Electronic cigarettes for smoking cessation in pregnancy: a nation–wide mixed methods study

An electronic cigarette (e-cigarette) is a device that delivers nicotine to the lungs in the form of a vapour (Hartmann-Boyce et al, 2018). There are around 3.6 million e-cigarette users in the UK, an increase from 700000 in 2012 (Action on Smoking and Health (ASH), 2017), while in 2019, there were approximately 6.9 million adults smoking conventional tobacco cigarettes in the UK (ASH, 2021a). Euromonitor, a market research group, predicted that the rapidly expanding market for e-cigarettes would have about 55 million users worldwide by the end of 2021 (BBC News, 2018). The UK is the second largest market in the world for e-cigarette products, generating an estimated £2.3 billion in 2020 (Statista, 2022).

Between 2020 and 2021, the NHS reported that in England, 9.6% of pregnant women were smokers at the time of birth and only 35% of pregnant smokers in England accessed stop smoking services (NHS Digital, 2021). In 2020, e-cigarettes were the most common type of smoking cessation strategy in England, with 27.2% of smokers using them to aid quitting conventional smoking (McNeill et al, 2021).

There has been a push for e-cigarettes to be recommended in pregnancy, leading to further exploration regarding the role healthcare professionals play in their recommendation (Royal College of Midwives (RCM), 2019; Schilling et al, 2020). The National Institute for Health and Care Excellence (2021) currently recommends several smoking cessation strategies for pregnant women, which include behavioural therapy, regular carbon monoxide monitoring and nicotine replacement therapy, with the latter to be considered at the earliest opportunity in pregnancy. E-cigarettes are not currently licenced or recommended for use as a smoking cessation strategy in pregnancy. Schilling et al (2020) concluded that there was no compelling evidence regarding the effectiveness of e-cigarettes for smoking cessation in pregnancy.

The literature on e-cigarettes as a smoking cessation strategy in pregnancy is updated frequently; yet there remains insufficient data to draw reliable conclusions (Calder et al, 2021). The RCM (2019) released a position statement regarding supporting pregnant women in quitting smoking. The statement highlights that e-cigarettes contain much lower levels of toxins than tobacco smoke and a pregnant woman should be supported if she wishes to switch to e-cigarettes during pregnancy. The RCM also states that there is no reason to believe e-cigarettes have an adverse effect on breastfeeding, and that they carry a smaller risk of harm compared with smoking to the health of the fetus. Similarly, the UK Health Security Agency (UKHSA), formerly known as Public Health England (PHE), releases periodic evidence updates on e-cigarettes, where the main message has shifted from the 2016 statement that ‘using an e-cigarette is 95% less harmful than smoking conventional cigarettes’ (PHE, 2016) to the latest communication that ‘vaping products are the most popular as well as most effective aid used by people trying to quit smoking’ (PHE, 2021). The NHS (2019) also reported that e-cigarettes are less risky than conventional cigarettes, and it is safer for both mother and fetus for a pregnant woman to vape than to continue smoking tobacco.

A 2018 cross-sectional study surveyed 4509 women who had recently given birth, finding that 2.8% used e-cigarettes at some point during pregnancy, with the majority using them as a smoking cessation strategy (Opondo et al, 2021). The most common reason for using e-cigarettes during pregnancy was to quit smoking. The general population, with 26% current smokers in the UK, also reported using e-cigarettes to cut down on the use of tobacco cigarettes, while 17% use e-cigarettes to quit smoking completely (ASH, 2021b).

