Results from these studies are reported by comparison in Effects of interventions. Further details on the intervention and comparator groups (where applicable) for each study can be found in the Characteristics of included studies tables.
Where reported in the primary research publications, details of the devices tested can also be found in the Characteristics of included studies tables. Of the studies with sufficient data with which to judge, 30 used cartridge devices, 30 used refillable devices, four used both types, four used a pod device, and the remainder did not report device type.
Outcomes
Of the 78 included studies:
-
32 reported data on abstinence at six months or longer
-
55 reported data on adverse events
-
38 reported data on serious adverse events
-
46 reported data on carbon monoxide
-
11 reported data on heart rate
-
13 reported data on blood pressure
-
4 reported data on blood oxygen saturation
-
14 reported data on at least one known toxin/carcinogen
-
7 reported data on at least one measure of lung function
-
14 reported data on study product use at six months or longer
One study (Skelton 2022) measured safety outcomes but did not report them in the text available at time of writing (they may be forthcoming), hence this study currently does not contribute any data to this review.
Study types and funding
Forty studies were RCTs, 22 of which contributed to cessation analyses. Seven studies used randomized cross‐over designs, and the remainder were uncontrolled cohort studies. Of the 65 studies which reported funding information, 47 had no EC industry funding or support.
Excluded studies
We list 91 studies excluded at full‐text stage, along with reasons for exclusion, in the Characteristics of excluded studies table. The most common reason for exclusion was that studies were short‐term, following up participants for periods of less than one week.
Risk of bias in included studies
Overall, we judged ten studies (Bullen 2013; Cobb 2021; Eisenberg 2020; Hajek 2019; Hajek 2022; Kerr 2020; Lee 2018; Lee 2019; Martinez 2021; Myers‐Smith 2022) to be at low risk of bias, 18 to be at unclear risk, and the remaining 50 at high risk of bias (this includes the non‐randomized studies, which we deemed to be at high risk due to this lack of randomization).
Details of risk of bias judgements for each domain of each included study can be found in the Characteristics of included studies table. Figure 7 and Figure 8 illustrate judgements for each included study.
Risk of bias summary: review authors' judgements about each risk of bias item for each included study.
Allocation
We judged 28 studies to be at high risk of selection bias; for the majority of cases this is because the study was not randomized. We rated a pilot cluster‐randomized trial to be at high risk as randomization was not carried out as intended for pragmatic reasons (Dawkins 2020). We judged 25 studies to be at low risk of selection bias, and the remainder to be at unclear risk as there was insufficient information with which to judge.
Blinding
Of the 40 studies assessed for these domains, we judged 28 to be at low risk for both performance and detection bias. We rated 19 to be at high risk for performance or detection bias, or both. In these studies, blinding was not used and different levels of support were provided; this alone or in conjunction with the outcome measures being used (subjective rather than objective measures) meant we thought there was a high risk of bias being introduced. We judged the rest to be at unclear risk, or ineligible for this domain due to single‐arm design.
Incomplete outcome data
We judged most studies (56 out of 78) to be at low risk of attrition bias. We rated nine studies with substantial loss to follow‐up at high risk of attrition bias. The remainder did not provide sufficient data on which to judge, and hence we judged them to be at unclear risk.
Selective reporting
Of the 78 studies, we considered that 40 were at low risk of reporting bias, as all prespecified or expected outcomes were reported. We rated eight as being at high risk, as data were not available as specified in the original protocols (note in some cases these are recent studies, and judgement on these may change as more publications emerge). We judged the rest to be at unclear risk, due to insufficient information with which to make a judgement.
Other potential sources of bias
We considered Ioakeimidis 2018 to be at high risk of other bias; data were from a conference poster and the associated abstract, and quit rates in the intervention arm differed between the two sources. Three further studies were considered to be at unclear risk in this domain.
Effects of interventions
See: Summary of findings 1 Nicotine EC compared to NRT for smoking cessation; Summary of findings 2 Nicotine EC compared to non‐nicotine EC for smoking cessation; Summary of findings 3 Nicotine EC compared to behavioural support only/no support for smoking cessation
Data on our outcomes of interest are summarized below. Due to the volume of data available, some relevant information is hosted on a companion repository; these data are open‐access and can be found at https://doi.org/10.5287/bodleian:JbB1VNgDq. They are referred to below as supplemental tables. Forest plots are available through 'analysis' links; for some outcomes, benefit is plotted on the right, for others on the left. This is due to direction of effect, e.g. an increase in cessation is a benefit, whereas an increase in a carcinogen is not.
Direct comparisons between nicotine EC and other pharmacotherapies
Comparisons reported here include cartridge and refillable nicotine ECs versus NRT, and cartridge nicotine ECs versus varenicline. Only randomized controlled trials contributed data.
Cessation
Pooled data from six studies (2 cartridges, 3 refillable, 1 pod), five of which were rated at low risk of bias and the sixth as unclear, showed higher quit rates in people randomized to nicotine EC than to NRT (risk ratio (RR) 1.63, 95% confidence interval (CI) 1.30 to 2.04; I2 = 10%; 2378 participants; Analysis 1.1). One study included in this analysis, Hajek 2022, was conducted in pregnant women. There was no evidence of a subgroup difference between this study and studies in participants not selected on the basis of pregnancy (P = 0.90, I2 for subgroup differences = 0%). Follow‐up time was based on end of pregnancy, and our primary analysis included only those participants with follow‐up of at least six months. Results were not sensitive to including all participants followed‐up at end of pregnancy (RR 1.49, 95% CI 1.21 to 1.84, I2 = 0%; analysis not shown).
One study (Ioakeimidis 2018), available as a conference presentation only and considered at high risk of bias due to inconsistencies in the data reported and an unclear definition of abstinence, found lower quit rates in people allocated to nicotine EC (cartridge) compared to those allocated to varenicline (RR 0.31, 95% CI 0.11 to 0.82; 54 participants; Analysis 2.1).
Adverse events
Pooled data from four studies (all considered at low risk of bias) showed no evidence of a difference in the number of participants reporting adverse events (AEs) between nicotine EC and NRT arms (RR 1.02, 95% CI 0.88 to 1.19; I2 = 0%; 1702 participants; Analysis 1.2). Hajek 2019 and Bonafont Reyes 2022 did not contribute data to this analysis due to the way in which events were recorded. In Hajek 2019's prespecified adverse reactions of interest, nausea was more frequent in the NRT group, throat/mouth irritation was more frequent in the nicotine EC group, and there was little difference in other reactions (see Supplemental Table 1 for more detail). Bonafont Reyes 2022 recruited participants with COPD and reported "a trend towards decreased dyspnoea and COPD symptoms...in the EC arm compared to the NRT arm", but did not provide further detail.
In Ioakeimidis 2018, reports of sleep disorders were evenly distributed between groups, and nausea was more common in the varenicline arm than in the nicotine EC arm (see Supplemental Table 1 for more detail).
Serious adverse events
Five studies at low risk of bias comparing nicotine ECs with NRT provided data on SAEs. In some studies, no events occurred. Pooled results showed a small increased number of events in the nicotine EC arms, but with wide CIs incorporating no difference, as well as clinically significant harm and clinically significant benefit (RR 1.12, 95% CI 0.82 to 1.52; I2 = 34%; 2411 participants; Analysis 1.3). In Hajek 2022 (conducted in pregnant women), the authors also reported no evidence of a difference in birth outcomes overall. However, low birthweight (< 2500 g) was less frequent in the EC than NRT arm (14.8% vs. 9.6%; RR 0.65, 95% CI 0.47 to 0.90).
No SAEs occurred in Ioakeimidis 2018 (Analysis 2.2).
Carbon monoxide (CO)
Pooled data from three studies (Hatsukami 2020; Kerr 2020; Lee 2018; none considered at high risk of bias) comparing nicotine EC with NRT found that CO levels decreased more in those randomized to nicotine EC (MD −2.74 ppm, 95% CI −5.42 to ‐0.07; I2 = 3%; 191 participants; Analysis 1.4). A fourth, small study (Eisenhofer 2015; n = 11) was reported as a conference abstract and hence had limited data available. At three weeks, this study showed that both EC and NRT groups had "significantly reduced" CO, but between‐group differences were not reported.
Heart rate, blood pressure, and oxygen saturation
Pooled data from two studies (166 participants; one study judged to be at unclear risk of bias, one at low risk) showed no clear evidence of a clinically meaningful difference in heart rate (MD 0.53 bpm, 95% CI ‐1.76 to 2.83; I2 = 0%; Analysis 1.5), systolic blood pressure (MD ‐1.62, 95% CI ‐3.59 to 0.36; I2 = 0%; Analysis 1.6), or blood oxygen saturation (MD ‐0.14, 95% CI ‐0.59 to 0.30; I2 = 0%; Analysis 1.7), although confidence intervals were wide.
Toxicants
Only Hatsukami 2020 (unclear risk of bias, n = 111) contributed data for these outcomes. For PheT, CEMA, and AAMA (Analysis 1.12; Analysis 1.13; Analysis 1.14), point estimates favoured NRT but CIs included no difference. For 3‐HPMA, 2‐HPMA, and HMPMA, point estimates favoured EC but CIs included no difference (Analysis 1.8; Analysis 1.10; Analysis 1.11). There was no evidence of a difference for NNAL (nitrosamine 4‐(methylnitrosamino)‐1‐(3‐pyridyl)‐1‐ butanol) but CIs were again wide (Analysis 1.9).
Lung function
Lee 2018 and Kerr 2020 measured change in FEV1 (forced expiratory volume) and FEV1/FVC (forced vital capacity) (both low risk of bias; n = 81). High statistical heterogeneity (I2 = 89%) precluded pooling for FEV1 (Analysis 1.15); the point estimate for Lee 2018 favoured EC and for Kerr 2020 favoured NRT, but in both cases CIs also included no difference. There was no evidence of a difference for FEV1/FVC, but there was moderate unexplained statistical heterogeneity, and again CIs were wide (MD ‐0.16%, 95% CI ‐1.83 to ‐1.50; I2 = 51%; Analysis 1.16).
Study product use
Five studies reported study product use at six months or longer, but statistical heterogeneity precluded pooling (I2 = 95%). Whereas Russell 2021 and Lee 2018 found no difference between EC and NRT arms, in the other three studies people in the EC arm were more likely to be continuing to use study product (EC) than those in the NRT arm (Analysis 1.18). A companion publication explored long‐term rates in more detail (Butler 2022).
Nicotine EC alone or versus control
Comparisons reported here include nicotine EC versus non‐nicotine EC, and nicotine EC compared to behavioural support only or to no support. In this section, we also reported results from studies in which all participants received nicotine EC (cohort studies and randomized studies which did not differ across arms in EC provision, device generation, or nicotine content).
Cessation
Randomized controlled trials
At six months or longer, quit rates were higher in nicotine EC groups than in comparator groups. Compared to EC without nicotine (placebo EC), pooled results showed nicotine EC produced higher quit rates (RR 1.94, 95% CI 1.21 to 3.13; I2 = 0%; 5 studies of cartridge devices, 1447 participants; Analysis 3.1). The effect size increased when we removed the one study at high risk of bias (Lucchiari 2020). The effect was more pronounced when comparing nicotine EC to behavioural support only or to no support (RR 2.66, 95% CI 1.52 to 4.65; I2 = 0%; 7 studies (4 refillable, 3 cartridge), 3126 participants; Analysis 4.1). As this involved unblinded comparisons with unequal levels of support, we judged all data contributing to this outcome to be at high risk of bias.
Pulvers 2020 (pod device) measured cessation at six months in the intervention group only, using self‐report. As they did not measure cessation at six months in the comparator group, we could not include these data in our meta‐analysis. At six months, 23 (24%) intervention participants were exclusively using EC and 10 (10.4%) reported using neither EC nor combustible cigarettes (making a combined quit rate of 34.4% in the intervention arm at six months).
Data from other studies
Nine studies provided all participants with nicotine EC and assessed abstinence at six months or longer (Table 1; 1 refillable, 6 cartridges, 1 pod, 1 not specified). The highest proportion of quitters was observed in Ely 2013 (cartridge), in which all participants (n = 48) used EC and 18 used additional pharmacotherapy: 44% of participants were abstinent at six months. The lowest quit rates were seen in two studies where participants were not motivated to quit at baseline: in Caponnetto 2013b, 14% of participants were abstinent at 12 months and, in Polosa 2011, 23% of participants were abstinent at six months, but this fell to 13% at 24 months (both studies used cartridge devices).