Current practice for antenatal care requires a midwife to ask if the woman smokes. If she does, the midwife will refer the woman to a smoking cessation service. Midwives are also able to access information and tools, such as leaflets, to advise pregnant women on smoking cessation. While there is a growing body of continually emerging research, much uncertainty remains because of conflicting findings and a lack of evidence surrounding the long-term effects of vaping on both the mother and fetus (Bruin et al, 2010). Further studies are required to confirm the effect e-cigarettes have on quit attempts in comparison to no smoking cessation treatment (Hartmann-Boyce et al, 2020). The dearth of evidence may translate into uncertainty among smoking cessation professionals, including midwives and specialist services, regarding how best to advise pregnant women (Broadfield, 2019; Cooper et al, 2019). This study aimed to explore midwives’ knowledge, attitudes and practice in relation to e-cigarette use as a smoking cessation strategy in pregnancy.

Methods

This nationwide mixed-methods study used a sequential quantitative-qualitative design. The quantitative elements were the main focus, complemented by a qualitative component. A survey tool was developed and validated to collect quantitative and qualitative data.

The knowledge, attitudes and practice theoretical framework was used to develop the survey tool (Cleland, 1973). The model states that by increasing a healthcare professional’s knowledge, there will be a change in attitude that will be reflected in overall practice (Cleland, 1973), ultimately positing that knowledge and attitudes are the driving forces of behaviour change (Alzghoul and Abdullah, 2016).

Sampling

A minimum sample of 264 participants was calculated based on the total number of midwives in England (RCM, 2018). Originally, 20 NHS trusts were selected by simple randomisation (Trivedi, 2017). The questionnaire was sent to NHS research departments, who were encouraged to forward the survey invitation to midwives employed in their trust. As a result of the COVID-19 pandemic, several NHS organisations were unable to approve or participate in non-COVID-19 research, and so additional participants were recruited via closed Facebook midwifery groups. Permission was gained from group admins to post the survey link in two professionalfacing closed groups.

Eligibility questions were included in the survey to ensure that participants met the inclusion criteria, which were that the participant be a qualified midwife in England. To increase the number of responses, the principles of snowball sampling were applied (Acharya et al, 2013), where participants were encouraged to share the survey link with their friends or colleagues who were qualified midwives in England. This strategy can result in participants having homogenous characteristics (Etikan, 2016). The main drawback of this type of sampling is that as a form of non-probability sampling, it can produce biased results as not all members of the population have the same probability of being chosen (Acharya et al, 2013).

Data collection

The survey tool was a questionnaire that contained 44 items that measured the knowledge, attitudes and practices of registered midwives in England regarding e-cigarette use as a smoking cessation strategy during pregnancy. All but three of the items were closed questions/statements. The survey was open for 72 days between 25 March 2020 and 5 June 2020.

Knowledge domain

Knowledge was interpreted in relation to aspects of e-cigarettes, vaping, safety and practice. Participants were asked questions such as ‘can nicotine cross the placental barrier by vaping?’. The answer categories were ‘yes’, ‘no’ or ‘don’t know’, with a correct answer scoring 1 and ‘don’t know’ or an incorrect answer scoring 0. The maximum knowledge score was 15 and scores were expressed as percentages for bivariate and multivariate analysis.

Attitude domain

Attitude was interpreted as attitude to e-cigarette use during pregnancy as a smoking cessation strategy. Partcipants were asked to respond to statements such as ‘electronic cigarettes should be suggested as a smoking strategy for women during pregnancy’. Answer categories were presented in a Likert scale ranging from ‘strongly agree’ to ‘strongly disagree’, with scores from 0–4. The maximum score was 72. A higher score correlated with a more negative attitude.

Reliability and validity

Cronbach’s alpha was used to evaluate the internal consistency of the knowledge and attitude domains of the survey tool (Vaske et al, 2016). The knowledge domain’s reliability score was 0.70, indicating good reliability, while the attitude domain’s score was 0.84, indicating very good reliability (Nunnally and Bernstein, 1994; Bolarinwa, 2015).

Face validity testing was conducted to improve the readability of the questionnaire, and content validity was tested to ensure the appropriateness of the items in achieving the objectives of the research. Subject experts and members of the public checked over the survey to highlight any errors or inconsistencies. One subject expert was selected who was a member of the midwifery department at the University of Lincoln. A member of the public was asked to read over the survey to ensure the questions were grammatically correct and comprehensible.