Study | Motivated or unmotivated to quit smoking? | % abstinent | ||||
---|---|---|---|---|---|---|
Cohort studies | 6‐month | 12‐month | 18‐month | 24‐month | Notes | |
Unmotivated to quit | 19.6% (10/51) | ‐ | ‐ | ‐ | Data from 8‐month follow‐up | |
"Willing to attempt to quit" | 26.6% (8/30) | ‐ | ‐ | ‐ | ‐ | |
Unmotivated to quit | ‐ | 14% (2/14) | ‐ | ‐ | ‐ | |
Unmotivated to quit | 35% (14/40) | ‐ | ‐ | ‐ | ‐ | |
Ely 2013 b | Motivated to quit | 44% (21/48) | ‐ | ‐ | ‐ | ‐ |
Unmotivated to quit | ‐ | 53% (18/34) | ‐ | ‐ | ‐ | |
Unmotivated to quit | 23% (9/40) | ‐ | 15% (6/40) | 13% (5/40) | ‐ | |
Unmotivated to quit | 36% (18/50) | ‐ | ‐ | ‐ | ‐ | |
Not defined | 42% (30/71) | 41% (29/71) | ‐ | ‐ | ‐ |
aTechnically an RCT but observational for purposes of EC analysis
bAll participants (N = 48) used an EC, but 16 also used bupropion and 2 used varenicline
Adverse events
Randomized controlled trials
Pooled data from five studies (none at high risk of bias) showed no evidence of a difference in the number of participants experiencing adverse events when comparing nicotine EC to non‐nicotine EC (RR 1.01, 95% CI 0.91 to 1.11; I2 = 0%; 840 participants; Analysis 3.2). When comparing nicotine EC to behavioural support only or to no support, more people in the groups randomized to nicotine EC reported experiencing adverse events (RR 1.22, 95% CI 1.12 to 1.32; I2 = 41%; 4 studies, 765 participants; Analysis 4.2). As this involved unblinded comparisons with unequal levels of support, we judged all data contributing to this outcome to be at high risk of bias.
A further ten randomized controlled trials provided adverse event or related data for this comparison, but could not be included in the meta‐analysis due to the way in which data were presented (see Supplemental Table 1). In the studies comparing nicotine EC to non‐nicotine EC, one found similar event rates across arms (Caponnetto 2013a), and two reported more events in the nicotine EC arms (Felicione 2019; Tseng 2016). In a further study comparing nicotine to non‐nicotine EC, events were reported by type, with an increase in some seen in the nicotine group and an increase in others seen in the non‐nicotine group (Lucchiari 2020). In the six studies comparing nicotine EC to behavioural support only or traditional cigarettes, Kumral 2016 found an increase in sinonasal symptoms in the group receiving nicotine EC compared to behavioural support only, and Ozga‐Hess 2019 found that throat irritation, cough, and dry mouth increased in the e‐cigarette group relative to the traditional cigarette group. By contrast, Pulvers 2020 found a reduction in respiratory symptoms in the e‐cigarettes compared to the traditional cigarettes group. Begh 2021 found an increase in throat irritation, palpitations and dizziness in the EC group, but decreases in cough, headache, nausea, dry mouth, shortness of breath, and stomach pain. Edmiston 2022 did not break down AEs by group but reported that three subjects experienced a non‐serious adverse event definitely related to study product. Pratt 2022 reported no statistically significant between‐group difference in AEs.
Data from other studies
Eighteen studies provided all participants with nicotine EC and assessed adverse events at one week or longer (see Supplemental Table 1). In the seven studies which tracked event rates over time, six showed adverse events reducing over time (Bell 2017; Caponnetto 2013b; Goniewicz 2017; Polosa 2011; Polosa 2014b; Pratt 2016). Hickling 2019 showed no change. The most commonly‐reported adverse events were throat/mouth irritation, headache, cough, and nausea.
Serious adverse events
Randomized controlled trials
Eight studies compared nicotine EC with non‐nicotine EC and reported data on SAEs; in four of these, no events occurred, so results could not contribute to the meta‐analysis, although they are included in the forest plots for descriptive purposes. In the four studies (three low risk of bias, one unclear) where events occurred, there was no evidence of a difference between groups, but CIs were wide (RR 1.00, 95% CI 0.56 to 1.79; 1272 participants; Analysis 3.3).
Nine studies compared nicotine EC with behavioural support only or no support and reported data on SAEs; in five of these, no events occurred. Pooled results from the four studies in which events occurred showed no clear evidence of a difference between arms, but CIs were wide (RR 1.03, 95% CI 0.54 to 1.97; I2 = 38%; 1993 participants; Analysis 4.3).
In a study in people experiencing homelessness (Dawkins 2020), SAEs were not reported, but authors reported that four to seven participants in the usual‐care arm and five to seven participants in the nicotine EC arm visited Accident & Emergency services at a hospital. The authors reported that these visits were unrelated to study treatment and were assessed to gather data for future economic evaluation. Further detail can be seen in Supplemental Table 2.
Data from other studies
Eight studies provided all participants with nicotine EC and reported SAEs at a week or longer (Supplemental Table 2.). In six of these (Bell 2017; Caponnetto 2013b; Caponnetto 2021; Humair 2014; Polosa 2011; Valentine 2018), authors reported that no SAEs occurred. In NCT02648178 (19 participants), one death occurred (no further detail provided). Hickling 2019 (50 participants) recruited participants from mental health settings; five SAEs were recorded during the study, all of which were psychiatric hospitalizations. None were considered related to study treatment.
Carbon monoxide
Randomized controlled trials
High statistical heterogeneity (I2 = 81%) precluded pooling CO data from the five trials (n = 511, none considered at high risk of bias) comparing nicotine EC with non‐nicotine EC (Analysis 3.4). Point estimates from three studies favoured nicotine EC and from two favoured non‐nicotine EC, but in all cases CIs were consistent with no clinically meaningful difference. Three further randomized studies measured CO levels in those assigned to nicotine EC and those assigned to non‐nicotine EC, but did not present data in a way that could be pooled: George 2019 did not compare data by group; Tseng 2016 reported no between‐group differences; and Meier 2017 found a slightly higher CO reading in those using nicotine EC, but the clinical and statistical significance of this difference was not clear (see Supplemental Table 3 for more detail). These data were from all study participants based on group randomized, not on subsequent EC or cigarette use.
Pooled data from 11 studies comparing nicotine EC to behavioural support alone or to no support resulted in a high I2 value (89%); thus, pooled results were not presented here (see Analysis 4.4 for individual study data). A funnel plot did not show asymmetry (Figure 9). Heterogeneity was primarily driven by magnitude rather than direction of effect, with results in 10 of 11 studies favouring nicotine EC. Three further trials reported data which could not be included in a meta‐analysis. Walele 2018 compared nicotine EC to cigarettes and found CO levels declined in the EC group and remained similar to baseline in the cigarette group. Czoli 2019 instructed baseline dual users to spend periods only using EC or only using traditional cigarettes; CO measured during sole EC use was lower than baseline and lower than during cigarette‐only periods. Further detail can be seen in Supplemental Table 3.
Funnel plot for comparison 4.4
Data from other studies
Nineteen studies provided all participants with nicotine EC and reported data on CO at one week or longer. In the 18 studies that presented change over time, CO declined from baseline although, in Ikonomidis 2018, CO levels were equivalent to baseline again at 24 weeks and, in Polosa 2014b, a decline was observed in people who quit smoking or reduced cigarette consumption by at least half, but not in those who continued smoking at least half as many cigarettes as they had from baseline.
Heart rate
Randomized controlled trials
One RCT (Caponnetto 2013a, unclear risk of bias, n = 141) provided data on heart rate and compared nicotine EC with non‐nicotine EC; there was no evidence of a clinically significant between‐group difference (Analysis 3.5). This was comparable with findings from the one RCT (Hatsukami 2020, unclear risk of bias, n = 90) comparing nicotine EC with no pharmacotherapy, which also found no evidence of a clinically significant difference (Analysis 4.5).
A further three RCTs provided data on heart rate which could not be used to calculate effect estimates. George 2019 compared nicotine to non‐nicotine EC and found no difference in heart rate between arms; Walele 2018 compared a nicotine EC with a traditional cigarette and reported "no clinically significant changes", and Cobb 2021 found decreases in both the EC and QuitSmart cigarette substitute groups, with the decrease being slightly greater in the latter group. See Supplemental Table 4 for further information.
Data from other studies
Six studies in which all participants received a nicotine EC also reported data on heart rate; for five, changes were minimal and directions of effect were mixed, and for Caponnetto 2021 (n = 40) the rate reduced by 9 bpm at 12 weeks (see Supplemental Table 4).
Blood pressure
Caponnetto 2013a found no evidence of a difference in the change in systolic blood pressure (BP) between nicotine EC and non‐nicotine EC arms (unclear risk of bias, 141 participants; Analysis 3.6). Three studies (2 at high risk of bias, 1 at unclear risk of bias) compared nicotine EC to behavioural support only and reported data on systolic blood pressure; there was a small difference favouring the EC arms (MD ‐2.3, 95% CI ‐3.9 to ‐0.7, I2 = 24%; 298 participants; Analysis 4.6). Three further RCTs measured change in blood pressure but presented results in such a way that they could not be pooled. George 2019 compared nicotine EC and non‐nicotine EC and combined data from both groups; BP declined over time. Compared to a QuitSmart cigarette substitute, Cobb 2021 found EC led to a greater reduction in BP. Walele 2018 found "no clinically significant changes" when comparing nicotine EC to a conventional cigarette at two weeks.
Five studies which provided nicotine EC to all participants reported change in blood pressure; results were clinically insignificant except for Caponnetto 2021 in which systolic BP reduced by 12 (from 134 to 122) at 12 weeks (see Supplemental Table 5 for further detail on all studies reporting this outcome).
Oxygen saturation
Hatsukami 2020 found no evidence of a difference in blood oxygen saturation when comparing nicotine EC to cigarettes (90 participants, Analysis 4.7). Van Staden 2013, a short‐term pre‐post study which measured outcomes after two weeks of EC use, found that people who smoked and switched to ECs had significant improvement in blood oxygen saturation (96.2% (SD 1.8) to 97.5% (SD 1.3); 1.3% increase, 95% CI 0.6 to 2.1; P = 0.002).
Toxicants
Unless stated otherwise, all randomized controlled trials measuring these outcomes compared nicotine EC with no pharmacotherapy.
Two trials measured change in 3‐HPMA (one at high risk of bias). In both, the point estimate favoured the EC arm, but pooling was precluded due to difference in measurement methods (Analysis 4.8). Five further studies, in which all participants were given nicotine EC, measured 3‐HPMA; all found reductions over time (Supplemental Table 6).
Five trials measured change in NNAL (four at high risk of bias; Analysis 4.9). Three of the five studies found results favouring nicotine EC, but the final two indicated no difference; statistical heterogeneity was high (I2 = 96%), so pooled results were not presented. Pulvers 2018 and Morris 2022, which provided all participants with nicotine EC, found a reduction in NNAL over time and Czoli 2019, which was a cross‐over trial, found NNAL decreased when using nicotine EC compared to using traditional cigarettes (Supplemental Table 6). An additional two RCTs (one unclear and one low risk of bias) compared nicotine EC versus non‐nicotine EC and found no evidence of difference, with wide CIs and moderate statistical heterogeneity (‐0.02 pmol/mg creatinine, 95% CI ‐0.45 to 0.41; I2 = 54%; 363 participants; Analysis 3.10).
One trial (n = 90, unclear risk of bias) found non‐statistically significant lower levels of 2‐HPMA, HMPMA, PhET and AAMA in nicotine EC arms compared to control (Analysis 4.10; Analysis 4.11; Analysis 4.12; Analysis 4.14). A further two studies in which all participants received nicotine EC found reductions in 2‐HPMA and AAMA measures over time (Supplemental Table 6). No difference was found in the one trial (n = 90, unclear risk of bias) evaluating CEMA (Analysis 4.13).
One trial (n = 90, unclear risk of bias) found reductions in S‐PMA compared to control (Analysis 4.15); this was consistent with the two studies in which all participants received nicotine EC that measured S‐PMA, where levels declined over time (Supplemental Table 6).
Of the 30 remaining measurements in single studies where all participants received a nicotine EC, 25 reduced over time and five increased (Supplemental Table 6).
Lung function
Caponnetto 2013a measured a number of lung function parameters. FeNO increased more in the nicotine EC than the non‐nicotine EC group (MD 2.35, 95% CI 1.78 to 2.92; 90 participants; Analysis 3.7). No difference was found between nicotine and non‐nicotine EC for FEV1 or FEV1/FVC (Analysis 3.8; Analysis 3.9).
Compared to behavioural support only/no support, pooled results from two studies (both high risk of bias) found improvements in FEV1 but with moderate statistical heterogeneity and CIs including no difference (SMD 0.15, 95% CI ‐0.01 to 0.31, I2 = 50%; 714 participants; Analysis 4.16). Data from one study at high risk of bias showed no difference in PEF (peak expiratory flow) 25‐75 (101 participants; Analysis 4.18). Pooled data from two studies (both high risk of bias) showed no difference in FEF (forced expiratory flow) 25‐75, with substantial levels of statistical heterogeneity (MD −0.06, 95% CI −0.18 to 0.06, I2 = 73%; 2 studies, 555 participants; Analysis 4.17). The one study (115 participants, high risk of bias) reporting FEV1/FVC favoured nicotine EC (Analysis 4.19).
Cobb 2021, which randomized participants to EC or the QuitSmart cigarette substitute, measured change in a number of lung function parameters: direction of effect was mixed across these, with no statistically or clinically significant between‐group differences at 12 weeks (Supplemental Table 7).
Two studies which provided all participants with nicotine EC measured change in lung function over time: Hickling 2019 found an increase in peak flow, and Oncken 2015 "no significant differences" in airway function (Supplemental Table 7).
Study product use
Three trials (all low risk), comparing nicotine EC with non‐nicotine EC, reported the number of participants still using EC at six months or longer. Slightly more participants were still using EC in the nicotine EC arms, but CIs were wide and included no difference (RR 1.15, 95% CI 0.94 to 1.41, I2 = 30%; 874 participants; Analysis 3.11). Data on this outcome from single‐arm studies can be found in a companion publication (Butler 2022).
Direct comparisons between nicotine EC
Note, studies reported in this section are only those where participants were randomized to different nicotine EC conditions.
Comparisons based on nicotine dose
Six trials provided data comparing different doses of nicotine in EC (although other studies provided a range of doses, these were not randomly assigned). Only one study provided data on abstinence; in Cobb 2021 (low risk of bias), quit rates were higher in the higher‐dose arm but the 95% CI included no difference (RR 2.50, 95% CI 0.80 to 7.77, 260 participants, Analysis 5.1).
The three studies that provided data on adverse events and contributed to this comparison provided them in such a way that the studies could not be pooled. Kimber 2021 reported "no changes over time or differences between condition", and Pratt 2022 and Morris 2022 did not compare AEs by nicotine strength (see Supplemental Table 1).
In Caponnetto 2013a, no serious adverse events were reported in either arm; in Cobb 2021, there were more events in the higher‐dose arm but CIs were wide (RR 1.51, 95% CI 0.51 to 4.42, 239 participants, Analysis 5.2). In Morris 2022, no serious adverse events occurred (Supplemental Table 2).
Point estimates favoured EC and CIs excluded no difference for carbon monoxide and FEV1/FVC (Analysis 5.3; Analysis 5.9). There were no clinical or statistically significant differences between arms for heart rate, blood pressure, other lung function measures, or NNAL (Analysis 5.4; Analysis 5.5; Analysis 5.6; Analysis 5.7; Analysis 5.8; Analysis 5.10). More participants in the higher‐dose nicotine group were still using EC at six months or longer, but data were from one study and CIs were wide and included no difference (Analysis 5.11). In Yingst 2020 (cross‐over, comparing different doses and different devices), exhaled CO and reported nausea did not differ between devices; self‐reported dizziness was low overall but slightly higher in the higher‐dose arm. Further detail can be found in Supplemental Table 1 and Supplemental Table 3. Morris 2022 measured a range of toxicants but did not compare these based on nicotine level assignments (Supplemental Table 6).
One further study, White 2021, also included comparisons based on nicotine levels (1.8% free‐base nicotine, designated by the authors as 'moderate', and 0.3% free‐base nicotine, designated by the authors as 'low'). This was a factorial trial (unpublished at the time of writing) which, in addition to e‐liquid nicotine content, also manipulated cigarette nicotine content and e‐liquid flavour availability. The authors reported no significant main effects for nicotine content on CO or CEMA, and no statistically significant interactions for these conditions. There also appear to have been no differences in proportions of people experiencing adverse events, but the study terminated early and was likely underpowered to detect differences.
Comparisons based on flavour
One study (Edmiston 2022, n = 300, high risk of bias) randomized participants to different flavours (tobacco versus menthol) and provided data in a way that could have been used to compute risk ratios, although no SAEs occurred in either arm (Analysis 6.1). NNAL and FEV1/FVC were lower in the tobacco flavour group but CIs were wide and included no difference (Analysis 6.2; Analysis 6.4). There was no evidence of a difference in FEV1 (Analysis 6.3). No other outcomes from this paper were eligible for inclusion in our review.
Morris 2022, a randomized cross‐over trial, tested the effect of 10 different flavours (as well as nicotine strengths and salt versus free‐base nicotine). Only their data on AE and SAE were eligible for inclusion in our review, but analyses were not reported by flavour (Supplemental Table 1; Supplemental Table 2).
White 2021 also contributed data to this comparison, with conditions being tobacco flavours only, or tobacco, fruit, dessert and mint flavours. No significant main effects or interactions were found for flavours on the outcomes relevant to this review, namely CO and CEMA, and no difference was discernable in occurrence of AEs. However, as noted above, the study terminated early and hence was underpowered to detect differences.
More information on flavour choices across the studies in this review can be found in a companion publication (Lindson 2022b).
Comparisons based on device type
Kimber 2021 (n = 50, high risk of bias) is the only study to directly compare device types (cartridge versus refillable). Outcomes eligible for this review were CO and AE. There was no difference between arms for CO, but CI were wide (Analysis 7.1). The authors reported "no changes over time or differences between condition" for AEs (see Supplemental Table 1).
Nicotine salt versus free‐base nicotine
One study (Russell 2021, unclear risk of bias) contributed data to this comparison. Quit rates and study product use were both similar between arms (RR 1.25, 95% CI 0.85 to 1.83, n = 285; Analysis 8.1; and RR 1.07, 95% CI 0.82 to 1.41, n = 227; Analysis 8.2, respectively).
As described above, Morris 2022 also tested salt versus free‐base nicotine, but did not provide data broken down by these characteristics for our outcomes of interest (Supplemental Table 1, Supplemental Table 2).
Non‐nicotine EC
Although non‐nicotine ECs serve as a 'control group' in our primary analysis, due to their behavioural properties, they can also be considered an intervention in and of themselves. Comparisons included here are: non‐nicotine EC versus NRT; non‐nicotine EC versus usual care; and non‐nicotine EC as an adjunct to NRT. All contributing data were from randomized controlled trials. None of these studies reported data on change in CO, heart rate, blood pressure, oxygen saturation, toxicants, or lung function.
Cessation
When comparing non‐nicotine EC to behavioural support only, pooled results from two studies (n = 388) found higher quit rates in participants randomized to non‐nicotine EC, but the confidence interval included the possibility of no difference (RR 1.74, 95% CI 0.76 to 3.96; I2 = 0%; Analysis 9.1). When evaluating non‐nicotine EC as an adjunct to NRT, Walker 2020 also found higher quit rates in participants randomized to non‐nicotine EC, although again the confidence interval included no difference (Analysis 10.1).
Lee 2019 compared non‐nicotine EC with NRT; the point estimate favoured NRT, but the confidence interval included no difference (Analysis 11.1).
Adverse events
Eisenberg 2020 found a higher rate of adverse events in the non‐nicotine EC arm than in behavioural support only, with the confidence interval excluding no difference (Analysis 9.2). By contrast, Walker 2020 found fewer adverse events in participants receiving non‐nicotine EC + NRT compared to NRT alone, with the confidence interval excluding no difference (Analysis 10.2). Lee 2019 also found that fewer participants receiving non‐nicotine EC reported adverse events than those receiving NRT, with the confidence interval excluding no difference (Analysis 11.2).
Serious adverse events
Eisenberg 2020 found a higher rate of SAEs in the non‐nicotine EC arm than in the behavioural support‐only arm, but confidence intervals were wide and incorporated clinically significant benefit and clinically significant harm (Analysis 9.3). In Walker 2020, more SAEs occurred in the group randomized to non‐nicotine EC + NRT than in the NRT‐alone group, but the confidence interval included no difference as well as the potential for a clinically significant difference in favour of the intervention (Analysis 10.3). No SAEs were reported in either arm of Lee 2019 (non‐nicotine EC versus NRT).
Advice to use own EC to quit
Three studies did not provide EC but instead provided dual users with advice on how to use their EC to stop smoking. Czoli 2019 and Vickerman 2022 were short‐term studies and contributed data to supplementary tables only. In Martinez 2021, people receiving tailored self‐help material with information on how to use EC to quit smoking had marginally higher long‐term quit rates than those receiving self‐help material without EC advice (RR 1.04, 95% CI 0.89 to 1.22; 2321 participants; Analysis 12.1). The RR was higher and CIs excluded one when compared to an assessment‐only control group. At six months, 64% in the targeted booklet arm, 66% in the generic booklet arm, and 68% in the assessment‐only arm were still using EC.
Combination therapy
Nicotine EC and NRT
This section covers two comparisons: studies in which all arms received NRT and participants were randomized to nicotine EC or non‐nicotine EC, and studies in which all participants received NRT and one arm was randomized to nicotine EC in addition. All studies contributing data were randomized controlled trials. No studies in this group reported data on heart rate, blood pressure, oxygen, or toxicants.
Cessation
Two trials (both at high risk of bias, both testing refillable devices) in which all participants received NRT compared nicotine EC to non‐nicotine EC; pooled results favoured nicotine EC, with the CI excluding no difference (RR 1.77, 95% CI 1.07 to 2.94; I2 = 0%; 1039 participants; Analysis 13.1).
Three studies (two high risk of bias, one unclear risk; two refillable, one cartridge) also compared nicotine EC + NRT to NRT alone. Pooling results from all three studies resulted in high statistical heterogeneity precluding meta‐analysis (I2 = 83%). This heterogeneity was driven by the study of a cartridge device (Morphett 2022a, RR 1.00, 95% CI 0.64 to 1.55, 1712 participants); historically cartridge devices have had poorer nicotine delivery than refillables. Once this study was removed, heterogeneity disappeared (I2 = 0%), but only two studies remained. In these two studies, pooled results showed more people quit in the refillable nicotine EC + NRT arm than in the NRT alone arm (RR 3.53, 95% CI 1.93 to 6.44; 908 participants; Analysis 14.1). In two of these studies, participants in both groups received nicotine patches but, in Morphett 2022b, participants in the NRT only arm also received a short‐acting form of NRT.
Adverse events
Three trials in which nicotine ECs were compared to non‐nicotine ECs reported data on AEs. Baldassarri 2018 reported results combined across groups but noted "no significant differences by treatment group" (Supplemental Table 1). Pooled data from the other two studies also showed no clear evidence of difference (RR 1.11, 95% CI 0.93 to 1.32, I2 = 0%; 677 participants; Analysis 13.2).
The three trials comparing nicotine EC + NRT to NRT alone that contributed data to this outcome were all at high risk of bias. Pooled results showed no evidence of a difference in AEs between arms, but with moderate statistical heterogeneity (RR 0.96, 95% CI 0.83 to 1.11, I2 = 64%; 1984 participants; Analysis 14.2).
Serious adverse events
Pooled data from two studies (one high risk, one unclear) comparing nicotine EC with non‐nicotine EC as adjuncts to NRT showed fewer SAEs in the nicotine EC group than in the non‐nicotine EC group, but the CI included no difference (RR 0.66, 95% CI 0.38 to 1.14, I2 = 0%; 1069 participants; Analysis 13.3).
Four studies (all high risk of bias) provided data on SAEs and compared nicotine EC + NRT to NRT alone. The pooled estimate favoured the NRT‐alone group, but the CI was wide and included no difference (RR 1.26, 95% CI 0.46 to 3.42: I2 = 0; 2245 participants; Analysis 14.3).
Carbon monoxide
Walker 2020 (which compared nicotine EC + NRT, non‐nicotine EC + NRT, and NRT alone) measured change in CO levels but did not report data in a way that could be pooled. CO declined over time, with the greatest reduction seen in the nicotine EC group (see Supplemental Table 3). Pooled data from two studies (one high risk of bias, one unclear) comparing nicotine and non‐nicotine EC as adjuncts to NRT found a slightly greater reduction in CO in the nicotine EC group, but the CI included no clear evidence of a difference (MD ‐1.73 ppm, 95% CI ‐4.44 to 0.98, I2 = 0%; 70 participants; Analysis 13.4) between groups.
Lung function
Baldassarri 2018, which compared nicotine EC to non‐nicotine EC, in which both groups received NRT, found no between‐group differences in FeNO, FEV1, or FVC (Analysis 13.5; Analysis 13.6; Analysis 13.7); confidence intervals were wide for all outcomes.
Study product use
In Walker 2020, at six months, 40% of the patches‐only arm (n = 52) were still using patches, and in the patches + nicotine EC group (n = 317), 22% were using patches only, 45% were using EC only, and 11% were using both patch and EC. In the patches + non‐nicotine EC group (n = 308), 29% were still using patches, 36% were using EC only, and 13% were using both patches and EC. In Baldassarri 2018, there was no difference between arms in product use, but only nine participants contributed data (Analysis 13.8).