Quantitative data analysis

All quantitiative data analysis was completed using RStudio, including chi squared tests, principle component analysis, bivariate analysis and multivariate analysis. Bivariate analysis was done to analyse the relationship between knowledge and attitude scores against covariates such as age, experience and smoking status. Bivariate analysis gave a result using the chi squared test to test relationships. Attitude was analysed against knowledge, likelihood of recommendation, age, smoking status and experience. Odds ratios were used to measure the outcome, likelihood of recommendation and exposure to a risk factor, such as experience and age. Multivariate analysis was done using linear regression models created in RStudio that presented predictors, interactions and covariates. Model 1 was used as a base model with no predictors present. This model explored the covariates of e-cigarette use in pregnancy, which were ‘knowledge’ and ‘likelihood’. All other models were compared to model 1.

Qualitative questions and response analysis

The open-ended qualitative questions were designed to allow participants to provide more in-depth responses, as well as expand on topics covered in previous questions. The qualitative analysis was intended to complement the findings from quantitative analysis (Kroll and Neri, 2009).

Conventional content analysis was used to analyse three open-ended questions with categories developed inductively based on responses (Hsieh and Shannon, 2005). These questions were:

• Please include any additional comments on resources given to midwives regarding e-cigarette use for smoking cessation in pregnancy
• Please include any additional comments regarding the safety of electronic cigarettes as smoking cessation strategy in pregnant women
• Please include any other additional comments on any other aspects of electronic cigarette use in pregnancy.

An excel spreadsheet was used to facilitate analysis; the response was noted and important information was extracted. This was given a code, a shorter explanation of what information the participant was conveying. Finally, similar codes were grouped and given a final category, allowing themes/categories to be extracted and refined.

Ethical considerations

The study protocol, participant information sheet and questionnaire received approval from the Integrated Research Application System (project ID: 274637) and Lincoln Ethics Apply System (reference code: 2020-0977) applications. The consent form was incorporated into the questionnaire, after the participant information sheet and privacy notice. Participants were asked to give consent by checking a box to confirm they had read and understood the information sheet and privacy notice and were happy to continue to the survey. Participants could not continue with the survey if they had not given consent.

Results

A total of 103 midwives completed the survey in full; there were 19 additional participants who did not complete the survey and whose responses were omitted. Table 1 shows the sociodemographic characteristics of the participants. More than half were 30 or more years old (66.0%), had more than 5 years’ experience (54.4%) and had received smoking cessation training (56.3%). The majority (63.1%) were neither smokers nor vapers while 14.6% were vapers. Over half of the participants (54.4%) had cared for a pregnant woman who had successfully quit smoking using e-cigarettes.

Attitude to vaping during pregnancy

Knowledge, likelihood of recommending e-cigarettes, having provided care for a pregnant woman who successfully quit smoking using e-cigarettes and smoking status were all analysed against attitude. All four factors were significantly associated with participants’ attitudes to e-cigarettes as a smoking cessation strategy during pregnancy.

Knowledge and attitude were significant negatively associated (P≤0.001), meaning as knowledge increased, attitude to e-cigarettes became more negative. The likelihood of recommending e-cigarettes and attitude were significantly negatively associated (P≤0.001). Participants who had a more negative attitude to smoking in pregnancy were less likely to recommend them as a smoking cessation strategy. Attitude was significantly negatively associated with whether a participant had provided care for a pregnant woman who successfully quit smoking by using e-cigarettes (P=0.035). Participants who had not provided this care had a more negative attitude to vaping during pregnancy.

Smoking status and attitude had a positive association that was statistically significant (P≤0.001). Participants who smoked and/or vaped themselves had a more positive attitude to vaping during pregnancy. The odds of recommending e-cigarettes were more than 10 times lower in midwives who did not smoke or vape, compared to those who did (odds ratio=0.07), and this was statistically significant (P<0.0001). The odds of recommending e-cigarettes was also lower in midwives who had not received training on smoking cessation in pregnancy, compared to midwives who had (odds ratio=0.18, P=0.581).