Nicotine EC and varenicline
One study, Tattan‐Birch 2022 (high risk of bias, 92 participants), evaluated nicotine EC and varenicline compared to varenicline alone. The study terminated early due to varenicline supply issues (an international recall), and the only data eligible for inclusion in this review related to AEs and SAEs. There was no evidence of a difference in AEs, though CIs were wide (Analysis 15.1), and no SAEs occurred (Analysis 15.2).
Discussion
Summary of main results
This update includes a further 17 studies published since the last version. Our three main comparisons, nicotine EC compared to NRT, nicotine EC compared to non‐nicotine EC, and nicotine EC compared to behavioural support only/no support still show increased quit rates in people assigned to nicotine EC arms; this is now high‐certainty for the comparison with NRT, moderate‐certainty for the comparison with non‐nicotine EC, and very low‐certainty for the comparison with behavioural support only/no support (summary of findings Table 1; summary of findings Table 2; summary of findings Table 3). In absolute terms, pooled data suggest an additional two to six people for every 100 would quit smoking with nicotine EC compared to NRT, an additional two to sixteen people for every 100 would quit smoking with nicotine EC compared to non‐nicotine EC, and an additional one to four people for every 100 would quit smoking with nicotine EC compared to behavioural support only or no support for smoking cessation. Most data come from studies of cartridge and refillable devices.
There remains moderate certainty of no difference in rates of adverse events in nicotine EC compared to non‐nicotine EC, and there is now also moderate‐certainty evidence of no difference in rates of adverse events in nicotine EC compared to NRT. Evidence on adverse events (AEs) and serious adverse events (SAEs) was of low to very low certainty across all other comparisons, due to a paucity of data. Many of the studies which measured SAEs reported no such events in either study arm. For nicotine EC compared to non‐nicotine EC, pooled data suggest no difference in the number of people experiencing AEs or SAEs. Conversely, data from comparisons between nicotine EC and behavioural support alone or no support suggest an additional 14 people per 100 assigned to nicotine EC may experience AEs, but with no difference in SAEs; this evidence was of low and very low certainty, respectively. As with AEs from other smoking cessation treatments (e.g. NRT, Hartmann‐Boyce 2018a), AEs in these studies typically related to irritation at site (e.g. dry mouth, cough) and resolved over time. No studies in any of the different comparison conditions detected serious harms considered to be related to EC use. No authors explicitly identified SAEs as attributable to treatment, but few studies reported detail on this.
Beyond AEs and SAEs, we consider data on a range of safety‐ and health‐related outcomes, including carbon monoxide and other toxins, lung function, blood pressure, pulse, and oxygen levels. Data on all of these outcome measures were limited; for most outcomes within most comparisons, only one or two studies currently contribute data. A companion paper provides more data on the measured toxicants, analysing studies based on actual use of ECs and combustible cigarettes (Hartmann‐Boyce 2022). Consistent with findings from this review, the companion paper found that most measured toxicants were lower in people exclusively using EC than those exclusively smoking or those both smoking and using EC. Most measured toxicants were lower in people using both EC and smoking compared to smoking only.
In this update, we also have more data from studies testing nicotine EC as adjuncts to other stop‐smoking treatments. As with the previous update, pooled data from two studies in which all participants received NRT showed that nicotine EC led to higher quit rates than non‐nicotine EC, but we judged both studies to be at high risk of bias, meaning the effect remains uncertain. Three studies now compare nicotine EC + NRT to NRT alone. Pooling results from all three studies resulted in high statistical heterogeneity precluding meta‐analysis, but this heterogeneity was driven by the one study of a cartridge device. When restricting the analyses to refillable devices, heterogeneity disappeared (I2 = 0%), and results showed more people quit in the nicotine EC + NRT arm than in the NRT alone arm. These results should also be treated with caution as one of the two studies was judged to be at high risk of bias, but they do suggest that this is an area where further research is warranted. It is well‐established that combining short and long‐acting forms of NRT ('combined NRT') leads to greater success than single‐form NRT (Lindson 2019) but, of note, one of the studies showing a benefit of nicotine EC in this comparison compared nicotine EC + patch to short‐acting NRT + patch, suggesting it is not just the 'combined NRT' effect that is driving increased effectiveness.
We also included data on the proportion of participants still using study product (EC or pharmacotherapy) at six months or longer. We introduced this new outcome in our last update after feedback from readers and key stakeholders. There remains no clear evidence of a between‐group difference for this outcome, which is also now explored further in a companion publication (Butler 2022).
Overall completeness and applicability of evidence
This field of research and EC devices themselves continue to evolve rapidly. This is the third update conducted as part of our 'living systematic review' approach, with which we will proceed until at least the end of 2022, meaning we can continue to rapidly incorporate new evidence (see Appendix 1). This is important, as all but two of our analyses currently demonstrate imprecision.
This update incorporates data from 1 June 2021 to 1 July 2022. Subsequent monthly searches will keep this review current. Although studies predominantly came from the USA and UK, overall this review covers data from 14 countries. Geographical range in studies may be particularly important in this area, due to the marked differences in EC regulation between countries; for example, studies conducted in countries that limit nicotine dose in EC or allow only certain EC devices to be tested may observe less pronounced effects on quitting. This review includes studies in some under‐researched populations, including people not motivated to quit smoking, people with substance misuse disorders, people with serious mental health conditions, and people experiencing homelessness. Quit rates in these groups are traditionally lower, which may make it more difficult to detect effects of interventions. However, it could be that these groups may particularly stand to benefit from EC if they are effective because, in absolute terms, conventional cessation methods are often not as effective for them.
As well as the rapid pace of research in this field, evolutions in EC technology pose a challenge when considering the applicability of our evidence to the present. We had downgraded the certainty of our data in the 2016 update, as the devices tested in the trials were first‐generation 'cig‐a‐like' devices which did not deliver nicotine well, meaning the studies may have yielded more conservative estimates than would be seen with newer models, as newer devices and models have tended towards improved nicotine delivery. Nicotine delivery is also relevant to the comparator NRT arms tested; use of both a shorter‐ and a longer‐acting form of NRT show the highest success, and it is important that, where possible, this be the comparator chosen for such trials (Lindson 2019). We no longer downgrade the evidence on this basis as studies with newer device types are now included, although there will always be a time lag between current devices and the research evidence available. Within our primary comparisons, none of the analyses of our primary outcomes signified substantial levels of statistical heterogeneity, despite the fact that different devices were used in the included studies. However, this could be because confidence intervals were wide for individual studies, and does not rule out clinically significant differences in effects between EC types. As further data emerge, we hope to be able to formally test for differences in subgroup analyses, and in head‐to‐head comparisons of different device types. As of this 2022 update, we continue to have only one study of a pod device contributing to our cessation analysis (Russell 2021, abstract only). No studies tested newer disposable devices, which data show are growing in popularity (Tattan‐Birch 2022). There also continues to be little evidence on the impact of different devices, flavours, and nicotine delivery profiles when directly compared to one another. A companion paper explores available data on flavours in more detail (Lindson 2022b).
The adverse effects described in both the RCT and cohort studies continue to look similar, regardless of the brand of EC used or nicotine content, with placebo and nicotine‐containing ECs showing similar numbers and types of adverse events in direct comparisons. They also reflect what is reported in survey data (Dawkins 2013b; Etter 2011).
The structure of our analyses follows standard practice of the Cochrane Tobacco Addiction Group, i.e. evaluating outcomes on an intention‐to‐treat basis, meaning our pooled results represent the effect of offering an EC intervention. This is different from evaluating the per protocol effect, or the effect only in those who use the EC to quit smoking entirely, or continue to smoke whilst also using EC. Although pragmatic and hopefully of use to those designing and delivering interventions, we acknowledge that our intention‐to‐treat approach limits the ability to use the data presented here to draw conclusions about biomarkers in subgroups of participants based on subsequent EC use/smoking profiles. A new companion publication attempts to address this deficit (Hartmann‐Boyce 2022).
Certainty of the evidence
We consider the certainty of the evidence below as it relates to primary outcomes for our three main comparisons: nicotine EC versus NRT; nicotine EC versus non‐nicotine EC; nicotine EC versus behavioural support only/no support (summary of findings Table 1; summary of findings Table 2; summary of findings Table 3). The certainty of evidence for all other comparisons and outcomes should be considered very low due to a paucity of data and issues with risk of bias.
Our summary of findings tables and assessments of certainty are based on the evidence from randomized controlled trials (RCTs). The cohort studies that we include are all deemed to have high risks of bias, which is inherent in the study design. Data presented from these studies need to be interpreted with caution. However, data from cohort studies were reassuringly consistent with data from RCTs.
Risk of bias did not impact on the certainty of evidence for comparisons between nicotine and non‐nicotine EC, or between nicotine EC and NRT. For the latter, we judged all three studies to be at low risk of bias overall. For the former, removing the one study at high risk of bias increased the effect estimate for our efficacy outcome. Risk of bias decreased our certainty in the effect estimates for our nicotine EC versus behavioural support only/no support comparison as, due to the nature of the comparison, blinding was not possible and different levels of support could lead to bias. All but two of our main comparisons were downgraded for imprecision, due to wide confidence intervals and few events. Other than risk of bias and imprecision, we identified no other issues which decreased the certainty of the primary outcomes for our main comparisons. Due to the small number of studies contributing to individual analyses, we were unable to formally test for publication bias and cannot rule this out.
Cessation
All three comparisons found effect estimates favouring nicotine EC for smoking cessation. For nicotine EC versus NRT, we now judge the evidence to be of high certainty, meaning we are now very confident that the true effect lies close to the estimate of the effect. For nicotine EC versus non‐nicotine EC, we continue to judge the evidence to be of moderate certainty, meaning we think the true effect is likely to be close to the estimate of effect. For nicotine EC versus behavioural support only/no support, we judged the evidence to be of very low certainty, meaning we have very little confidence in the effect estimate. Another way to look at this, however, is to consider that nicotine EC versus non‐nicotine EC comparisons isolate the effect of nicotine as provided by an EC, and nicotine EC versus NRT comparisons isolate the effect of the sensorimotor elements provided by an EC. Given that both of these comparisons find a benefit of nicotine EC for smoking cessation, it might logically follow that the comparison between nicotine EC and behavioural support only/no support would find a benefit in favour of nicotine EC, since this comparison would capture both pharmacological and sensorimotor mechanisms of effect. This increases our confidence in the effect of nicotine EC when compared to behavioural support alone or to no support. Nicotine replacement therapy has also been shown to be more effective than behavioural support alone, further supporting the likelihood that nicotine EC would be more effective than behavioural support alone (Hartmann‐Boyce 2018a).
Adverse and serious adverse events
We now have moderate‐certainty evidence of no difference in adverse events for nicotine EC compared to NRT as well as to non‐nicotine EC. For all other outcomes in this category, evidence is of low or very low certainty. Imprecision remains a key issue for these outcomes, and particularly for SAEs. None of the analyses signalled serious harm, nor did complementary data from cohort studies but, unlike our cessation analyses, many of the confidence intervals encompassed the possibility of both clinically significant harm and clinically significant benefit. This uncertainty should reduce as more studies become available.
Potential biases in the review process
We consider the review process we used to be robust. For outcome assessment, we followed the standard methods used for Cochrane Tobacco Addiction Review Group cessation reviews. Our search strategy included the Cochrane Tobacco Addiction Group Specialized Register and we were able to capture a number of ongoing studies. However, there may be unpublished data that our searches did not uncover. We also considered participants lost to follow‐up as continuing to smoke, which is standard practice in this field. There are concerns that frequently updating meta‐analyses can lead to issues with multiple testing; we followed Cochrane guidance in conducting this living systematic review and hence do not adjust for multiple testing (Brooker 2019).
Four of our review authors are authors of the included studies. These authors were not involved in the decisions about inclusion of their studies, or in risk of bias assessment for these studies.
Agreements and disagreements with other studies or reviews
This Cochrane Review aligns with but updates the conclusions of the 2018 U.S. National Academies of Science, Engineering, and Medicine Consensus Study Report, Public Health Consequences of E‐cigarettes (NASEM 2018), which reviewed literature published through August 2017 to address the question, “Do e‐cigarettes help smokers quit smoking combustible tobacco cigarettes?”. Focusing on RCTs and existing systematic reviews, it used a prespecified Level of Evidence framework to develop conclusions. The report’s overall conclusion was that there was “limited evidence that e‐cigarettes may be effective aids to promote smoking cessation.” Based on the RCTs available, it concluded that there was “moderate evidence” that e‐cigarettes containing nicotine were more effective for cessation than e‐cigarettes without nicotine, but “insufficient evidence” about the effectiveness of e‐cigarettes compared to no treatment or to FDA‐approved smoking cessation treatments. Our review contradicts this latter point, as we now find high‐certainty evidence of benefit when comparing nicotine EC with NRT; this is due to the inclusion of studies published after NASEM 2018. Reviews from the Office for Health Improvements and Disparities (formerly Public Health England) conclude that, compared to their 2018 review, there is now stronger evidence that nicotine vaping products are effective for smoking cessation (McNeill 2021, McNeill 2022).
Findings are also broadly consistent with those from other recent reviews, with some exceptions. Amato 2020 did not evaluate effectiveness and focused only on safety; consistent with our review, they found very low‐ to moderate‐certainty evidence on a range of possible adverse effects, with the most frequently reported being cough, dry mouth, shortness of breath, irritation of the mouth and throat, and headache. Consistent with our review, the studies reviewed by McNeill 2022 showed that, compared to combustible cigarettes, using ECs led to a substantial reduction in biomarkers of toxicant exposure associated with cigarette smoking; Wilson 2021 also agrees with this finding. Akiyama 2021 reviewed biomarker findings from clinical studies and also concluded that the use of EC could lead to a significant reduction in exposure to harmful substances compared to traditional cigarettes; this is again consistent with findings from our review. A systematic review of 22 studies found that several carcinogens with a known link to bladder cancer were present in the urine of EC users and recommended further study on the urological safety of ECs (Bjurlin 2021). We will continue to gather information on biomarkers of harm.
Martinez‐Morata 2021 reviewed blood pressure findings and concluded that EC may result in short‐term elevations, but that more data are needed; our review also lacks sufficient data to draw any conclusions about blood pressure at one week or longer. A scoping review by Gugala 2022 looked at the pulmonary health effects of EC and found an association between EC use and negative pulmonary symptoms. EC use resulted in worse outcomes than nonsmoking, but resulted in improved outcomes when compared with combustible cigarette use or dual use of combustible cigarettes and EC. The review by McNeill 2022 found acute and short to medium exposure to most potential respiratory toxicants from ECs to be significantly lower than combustible cigarettes, with substantial reductions in some biomarkers. For the respiratory toxicants assessed at long‐term exposure, evidence was moderate. McNeill 2022 found moderate evidence that exposure to most respiratory toxicants from ECs was similar to non‐use of tobacco or nicotine products. Banks 2022 focused on the absolute risks of EC; in this review, we are interested in both their absolute and relative risks in comparison to smoking.
Zhang 2021 conducted a rapid review; while their pooled analysis also suggested that EC increased quit rates compared to NRT or non‐nicotine EC, they judged the evidence to be of low certainty according to GRADE, driven by imprecision and inconsistency. Zhang 2021 combined studies with NRT comparators and those with non‐nicotine EC comparators in the same analysis and found moderate statistical heterogeneity; we evaluated these two comparisons separately and did not find evidence of statistical heterogeneity. We now include more studies than Zhang 2021 and have no longer downgraded our finding for nicotine EC compared to NRT based on imprecision. Patnode 2021 reviewed evidence on tobacco cessation interventions for the US Preventive Services Task Force (USPFTS 2021). The authors stated that none of their included EC trials suggested higher rates of serious adverse events; this is in line with our analyses. However, they reported that findings across EC trials were inconsistent for effectiveness, with some finding statistically significant evidence of benefit and some finding no statistically significant difference. They did not conduct statistical meta‐analyses and included five trials, all of which are included in our cessation meta‐analyses. None of our cessation meta‐analyses, which include these trials, detected levels of heterogeneity beyond what would be expected from chance alone. Wang 2021 reviewed data both from observational studies and from randomized controlled trials; in the trials, e‐cigarettes were associated with increased smoking cessation (as with our review). In observational studies, ECs were not associated with increased smoking cessation. As discussed in Methods, although we included non‐randomized studies in which an EC intervention is provided in this review, we did not include observational studies in which no EC intervention is provided, due to known issues with confounding.
Chan 2021, Grabovac 2021 and Vanderkam 2022 also reviewed evidence from randomized controlled trials and found higher quit rates in people assigned to nicotine EC than to NRT or non‐nicotine EC, although Grabovac 2021 noted that evidence was less clear at longer follow‐up when comparing nicotine EC to counselling alone. Pound 2021 compared only nicotine EC with NRT; their pooled estimate showed a higher quit rate with nicotine EC (RR 1.42) but 95% CIs were wide and included the possibility of no difference. They included two studies in their comparison that we do not: one which measured cessation at less than six months and hence was not eligible for inclusion in our cessation analysis, and one in which the nicotine level was so low that we classify the study as non‐nicotine (Lee 2019). The latter introduced statistical heterogeneity to their pooled results. We also include additional studies not available at the time of their analyses.
A network meta‐analysis, with searches up‐to‐date until February 2019, used direct and indirect evidence to compare the effectiveness and safety of ECs to placebo, bupropion, NRT and varenicline (Thomas 2022). The evidence was imprecise, however, there was evidence of a benefit of ECs with a nicotine level of 15 mg over placebo. The effect estimate also suggested a benefit of ECs with a 10 mg nicotine level, but the credibility interval indicated the possibility of both benefit and harm. Similarly, when EC was compared with individual pharmacotherapies, the direction of effect was in favour of ECs, however, imprecision means further evidence may change the interpretation of the effect. The safety data for ECs was inconclusive. A second network meta‐analysis also suffered from imprecision when comparing EC and NRT, though CIs were consistent with our results (Quigley 2021). A component network analysis is currently underway that will update this evidence and investigate additional components of the interventions and studies (Lindson 2022a).
Reviews of ECs for policymaking are often broader in scope than our review, which focuses exclusively on their role in supporting smoking cessation in people who smoke. Outside of smoking cessation, there remain unanswered questions about the impact of EC availability and use on young people; we will be evaluating this in a separate review.
PRISMA diagram for 2022 update
PRISMA diagram for 2021 update (Autumn update)
2021 update flow diagram (Spring update)
Risk of bias summary: review authors' judgements about each risk of bias item for each included study.
Funnel plot for comparison 4.4
Comparison 1: Nicotine EC versus NRT, Outcome 1: Smoking cessation
Comparison 1: Nicotine EC versus NRT, Outcome 2: Adverse events
Comparison 1: Nicotine EC versus NRT, Outcome 3: Serious adverse events
Comparison 1: Nicotine EC versus NRT, Outcome 4: Carbon monoxide (ppm)
Comparison 1: Nicotine EC versus NRT, Outcome 5: Heart rate (bpm)
Comparison 1: Nicotine EC versus NRT, Outcome 6: Systolic blood pressure
Comparison 1: Nicotine EC versus NRT, Outcome 7: Blood oxygen saturation
Comparison 1: Nicotine EC versus NRT, Outcome 8: 3‐HPMA (pmol/mg creatinine)
Comparison 1: Nicotine EC versus NRT, Outcome 9: NNAL (pmol/mg creatinine))
Comparison 1: Nicotine EC versus NRT, Outcome 10: 2‐HPMA (pmol/mg creatinine)
Comparison 1: Nicotine EC versus NRT, Outcome 11: HMPMA (pmol/mg creatinine)
Comparison 1: Nicotine EC versus NRT, Outcome 12: PheT (pmol/mg creatinine)
Comparison 1: Nicotine EC versus NRT, Outcome 13: CEMA (pmol/mg creatinine)
Comparison 1: Nicotine EC versus NRT, Outcome 14: AAMA (pmol/mg creatinine)
Comparison 1: Nicotine EC versus NRT, Outcome 15: FEV1
Comparison 1: Nicotine EC versus NRT, Outcome 16: FEV1/FVC (%)
Comparison 1: Nicotine EC versus NRT, Outcome 17: PEF (L/min)
Comparison 1: Nicotine EC versus NRT, Outcome 18: Product use at 6+ months
Comparison 2: Nicotine EC versus varenicline, Outcome 1: Smoking cessation
Comparison 2: Nicotine EC versus varenicline, Outcome 2: Serious adverse events
Comparison 3: Nicotine EC versus non‐nicotine EC, Outcome 1: Smoking cessation
Comparison 3: Nicotine EC versus non‐nicotine EC, Outcome 2: Adverse events
Comparison 3: Nicotine EC versus non‐nicotine EC, Outcome 3: Serious adverse events
Comparison 3: Nicotine EC versus non‐nicotine EC, Outcome 4: Carbon monoxide (ppm)
Comparison 3: Nicotine EC versus non‐nicotine EC, Outcome 5: Heart rate
Comparison 3: Nicotine EC versus non‐nicotine EC, Outcome 6: Systolic blood pressure
Comparison 3: Nicotine EC versus non‐nicotine EC, Outcome 7: FeNO (ppb)
Comparison 3: Nicotine EC versus non‐nicotine EC, Outcome 8: FEV1 (l)
Comparison 3: Nicotine EC versus non‐nicotine EC, Outcome 9: FEV1/FVC
Comparison 3: Nicotine EC versus non‐nicotine EC, Outcome 10: NNAL (pmol/mg creatinine)
Comparison 3: Nicotine EC versus non‐nicotine EC, Outcome 11: Product use at 6+ months
Comparison 4: Nicotine EC versus behavioural support only/no support, Outcome 1: Smoking cessation
Comparison 4: Nicotine EC versus behavioural support only/no support, Outcome 2: Adverse events
Comparison 4: Nicotine EC versus behavioural support only/no support, Outcome 3: Serious adverse events
Comparison 4: Nicotine EC versus behavioural support only/no support, Outcome 4: Carbon monoxide (ppm)
Comparison 4: Nicotine EC versus behavioural support only/no support, Outcome 5: Heart rate (bpm)
Comparison 4: Nicotine EC versus behavioural support only/no support, Outcome 6: Systolic blood pressure
Comparison 4: Nicotine EC versus behavioural support only/no support, Outcome 7: Blood oxygen saturation
Comparison 4: Nicotine EC versus behavioural support only/no support, Outcome 8: 3‐HPMA (SMD)
Comparison 4: Nicotine EC versus behavioural support only/no support, Outcome 9: NNAL (SMD)
Comparison 4: Nicotine EC versus behavioural support only/no support, Outcome 10: 2‐HPMA (pmol/mg creatinine)
Comparison 4: Nicotine EC versus behavioural support only/no support, Outcome 11: HMPMA (pmol/mg creatinine)