Attitude to vaping vs covariates

Four multivariable models to describe the association between attitudes to e-cigarettes and other covariates were developed (Tables 2–5); two are discussed here in detail, as they presented the most significant findings. The first model, shown in Table 2, was based on the knowledge, attitude and practices theoretical framework. This model showed that knowledge and likelihood of recommending e-cigarettes were significantly associated with attitude to e-cigarettes as a smoking cessation strategy in pregnancy (P=0.0012). Midwives with higher knowledge had a less favourable attitude to e-cigarettes, while those who were unlikely to recommend e-cigarettes had even less favourable attitudes compared to those who were likely to recommend e-cigarettes.

The second model, outlined in Table 4, showed that knowledge, providing care for a pregnant women who had successfully quit smoking through e-cigarette use and smoking status were non-significant factors influencing attitude. Increased knowledge led to a more negative attitude to recommending e-cigarettes (P=0.317). Whether a participant had provided care for a pregnant woman who had successfully quit smoking through e-cigarette use led to a more negative attitude to recommending e-cigarettes (P=0.061). Current smoking status positively affected attitude to e-cigarette recommendation during pregnancy (P=0.184).

Content analysis of open-ended survey questions

The categories derived from responses to qualitative questions are summarised in Table 6. Although it is not possible to fully discuss the qualitative findings, commonalities appeared in responses to all three questions, and these commonalities are outlined.

Midwives reported receiving little to no training, information or resources to support their practice. The participants’ free text responses that indicated this included ‘never been given any info or training on e-cigarettes’, ‘haven’t been given any resources’, ‘very little information given at a local level’ and ‘I’ve had no education or training and feel unconfident on the issue’.

Where training, information or resources were reportedly available, these appeared to vary in terms of quantity, quality and content. The participants’ responses ranged from midwives who reported receiving study sessions, advice from smoking cessation midwives and having ‘been given leaflets and information to tell pregnant women’ through to responses indicating minimal guidance such as ‘provided with one A5 poster with minimal information, no training’ and ‘it was mentioned at a conference and we were given information leaflets’. Responses suggesting variation in the content of guidance received included ‘I have been advised by a smoking cessation midwife that vaping is safer to the baby but we don’t know the long term health effects to the woman’ and ‘I have not received enough information to be able to advise one way or another and ensuring it is evidence-based’.

Variations in practice were also evident. One midwife reported having been ‘told not to recommend vaping and to discourage women from doing it due to insufficient evidence regarding safety in pregnancy’, and it was evident that others practised similarly. One reported that since ‘there isn’t enough research to say that women should use them, I recommend quitting’. Conversely, others reported a more pro-vaping stance, including the response ‘we have to advise women there is not lots of evidence of e-cigarettes in pregnancy but we discuss that they are safer and better than normal cigarettes’. Many others reported referring to other, more specialist, professionals: ‘in practice, we had been informed to signpost to a help to quit service’, and ‘at my trust, we have a designated smoking cessation midwife whom we refer women to for the most up to date information regarding [nicotine replacement therapy]/vapes’.

Responses across all three qualitative questions suggested that there was uncertainty and varying opinions among midwives about the safety of vaping in pregnancy and the underlying evidence base. Responses to the final question on other aspects of of e-cigarette use in pregnancy were grouped into one of three categories reflecting midwives’ attitudes and practice in relation to the recommendation of e-cigarettes during pregnancy. Some midwives reported that they would recommend it as a safer alternative to smoking, with responses such as ‘would recommend instead of smoking’ and ‘while we don’t have much research yet about vaping how can something with five or six ingredients compare to the carbon monoxide and 400 poisons cigarettes contain’. There were those who would not recommend it and/or felt it was unsafe, whose responses included ‘it is unsafe’ and ‘on talking to staff many still feel uncomfortable suggesting vaping even though they are aware it is safer than smoking’. Finally, there were those who were uncertain or felt that there was not enough information to inform their practice, with their responses including ‘if there was more evidence, I’d be more inclined to hold discussions with expectant mothers about vaping’ and ‘I need more guidance as I don’t agree with my trust saying if a pregnant woman vapes, she is classed as non-smoker’.