Comparison 4: Nicotine EC versus behavioural support only/no support, Outcome 12: PheT (pmol/mg creatinine)
Comparison 4: Nicotine EC versus behavioural support only/no support, Outcome 13: CEMA (pmol/mg creatinine)
Comparison 4: Nicotine EC versus behavioural support only/no support, Outcome 14: AAMA (pmol/mg creatinine)
Comparison 4: Nicotine EC versus behavioural support only/no support, Outcome 15: S‐PMA (nanograms)
Comparison 4: Nicotine EC versus behavioural support only/no support, Outcome 16: FEV1 (SMD)
Comparison 4: Nicotine EC versus behavioural support only/no support, Outcome 17: FEF 25‐75 (litres/second))
Comparison 4: Nicotine EC versus behavioural support only/no support, Outcome 18: PEF 25‐75 (litres/minute)
Comparison 4: Nicotine EC versus behavioural support only/no support, Outcome 19: FEV1/FVC
Comparison 5: Higher versus lower nicotine content, Outcome 1: Smoking cessation
Comparison 5: Higher versus lower nicotine content, Outcome 2: Serious adverse events
Comparison 5: Higher versus lower nicotine content, Outcome 3: Carbon monoxide (ppm)
Comparison 5: Higher versus lower nicotine content, Outcome 4: Heart rate
Comparison 5: Higher versus lower nicotine content, Outcome 5: Systolic blood pressure
Comparison 5: Higher versus lower nicotine content, Outcome 6: FeNO (ppb)
Comparison 5: Higher versus lower nicotine content, Outcome 7: FEV1 (l)
Comparison 5: Higher versus lower nicotine content, Outcome 8: FVC (l)
Comparison 5: Higher versus lower nicotine content, Outcome 9: FEV1/FVC
Comparison 5: Higher versus lower nicotine content, Outcome 10: NNAL (pg/mg creatinine) at 24 weeks
Comparison 5: Higher versus lower nicotine content, Outcome 11: Product use at 6+ months
Comparison 6: Tobacco vs. menthol flavour, Outcome 1: Serious adverse events
Comparison 6: Tobacco vs. menthol flavour, Outcome 2: NNAL (ng/g)
Comparison 6: Tobacco vs. menthol flavour, Outcome 3: FEV1 (% predicted)
Comparison 6: Tobacco vs. menthol flavour, Outcome 4: FEV1/FVC
Comparison 7: Refillable versus cartridge, Outcome 1: Exhaled CO
Comparison 8: Nicotine salt EC versus free‐base nicotine EC, Outcome 1: Smoking cessation
Comparison 8: Nicotine salt EC versus free‐base nicotine EC, Outcome 2: Product use at 6+ months
Comparison 9: Non‐nicotine EC versus behavioural support only/no support, Outcome 1: Smoking cessation
Comparison 9: Non‐nicotine EC versus behavioural support only/no support, Outcome 2: Adverse events at 12 weeks
Comparison 9: Non‐nicotine EC versus behavioural support only/no support, Outcome 3: Serious adverse events at 24 weeks
Comparison 10: Non‐nicotine EC + NRT versus NRT, Outcome 1: Smoking cessation
Comparison 10: Non‐nicotine EC + NRT versus NRT, Outcome 2: Adverse events
Comparison 10: Non‐nicotine EC + NRT versus NRT, Outcome 3: Serious adverse events
Comparison 11: Non‐nicotine EC versus NRT, Outcome 1: Smoking cessation
Comparison 11: Non‐nicotine EC versus NRT, Outcome 2: Adverse events
Comparison 11: Non‐nicotine EC versus NRT, Outcome 3: Serious adverse events
Comparison 12: Advice to use e‐cigarettes compared to no advice to use e‐cigarettes, Outcome 1: Smoking cessation
Comparison 12: Advice to use e‐cigarettes compared to no advice to use e‐cigarettes, Outcome 2: Adverse events at 3 months
Comparison 12: Advice to use e‐cigarettes compared to no advice to use e‐cigarettes, Outcome 3: Serious adverse events at 3 months
Comparison 13: Nicotine EC + NRT versus non‐nicotine EC + NRT, Outcome 1: Smoking cessation
Comparison 13: Nicotine EC + NRT versus non‐nicotine EC + NRT, Outcome 2: Adverse events
Comparison 13: Nicotine EC + NRT versus non‐nicotine EC + NRT, Outcome 3: Serious adverse events
Comparison 13: Nicotine EC + NRT versus non‐nicotine EC + NRT, Outcome 4: Carbon monoxide (ppm)
Comparison 13: Nicotine EC + NRT versus non‐nicotine EC + NRT, Outcome 5: FeNO (ppb)
Comparison 13: Nicotine EC + NRT versus non‐nicotine EC + NRT, Outcome 6: FEV1 (%)
Comparison 13: Nicotine EC + NRT versus non‐nicotine EC + NRT, Outcome 7: FVC (%)
Comparison 13: Nicotine EC + NRT versus non‐nicotine EC + NRT, Outcome 8: Study product use at 6+ months
Comparison 14: Nicotine EC + NRT versus NRT, Outcome 1: Smoking cessation
Comparison 14: Nicotine EC + NRT versus NRT, Outcome 2: Adverse events
Comparison 14: Nicotine EC + NRT versus NRT, Outcome 3: Serious adverse events
Comparison 15: Nicotine EC + varenicline vs. varenicline, Outcome 1: Adverse events at 12 weeks
Comparison 15: Nicotine EC + varenicline vs. varenicline, Outcome 2: Serious adverse events at 12 weeks
Nicotine EC compared to NRT for smoking cessation | ||||||
Patient or population: People who smoke | ||||||
Outcomes | Anticipated absolute effects* (95% CI) | Relative effect | № of participants | Certainty of the evidence | Comments | |
---|---|---|---|---|---|---|
Risk with NRT | Risk with Nicotine EC | |||||
Smoking cessation at 6 months to 1 year Assessed with biochemical validation | Study population | RR 1.63 | 2378 | ⊕⊕⊕⊕ | ‐ | |
6 per 100 | 10 per 100 | |||||
Adverse events at 4 weeks to 6‐9 months Assessed by self‐report | Study population | RR 1.02 | 1702 | ⊕⊕⊕⊝ | ‐ | |
27 per 100 | 27 per 100 | |||||
Serious adverse events at 4 weeks to 1 year Assessed via self‐report and medical records | Study population | RR 1.12 | 2411 | ⊕⊕⊝⊝ | 2 studies reported no events; effect estimate based on the three studies in which events were reported | |
6 per 100 | 7 per 100 | |||||
*The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). For cessation, the assumed risk in the control group is based on assumed quit rates for NRT assuming receipt of limited behavioural stop‐smoking support (as per Hartmann‐Boyce 2018a). The assumed risk for adverse events and serious adverse events is a weighted mean average of quit rates across control groups in contributing studies. | ||||||
GRADE Working Group grades of evidence | ||||||
aDowngraded one level due to imprecision; CIs consistent with benefit and harm |
Nicotine EC compared to non‐nicotine EC for smoking cessation | ||||||
Patient or population: People who smoke cigarettes | ||||||
Outcomes | Anticipated absolute effects* (95% CI) | Relative effect | № of participants | Certainty of the evidence | Comments | |
---|---|---|---|---|---|---|
Risk with non‐nicotine EC | Risk with Nicotine EC | |||||
Smoking cessation at 6‐12 months Assessed with biochemical validation | Study population | RR 1.94 | 1447 | ⊕⊕⊕⊝ | ‐ | |
7 per 100 | 14 per 100 | |||||
Adverse events at 1 week to 6 months Assessed via self‐report | Study population | RR 1.01 | 840 | ⊕⊕⊕⊝ | ‐ | |
9 per 100 | 9 per 100 | |||||
Serious adverse events at 1 week to 1 year Assessed via self‐report and medical records | Study population | RR 1.00 | 1272 | ⊕⊕⊝⊝ | 4 studies reported no events; effect estimate based on the 3 studies in which events were reported | |
3 per 100 | 3 per 100 | |||||
*The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). For cessation, the assumed risk in the control group is based on receipt of moderate‐intensity behavioural stop‐smoking support. The assumed risk for adverse events and serious adverse events is a weighted mean average of quit rates across control groups in contributing studies. | ||||||
GRADE Working Group grades of evidence | ||||||
aNot downgraded for risk of bias. One of four studies considered high risk of bias; removing this study increased the direction of the effect in favour of the intervention. |
Nicotine EC compared to behavioural support only/no support for smoking cessation | ||||||
Patient or population: People who smoke | ||||||
Outcomes | Anticipated absolute effects* (95% CI) | Relative effect | № of participants | Certainty of the evidence | Comments | |
---|---|---|---|---|---|---|
Risk with behavioural support only/no support | Risk with Nicotine EC | |||||
Smoking cessation at 6 to 12 months Assessed using biochemical validation | Study population | RR 2.66 | 3126 | ⊕⊝⊝⊝ | ‐ | |
1 per 100 | 3 per 100 | |||||
Adverse events at 12 weeks to 6 months Assessed via self‐report | Study population | RR 1.22 | 765 | ⊕⊕⊝⊝ | ‐ | |
66 per 100 | 80 per 100 | |||||
Serious adverse events at 4 weeks to 8 months Assessed via self‐report and medical records | Study population | RR 1.03 | 1993 | ⊕⊝⊝⊝ | 5 of the 9 studies reported no SAEs; MA is based on pooled results from 4 studies. | |
2 per 100 | 2 per 100 | |||||
*The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). For cessation, the assumed risk in the control group is based on receipt of limited stop‐smoking support. The assumed risk for adverse events and serious adverse events is a weighted mean average of quit rates across control groups in contributing studies. | ||||||
GRADE Working Group grades of evidence | ||||||
aDowngraded two levels due to risk of bias. Due to lack of blinding and differential support between arms, judged to be at high risk of bias. |
Study | Motivated or unmotivated to quit smoking? | % abstinent | ||||
---|---|---|---|---|---|---|
Cohort studies | 6‐month | 12‐month | 18‐month | 24‐month | Notes | |
Unmotivated to quit | 19.6% (10/51) | ‐ | ‐ | ‐ | Data from 8‐month follow‐up | |
"Willing to attempt to quit" | 26.6% (8/30) | ‐ | ‐ | ‐ | ‐ | |
Unmotivated to quit | ‐ | 14% (2/14) | ‐ | ‐ | ‐ | |
Unmotivated to quit | 35% (14/40) | ‐ | ‐ | ‐ | ‐ | |
Ely 2013 b | Motivated to quit | 44% (21/48) | ‐ | ‐ | ‐ | ‐ |
Unmotivated to quit | ‐ | 53% (18/34) | ‐ | ‐ | ‐ | |
Unmotivated to quit | 23% (9/40) | ‐ | 15% (6/40) | 13% (5/40) | ‐ | |
Unmotivated to quit | 36% (18/50) | ‐ | ‐ | ‐ | ‐ | |
Not defined | 42% (30/71) | 41% (29/71) | ‐ | ‐ | ‐ | |
aTechnically an RCT but observational for purposes of EC analysis |
Outcome or subgroup title | No. of studies | No. of participants | Statistical method | Effect size |
1.1 Smoking cessation Show forest plot | 6 | 2378 | Risk Ratio (M‐H, Fixed, 95% CI) | 1.63 [1.30, 2.04] |
1.1.1 Not selected on pregnancy | 5 | 2059 | Risk Ratio (M‐H, Fixed, 95% CI) | 1.62 [1.29, 2.04] |
1.1.2 Pregnant population | 1 | 319 | Risk Ratio (M‐H, Fixed, 95% CI) | 1.78 [0.45, 6.97] |
1.2 Adverse events Show forest plot | 4 | 1702 | Risk Ratio (M‐H, Fixed, 95% CI) | 1.02 [0.88, 1.19] |
1.2.1 4 weeks | 1 | 29 | Risk Ratio (M‐H, Fixed, 95% CI) | 0.74 [0.31, 1.73] |
1.2.2 6 months | 2 | 563 | Risk Ratio (M‐H, Fixed, 95% CI) | 1.01 [0.82, 1.24] |
1.2.3 3 months after end of pregnancy | 1 | 1110 | Risk Ratio (M‐H, Fixed, 95% CI) | 1.05 [0.84, 1.31] |
1.3 Serious adverse events Show forest plot | 5 | 2411 | Risk Ratio (M‐H, Fixed, 95% CI) | 1.12 [0.82, 1.52] |
1.3.1 4 weeks | 1 | 29 | Risk Ratio (M‐H, Fixed, 95% CI) | Not estimable |
1.3.2 6 months | 2 | 563 | Risk Ratio (M‐H, Fixed, 95% CI) | 1.53 [0.81, 2.88] |
1.3.3 1 year | 1 | 698 | Risk Ratio (M‐H, Fixed, 95% CI) | 1.37 [0.77, 2.41] |
1.3.4 3 months after end of pregnancy | 1 | 1121 | Risk Ratio (M‐H, Fixed, 95% CI) | 0.83 [0.52, 1.31] |
1.4 Carbon monoxide (ppm) Show forest plot | 3 | 191 | Mean Difference (IV, Fixed, 95% CI) | ‐2.74 [‐5.42, ‐0.07] |
1.4.1 Absolute values at follow‐up | 1 | 110 | Mean Difference (IV, Fixed, 95% CI) | ‐1.87 [‐5.15, 1.41] |
1.4.2 Change from baseline | 2 | 81 | Mean Difference (IV, Fixed, 95% CI) | ‐4.47 [‐9.09, 0.15] |
1.5 Heart rate (bpm) Show forest plot | 2 | 166 | Mean Difference (IV, Fixed, 95% CI) | 0.53 [‐1.76, 2.83] |
1.5.1 Absolute values at follow‐up | 1 | 111 | Mean Difference (IV, Fixed, 95% CI) | ‐0.74 [‐5.17, 3.69] |
1.5.2 Change from baseline | 1 | 55 | Mean Difference (IV, Fixed, 95% CI) | 1.00 [‐1.69, 3.69] |
1.6 Systolic blood pressure Show forest plot | 2 | 166 | Mean Difference (IV, Fixed, 95% CI) | ‐1.62 [‐3.59, 0.36] |
1.6.1 Absolute values at follow‐up | 1 | 111 | Mean Difference (IV, Fixed, 95% CI) | 1.00 [‐4.54, 6.54] |
1.6.2 Change from baseline | 1 | 55 | Mean Difference (IV, Fixed, 95% CI) | ‐2.00 [‐4.11, 0.11] |
1.7 Blood oxygen saturation Show forest plot | 2 | 165 | Mean Difference (IV, Fixed, 95% CI) | ‐0.14 [‐0.59, 0.30] |
1.7.1 Absolute values at follow‐up | 1 | 110 | Mean Difference (IV, Fixed, 95% CI) | ‐0.20 [‐0.72, 0.32] |
1.7.2 Change from baseline | 1 | 55 | Mean Difference (IV, Fixed, 95% CI) | 0.00 [‐0.83, 0.83] |
1.8 3‐HPMA (pmol/mg creatinine) Show forest plot | 1 | Mean Difference (IV, Fixed, 95% CI) | Totals not selected | |
1.8.1 Absolute values at follow‐up | 1 | Mean Difference (IV, Fixed, 95% CI) | Totals not selected | |
1.9 NNAL (pmol/mg creatinine)) Show forest plot | 1 | Mean Difference (IV, Fixed, 95% CI) | Totals not selected | |
1.9.1 Absolute values at follow‐up | 1 | Mean Difference (IV, Fixed, 95% CI) | Totals not selected | |
1.10 2‐HPMA (pmol/mg creatinine) Show forest plot | 1 | Mean Difference (IV, Fixed, 95% CI) | Totals not selected | |
1.10.1 Absolute values at follow‐up | 1 | Mean Difference (IV, Fixed, 95% CI) | Totals not selected | |
1.11 HMPMA (pmol/mg creatinine) Show forest plot | 1 | Mean Difference (IV, Fixed, 95% CI) | Totals not selected | |
1.11.1 Absolute values at follow‐up | 1 | Mean Difference (IV, Fixed, 95% CI) | Totals not selected | |
1.12 PheT (pmol/mg creatinine) Show forest plot | 1 | Mean Difference (IV, Fixed, 95% CI) | Totals not selected | |
1.12.1 Absolute values at follow‐up | 1 | Mean Difference (IV, Fixed, 95% CI) | Totals not selected | |
1.13 CEMA (pmol/mg creatinine) Show forest plot | 1 | Mean Difference (IV, Fixed, 95% CI) | Totals not selected | |
1.13.1 Absolute values at follow‐up | 1 | Mean Difference (IV, Fixed, 95% CI) | Totals not selected | |
1.14 AAMA (pmol/mg creatinine) Show forest plot | 1 | Mean Difference (IV, Fixed, 95% CI) | Totals not selected | |
1.14.1 Absolute values at follow‐up | 1 | Mean Difference (IV, Fixed, 95% CI) | Totals not selected | |
1.15 FEV1 Show forest plot | 2 | Std. Mean Difference (IV, Fixed, 95% CI) | Totals not selected | |
1.15.1 Change from baseline | 2 | Std. Mean Difference (IV, Fixed, 95% CI) | Totals not selected | |
1.16 FEV1/FVC (%) Show forest plot | 2 | 81 | Mean Difference (IV, Fixed, 95% CI) | ‐0.16 [‐1.83, 1.50] |
1.16.1 Change from baseline | 2 | 81 | Mean Difference (IV, Fixed, 95% CI) | ‐0.16 [‐1.83, 1.50] |
1.17 PEF (L/min) Show forest plot | 1 | Mean Difference (IV, Fixed, 95% CI) | Totals not selected | |
1.17.1 Change from baseline | 1 | Mean Difference (IV, Fixed, 95% CI) | Totals not selected | |
1.18 Product use at 6+ months Show forest plot | 5 | Risk Ratio (M‐H, Fixed, 95% CI) | Totals not selected | |
Outcome or subgroup title | No. of studies | No. of participants | Statistical method | Effect size |
2.1 Smoking cessation Show forest plot | 1 | Risk Ratio (M‐H, Fixed, 95% CI) | Totals not selected | |
2.2 Serious adverse events Show forest plot | 1 | 54 | Risk Ratio (M‐H, Fixed, 95% CI) | Not estimable |
2.2.1 12 weeks | 1 | 54 | Risk Ratio (M‐H, Fixed, 95% CI) | Not estimable |
Outcome or subgroup title | No. of studies | No. of participants | Statistical method | Effect size |
3.1 Smoking cessation Show forest plot | 5 | 1447 | Risk Ratio (M‐H, Fixed, 95% CI) | 1.94 [1.21, 3.13] |
3.2 Adverse events Show forest plot | 5 | 840 | Risk Ratio (M‐H, Fixed, 95% CI) | 1.01 [0.91, 1.11] |
3.2.1 1 week | 1 | 48 | Risk Ratio (M‐H, Fixed, 95% CI) | 1.50 [0.27, 8.19] |
3.2.2 8 weeks | 1 | 24 | Risk Ratio (M‐H, Fixed, 95% CI) | 1.18 [0.38, 3.66] |
3.2.3 12 weeks | 1 | 255 | Risk Ratio (M‐H, Fixed, 95% CI) | 1.01 [0.94, 1.08] |
3.2.4 6 months | 2 | 513 | Risk Ratio (M‐H, Fixed, 95% CI) | 0.97 [0.71, 1.34] |
3.3 Serious adverse events Show forest plot | 8 | 1272 | Risk Ratio (M‐H, Fixed, 95% CI) | 1.00 [0.56, 1.79] |
3.3.1 1 week | 1 | 48 | Risk Ratio (M‐H, Fixed, 95% CI) | Not estimable |
3.3.2 4 weeks | 1 | 74 | Risk Ratio (M‐H, Fixed, 95% CI) | Not estimable |
3.3.3 8 weeks | 1 | 24 | Risk Ratio (M‐H, Fixed, 95% CI) | 3.50 [0.16, 78.19] |
3.3.4 6 months | 4 | 1009 | Risk Ratio (M‐H, Fixed, 95% CI) | 0.95 [0.52, 1.72] |
3.3.5 1 year | 1 | 117 | Risk Ratio (M‐H, Fixed, 95% CI) | Not estimable |
3.4 Carbon monoxide (ppm) Show forest plot | 5 | Mean Difference (IV, Fixed, 95% CI) | Totals not selected | |
3.4.1 Change from baseline | 1 | Mean Difference (IV, Fixed, 95% CI) | Totals not selected | |
3.4.2 Absolute values at follow‐up | 4 | Mean Difference (IV, Fixed, 95% CI) | Totals not selected | |
3.5 Heart rate Show forest plot | 1 | Mean Difference (IV, Fixed, 95% CI) | Totals not selected | |
3.5.1 Absolute values at follow‐up | 1 | Mean Difference (IV, Fixed, 95% CI) | Totals not selected | |
3.6 Systolic blood pressure Show forest plot | 1 | Mean Difference (IV, Fixed, 95% CI) | Totals not selected | |
3.6.1 Absolute values at follow‐up | 1 | Mean Difference (IV, Fixed, 95% CI) | Totals not selected | |
3.7 FeNO (ppb) Show forest plot | 1 | Mean Difference (IV, Fixed, 95% CI) | Totals not selected | |
3.7.1 Change from baseline | 1 | Mean Difference (IV, Fixed, 95% CI) | Totals not selected | |
3.8 FEV1 (l) Show forest plot | 1 | Std. Mean Difference (IV, Fixed, 95% CI) | Totals not selected | |
3.8.1 Absolute values at follow‐up | 1 | Std. Mean Difference (IV, Fixed, 95% CI) | Totals not selected | |
3.9 FEV1/FVC Show forest plot | 1 | Mean Difference (IV, Fixed, 95% CI) | Totals not selected | |
3.9.1 Absolute values at follow‐up | 1 | Mean Difference (IV, Fixed, 95% CI) | Totals not selected | |
3.10 NNAL (pmol/mg creatinine) Show forest plot | 2 | 363 | Mean Difference (IV, Fixed, 95% CI) | ‐0.02 [‐0.45, 0.41] |
3.10.1 Change from baseline | 1 | 148 | Mean Difference (IV, Fixed, 95% CI) | 15.27 [‐4.98, 35.52] |
3.10.2 Absolute values at follow‐up | 1 | 215 | Mean Difference (IV, Fixed, 95% CI) | ‐0.03 [‐0.46, 0.40] |
3.11 Product use at 6+ months Show forest plot | 3 | 874 | Risk Ratio (M‐H, Fixed, 95% CI) | 1.15 [0.94, 1.41] |
Outcome or subgroup title | No. of studies | No. of participants | Statistical method | Effect size |
4.1 Smoking cessation Show forest plot | 7 | 3126 | Risk Ratio (M‐H, Fixed, 95% CI) | 2.66 [1.52, 4.65] |
4.2 Adverse events Show forest plot | 4 | 765 | Risk Ratio (M‐H, Fixed, 95% CI) | 1.22 [1.12, 1.32] |
4.2.1 12 weeks | 2 | 657 | Risk Ratio (M‐H, Fixed, 95% CI) | 1.20 [1.11, 1.30] |
4.2.2 16 weeks | 1 | 50 | Risk Ratio (M‐H, Fixed, 95% CI) | 1.18 [0.67, 2.07] |
4.2.3 6 months | 1 | 58 | Risk Ratio (M‐H, Fixed, 95% CI) | 11.00 [0.64, 190.26] |
4.3 Serious adverse events Show forest plot | 9 | 1993 | Risk Ratio (M‐H, Fixed, 95% CI) | 1.03 [0.54, 1.97] |
4.3.1 4 to 6 weeks | 2 | 246 | Risk Ratio (M‐H, Fixed, 95% CI) | Not estimable |
4.3.2 8 weeks | 1 | 240 | Risk Ratio (M‐H, Fixed, 95% CI) | 0.29 [0.06, 1.35] |
4.3.3 12 weeks | 2 | 858 | Risk Ratio (M‐H, Fixed, 95% CI) | 3.69 [0.21, 66.17] |
4.3.4 16 weeks | 1 | 50 | Risk Ratio (M‐H, Fixed, 95% CI) | Not estimable |
4.3.5 6 months | 2 | 307 | Risk Ratio (M‐H, Fixed, 95% CI) | 0.71 [0.16, 3.10] |
4.3.6 8 months | 1 | 292 | Risk Ratio (M‐H, Fixed, 95% CI) | 1.78 [0.68, 4.70] |
4.4 Carbon monoxide (ppm) Show forest plot | 11 | Mean Difference (IV, Fixed, 95% CI) | Totals not selected | |
4.4.1 Change from baseline | 2 | Mean Difference (IV, Fixed, 95% CI) | Totals not selected | |
4.4.2 Absolute values at follow‐up | 9 | Mean Difference (IV, Fixed, 95% CI) | Totals not selected | |
4.5 Heart rate (bpm) Show forest plot | 1 | Mean Difference (IV, Fixed, 95% CI) | Totals not selected | |
4.5.1 Absolute values at follow‐up | 1 | Mean Difference (IV, Fixed, 95% CI) | Totals not selected | |
4.6 Systolic blood pressure Show forest plot | 3 | 298 | Mean Difference (IV, Fixed, 95% CI) | ‐2.30 [‐3.91, ‐0.69] |
4.6.1 Change from baseline | 1 | 168 | Mean Difference (IV, Fixed, 95% CI) | ‐2.68 [‐4.38, ‐0.98] |
4.6.2 Absolute values at follow‐up | 2 | 130 | Mean Difference (IV, Fixed, 95% CI) | 1.11 [‐3.95, 6.18] |
4.7 Blood oxygen saturation Show forest plot | 1 | Mean Difference (IV, Fixed, 95% CI) | Totals not selected | |
4.7.1 Absolute values at follow‐up | 1 | Mean Difference (IV, Fixed, 95% CI) | Totals not selected | |
4.8 3‐HPMA (SMD) Show forest plot | 2 | Std. Mean Difference (IV, Fixed, 95% CI) | Totals not selected | |
4.8.1 Absolute values at follow‐up | 1 | Std. Mean Difference (IV, Fixed, 95% CI) | Totals not selected | |
4.8.2 Change from baseline | 1 | Std. Mean Difference (IV, Fixed, 95% CI) | Totals not selected | |
4.9 NNAL (SMD) Show forest plot | 5 | Std. Mean Difference (IV, Fixed, 95% CI) | Totals not selected | |
4.9.1 Absolute values at follow‐up | 2 | Std. Mean Difference (IV, Fixed, 95% CI) | Totals not selected | |
4.9.2 Change from baseline | 3 | Std. Mean Difference (IV, Fixed, 95% CI) | Totals not selected | |
4.10 2‐HPMA (pmol/mg creatinine) Show forest plot | 1 | Mean Difference (IV, Fixed, 95% CI) | Totals not selected | |