Discussion

The present study explored midwives’ knowledge, attitudes and practices regarding e-cigarettes as a smoking cessation strategy in pregnancy. A key finding of this research was that many respondents (62.2%) reported that they were neither likely nor unlikely to recommend e-cigarette use as a smoking cessation method during pregnancy. Combined with the qualitative findings, which indicated that midwives felt there was insufficient evidence on the safety of e-cigarettes and inadequate guidance to support midwives in this aspect of practice, this is strongly suggestive of uncertainty among midwives, which impacts their practice.

The knowledge, attitude and practices theoretical framework helped in the design and validation of the measurement tool and offered insight into variables influencing practice in participants. Valid and reliable survey tools are of paramount importance in precise measurement of outcomes in social and clinical research (Abowitz and Toole, 2009). A valid and reliable tool is shown in the present study, with acceptable psychometric properties that could be used by other researchers who would like to further investigate the knowledge, attitudes and practices of midwives or other healthcare professionals who play an important role in the provision of smoking cessation services to pregnant women.

The framework was also reflected in the multivariate linear regression model of attitude against knowledge and likelihood of recommendation (Cleland, 1973). The model posits that by increasing an individual’s knowledge, there will be a change in attitude that will be reflected in overall practice (World Health Organization, 2008). The present study’s findings suggested that higher knowledge and more negative attitude scores indicated that a participant was less likely to recommend e-cigarettes as a smoking cessation strategy in pregnancy. This has potential implications for future research and practice, as current guidelines state that midwives should support the use of e-cigarettes as a smoking cessation strategy during pregnancy if the woman has chosen to use this method (RCM, 2019). However, in the present study, increased knowledge appeared to be associated with a more negative attitude. Providing appropriate and effective training for midwives to increase their knowledge may therefore lead to midwives having a more negative attitude to vaping during pregnancy, which in turn may reduce the likelihood that they recommend their use for smoking cessation.

Recommendations by the National Institute for Health and Care Excellence (2018) and the RCM (2019) suggest that brief advice should be given to pregnant women by midwives, who should be ‘confident and competent in discussing smoking in pregnancy’, but that there should be a referral to smoking cessation services for additional support. The qualitative findings suggest that this practice is promoted in some, but not all areas.

Participants felt that they had not received sufficient training to enable them to confidently advise pregnant women on smoking cessation and the use of e-cigarettes in pregnancy. This was also reported in a study of midwives’ attitudes to smoking cessation more generally (Randall, 2009), where it was found that some midwives would not consider themselves qualified to discuss smoking cessation with a pregnant woman. This was further limited by the lack of resources available for them to do so. Where possible, midwives relied on smoking cessation services to provide care to pregnant women looking to quit smoking, as per current guidelines. However, existing research suggests that specialist smoking cessation practitioners also feel ill-equipped to advise pregnant women in relation to the use of e-cigarettes (Hartmann-Boyce et al, 2020).

It could be argued that the provision of training to improve midwives’ knowledge would lead to more support of current guidance, which states that vaping should continue if helpful to quit smoking (RCM, 2019), as many midwives stated that they did not know enough about e-cigarettes to advise correctly and confidently. However, it is important to note that according to the quantitative results, when a participant received training on smoking cessation in pregnancy, their knowledge score increased but their attitude to vaping in pregnancy became more negative. Therefore, it could be argued that there are confounders that interfere with the interplay between attitude to e-cigarettes as a smoking cessation strategy and knowledge. Future research should investigate the complex relationship between attitude and knowledge to further improve understanding of these factors.