4.10.1 Absolute values at follow‐up | 1 | Mean Difference (IV, Fixed, 95% CI) | Totals not selected | |
4.11 HMPMA (pmol/mg creatinine) Show forest plot | 1 | Mean Difference (IV, Fixed, 95% CI) | Totals not selected | |
4.11.1 Absolute values at follow‐up | 1 | Mean Difference (IV, Fixed, 95% CI) | Totals not selected | |
4.12 PheT (pmol/mg creatinine) Show forest plot | 1 | Mean Difference (IV, Fixed, 95% CI) | Totals not selected | |
4.12.1 Absolute values at follow‐up | 1 | Mean Difference (IV, Fixed, 95% CI) | Totals not selected | |
4.13 CEMA (pmol/mg creatinine) Show forest plot | 1 | Mean Difference (IV, Fixed, 95% CI) | Totals not selected | |
4.13.1 Absolute values at follow‐up | 1 | Mean Difference (IV, Fixed, 95% CI) | Totals not selected | |
4.14 AAMA (pmol/mg creatinine) Show forest plot | 1 | Mean Difference (IV, Fixed, 95% CI) | Totals not selected | |
4.14.1 Absolute values at follow‐up | 1 | Mean Difference (IV, Fixed, 95% CI) | Totals not selected | |
4.15 S‐PMA (nanograms) Show forest plot | 1 | Mean Difference (IV, Fixed, 95% CI) | Totals not selected | |
4.15.1 12 weeks | 1 | Mean Difference (IV, Fixed, 95% CI) | Totals not selected | |
4.16 FEV1 (SMD) Show forest plot | 2 | 714 | Std. Mean Difference (IV, Fixed, 95% CI) | 0.15 [‐0.01, 0.31] |
4.16.1 Change from baseline | 2 | 714 | Std. Mean Difference (IV, Fixed, 95% CI) | 0.15 [‐0.01, 0.31] |
4.17 FEF 25‐75 (litres/second)) Show forest plot | 2 | 555 | Mean Difference (IV, Fixed, 95% CI) | ‐0.06 [‐0.18, 0.06] |
4.17.1 Change from baseline | 2 | 555 | Mean Difference (IV, Fixed, 95% CI) | ‐0.06 [‐0.18, 0.06] |
4.18 PEF 25‐75 (litres/minute) Show forest plot | 1 | Mean Difference (IV, Fixed, 95% CI) | Totals not selected | |
4.18.1 Change from baseline | 1 | Mean Difference (IV, Fixed, 95% CI) | Totals not selected | |
4.19 FEV1/FVC Show forest plot | 1 | Mean Difference (IV, Fixed, 95% CI) | Totals not selected | |
4.19.1 Change from baseline | 1 | Mean Difference (IV, Fixed, 95% CI) | Totals not selected |
Outcome or subgroup title | No. of studies | No. of participants | Statistical method | Effect size |
5.1 Smoking cessation Show forest plot | 1 | Risk Ratio (M‐H, Fixed, 95% CI) | Totals not selected | |
5.2 Serious adverse events Show forest plot | 2 | Risk Ratio (M‐H, Fixed, 95% CI) | Totals not selected | |
5.2.1 1 year | 1 | Risk Ratio (M‐H, Fixed, 95% CI) | Totals not selected | |
5.2.2 6 months | 1 | Risk Ratio (M‐H, Fixed, 95% CI) | Totals not selected | |
5.3 Carbon monoxide (ppm) Show forest plot | 3 | 348 | Mean Difference (IV, Fixed, 95% CI) | ‐0.92 [‐1.71, ‐0.13] |
5.3.1 Change from baseline | 2 | 309 | Mean Difference (IV, Fixed, 95% CI) | ‐0.90 [‐1.70, ‐0.10] |
5.3.2 Absolute values at follow‐up | 1 | 39 | Mean Difference (IV, Fixed, 95% CI) | ‐1.66 [‐6.65, 3.33] |
5.4 Heart rate Show forest plot | 1 | Mean Difference (IV, Fixed, 95% CI) | Totals not selected | |
5.4.1 12 weeks | 1 | Mean Difference (IV, Fixed, 95% CI) | Totals not selected | |
5.5 Systolic blood pressure Show forest plot | 1 | Mean Difference (IV, Fixed, 95% CI) | Totals not selected | |
5.5.1 12 weeks | 1 | Mean Difference (IV, Fixed, 95% CI) | Totals not selected | |
5.6 FeNO (ppb) Show forest plot | 1 | Mean Difference (IV, Fixed, 95% CI) | Totals not selected | |
5.6.1 12 weeks | 1 | Mean Difference (IV, Fixed, 95% CI) | Totals not selected | |
5.7 FEV1 (l) Show forest plot | 1 | Mean Difference (IV, Fixed, 95% CI) | Totals not selected | |
5.7.1 12 weeks | 1 | Mean Difference (IV, Fixed, 95% CI) | Totals not selected | |
5.8 FVC (l) Show forest plot | 1 | Mean Difference (IV, Fixed, 95% CI) | Totals not selected | |
5.8.1 12 weeks | 1 | Mean Difference (IV, Fixed, 95% CI) | Totals not selected | |
5.9 FEV1/FVC Show forest plot | 1 | Mean Difference (IV, Fixed, 95% CI) | Totals not selected | |
5.9.1 12 weeks | 1 | Mean Difference (IV, Fixed, 95% CI) | Totals not selected | |
5.10 NNAL (pg/mg creatinine) at 24 weeks Show forest plot | 1 | Mean Difference (IV, Fixed, 95% CI) | Totals not selected | |
5.11 Product use at 6+ months Show forest plot | 1 | Risk Ratio (M‐H, Fixed, 95% CI) | Totals not selected | |
Outcome or subgroup title | No. of studies | No. of participants | Statistical method | Effect size |
6.1 Serious adverse events Show forest plot | 1 | Risk Ratio (M‐H, Fixed, 95% CI) | Totals not selected | |
6.2 NNAL (ng/g) Show forest plot | 1 | Mean Difference (IV, Fixed, 95% CI) | Totals not selected | |
6.2.1 Change from baseline | 1 | Mean Difference (IV, Fixed, 95% CI) | Totals not selected | |
6.3 FEV1 (% predicted) Show forest plot | 1 | Mean Difference (IV, Fixed, 95% CI) | Totals not selected | |
6.3.1 Change from baseline | 1 | Mean Difference (IV, Fixed, 95% CI) | Totals not selected | |
6.4 FEV1/FVC Show forest plot | 1 | Mean Difference (IV, Fixed, 95% CI) | Totals not selected | |
6.4.1 Change from baseline | 1 | Mean Difference (IV, Fixed, 95% CI) | Totals not selected |
Outcome or subgroup title | No. of studies | No. of participants | Statistical method | Effect size |
7.1 Exhaled CO Show forest plot | 1 | Mean Difference (IV, Fixed, 95% CI) | Totals not selected | |
Outcome or subgroup title | No. of studies | No. of participants | Statistical method | Effect size |
8.1 Smoking cessation Show forest plot | 1 | Risk Ratio (M‐H, Fixed, 95% CI) | Totals not selected | |
8.2 Product use at 6+ months Show forest plot | 1 | Risk Ratio (M‐H, Fixed, 95% CI) | Totals not selected | |
Outcome or subgroup title | No. of studies | No. of participants | Statistical method | Effect size |
9.1 Smoking cessation Show forest plot | 2 | 388 | Risk Ratio (M‐H, Fixed, 95% CI) | 1.74 [0.76, 3.96] |
9.2 Adverse events at 12 weeks Show forest plot | 1 | Risk Ratio (M‐H, Fixed, 95% CI) | Totals not selected | |
9.3 Serious adverse events at 24 weeks Show forest plot | 1 | Risk Ratio (M‐H, Fixed, 95% CI) | Totals not selected | |
Outcome or subgroup title | No. of studies | No. of participants | Statistical method | Effect size |
10.1 Smoking cessation Show forest plot | 1 | Risk Ratio (M‐H, Fixed, 95% CI) | Totals not selected | |
10.2 Adverse events Show forest plot | 1 | Risk Ratio (M‐H, Fixed, 95% CI) | Totals not selected | |
10.3 Serious adverse events Show forest plot | 1 | Risk Ratio (M‐H, Fixed, 95% CI) | Totals not selected | |
10.3.1 6 months | 1 | Risk Ratio (M‐H, Fixed, 95% CI) | Totals not selected |
Outcome or subgroup title | No. of studies | No. of participants | Statistical method | Effect size |
11.1 Smoking cessation Show forest plot | 1 | Risk Ratio (M‐H, Fixed, 95% CI) | Totals not selected | |
11.2 Adverse events Show forest plot | 1 | Risk Ratio (M‐H, Fixed, 95% CI) | Totals not selected | |
11.2.1 6 months | 1 | Risk Ratio (M‐H, Fixed, 95% CI) | Totals not selected | |
11.3 Serious adverse events Show forest plot | 1 | 132 | Risk Ratio (M‐H, Fixed, 95% CI) | Not estimable |
11.3.1 6 months | 1 | 132 | Risk Ratio (M‐H, Fixed, 95% CI) | Not estimable |
Outcome or subgroup title | No. of studies | No. of participants | Statistical method | Effect size |
12.1 Smoking cessation Show forest plot | 1 | Risk Ratio (M‐H, Fixed, 95% CI) | Totals not selected | |
12.2 Adverse events at 3 months Show forest plot | 1 | Risk Ratio (M‐H, Fixed, 95% CI) | Totals not selected | |
12.3 Serious adverse events at 3 months Show forest plot | 1 | Risk Ratio (M‐H, Fixed, 95% CI) | Totals not selected | |
Outcome or subgroup title | No. of studies | No. of participants | Statistical method | Effect size |
13.1 Smoking cessation Show forest plot | 2 | 1039 | Risk Ratio (M‐H, Fixed, 95% CI) | 1.77 [1.07, 2.94] |
13.2 Adverse events Show forest plot | 2 | 677 | Risk Ratio (M‐H, Fixed, 95% CI) | 1.11 [0.93, 1.32] |
13.2.1 8 weeks | 1 | 70 | Risk Ratio (M‐H, Fixed, 95% CI) | 1.25 [0.78, 1.99] |
13.2.2 12 weeks | 1 | 607 | Risk Ratio (M‐H, Fixed, 95% CI) | 1.09 [0.90, 1.31] |
13.3 Serious adverse events Show forest plot | 2 | 1069 | Risk Ratio (M‐H, Fixed, 95% CI) | 0.66 [0.38, 1.14] |
13.3.1 8 weeks | 1 | 70 | Risk Ratio (M‐H, Fixed, 95% CI) | 0.59 [0.11, 3.34] |
13.3.2 6 months | 1 | 999 | Risk Ratio (M‐H, Fixed, 95% CI) | 0.67 [0.37, 1.19] |
13.4 Carbon monoxide (ppm) Show forest plot | 2 | 70 | Mean Difference (IV, Fixed, 95% CI) | ‐1.73 [‐4.44, 0.98] |
13.4.1 change from baseline | 1 | 25 | Mean Difference (IV, Fixed, 95% CI) | ‐1.40 [‐4.26, 1.46] |
13.4.2 absolute values at follow‐up | 1 | 45 | Mean Difference (IV, Fixed, 95% CI) | ‐4.60 [‐13.02, 3.82] |
13.5 FeNO (ppb) Show forest plot | 1 | Mean Difference (IV, Fixed, 95% CI) | Totals not selected | |
13.5.1 6 months | 1 | Mean Difference (IV, Fixed, 95% CI) | Totals not selected | |
13.6 FEV1 (%) Show forest plot | 1 | Mean Difference (IV, Fixed, 95% CI) | Totals not selected | |
13.6.1 6 months | 1 | Mean Difference (IV, Fixed, 95% CI) | Totals not selected | |
13.7 FVC (%) Show forest plot | 1 | Mean Difference (IV, Fixed, 95% CI) | Totals not selected | |
13.7.1 6 months | 1 | Mean Difference (IV, Fixed, 95% CI) | Totals not selected | |
13.8 Study product use at 6+ months Show forest plot | 1 | Risk Ratio (M‐H, Fixed, 95% CI) | Totals not selected | |
Outcome or subgroup title | No. of studies | No. of participants | Statistical method | Effect size |
14.1 Smoking cessation Show forest plot | 2 | 980 | Risk Ratio (M‐H, Fixed, 95% CI) | 3.53 [1.93, 6.44] |
14.2 Adverse events Show forest plot | 3 | 1984 | Risk Ratio (M‐H, Fixed, 95% CI) | 0.96 [0.83, 1.11] |
14.2.1 12 weeks | 2 | 421 | Risk Ratio (M‐H, Fixed, 95% CI) | 0.88 [0.69, 1.11] |
14.2.2 7 months | 1 | 1563 | Risk Ratio (M‐H, Fixed, 95% CI) | 0.99 [0.82, 1.19] |
14.3 Serious adverse events Show forest plot | 4 | 2245 | Risk Ratio (M‐H, Fixed, 95% CI) | 1.26 [0.46, 3.42] |
14.3.1 5 weeks | 1 | 7 | Risk Ratio (M‐H, Fixed, 95% CI) | Not estimable |
14.3.2 12 weeks | 1 | 50 | Risk Ratio (M‐H, Fixed, 95% CI) | 3.00 [0.13, 70.30] |
14.3.3 6 months | 1 | 625 | Risk Ratio (M‐H, Fixed, 95% CI) | 1.12 [0.39, 3.27] |
14.3.4 7 months | 1 | 1563 | Risk Ratio (M‐H, Fixed, 95% CI) | Not estimable |
Outcome or subgroup title | No. of studies | No. of participants | Statistical method | Effect size |
15.1 Adverse events at 12 weeks Show forest plot | 1 | Risk Ratio (M‐H, Fixed, 95% CI) | Totals not selected | |
15.2 Serious adverse events at 12 weeks Show forest plot | 1 | Risk Ratio (M‐H, Fixed, 95% CI) | Totals not selected | |