Participants who had previously provided care for a pregnant woman who had successfully quit smoking using e-cigarettes were more likely to recommend the use of e-cigarettes during pregnancy in the future. Similarly, Tzelepis et al (2017) assessed midwives’ confidence, attitudes and practices and reported that midwives who had previously delivered smoking cessation care to pregnant women were more confident in their provision of care. In a similar study exploring the use and perceptions of smoking cessation in midwives’ practice, it was found that midwives who had more effective communication with patients about smoking cessation were more likely to have higher outcome expectations (Price et al, 2006). These findings could be linked to the perception that being involved with successful care that resulted in a good outcome influenced attitude and practice more favourably to e-cigarettes. Similarly, in the ‘safety’ question of the qualitative section in the present study, one participant noted anecdotally that she had witnessed a fetus’ growth markedly improve after the switch to vaping from conventional cigarette smoking. This suggests that previous successful care influenced attitudes and practice, and that training that incorporates sharing experiences and success stories in addition to practical methods for supporting women may support midwives to feel more confident in recommending e-cigarettes for smoking cessation.

These findings explore a new area of research regarding the association of midwives’ knowledge and likelihood of recommending e-cigarettes as a smoking cessation strategy in pregnancy to attitude, specifically in relation to variables such as whether a midwife has provided care for a pregnant woman who has successfully quit smoking through e-cigarette use and current smoking status. The findings suggest that midwives would be more comfortable discussing e-cigarettes as a smoking cessation strategy with pregnant women if there were more education and resources available on e-cigarette use in pregnancy.

The Nursing and Midwifery Council (2020) code of conduct states that midwives should always practise in line with the best available evidence, and that any advice given must be evidence based. Given that a perceived absence of evidence was prevalent among the midwives participating in the present study, this may explain some midwives’ reluctance to advise e-cigarette use. Currently, the best available evidence suggests that vaping is safer and should be recommended by healthcare professionals and stop smoking services (PHE, 2019; RCM, 2019). This is reflected in the qualitative responses where some participants stated that they recommend e-cigarettes to pregnant women as they were told that they are safer than conventional cigarettes.

There is a clear need for more research regarding the potential risks and benefits of e-cigarette use during pregnancy, as well as more resources for midwives to allow for informed discussions about e-cigarettes as a smoking cessation strategy. It would also be advantageous to create a clear resource to aid midwives in approaching the subject of e-cigarette use in pregnancy, as this may improve signposting to smoking cessation services. Similarly, it could be advantageous to implement case studies or success stories into training, to create a more positive attitude to vaping during pregnancy, increasing the likelihood of recommendation.

The use of e-cigarettes in pregnancy is becoming an increasingly important subject, as more research is being undertaken relating to the safety to both mother and fetus (Bruin et al, 2010). Therefore, it is vital that midwives are supplied with up-to-date information to improve their knowledge and allow for more informed discussions in practice.

Limitations

As a result of the COVID-19 pandemic, the desired number of participants was not met. Therefore, the study was underpowered and as a result, the results may not be generalisable. This research could be viewed as a feasibility study, with the view to scale up in future research to increase responses, improving generalisability. This could be done by comparing the demographics of the study sample with demographic data of the midwifery workforce in England, to see whether a representative sample was achieved. The average age of midwives was demographically representative compared to data published by the RCM (2018), and so it could be argued that the non-probability sampling method achieved a representative sample, and could be used in a larger study to recruit a cohort from which findings could be generalised to the midwifery workforce in England.

Conclusions

There was a perceived insufficiency of evidence associated with the safety of e-cigarettes during pregnancy, and midwives felt that they were not provided with enough information and resources to allow them to recommend e-cigarettes for smoking cessation. It is important to improve on the knowledge midwives hold regarding e-cigarettes by providing clear, informative resources and a clear approach in discussions with pregnant women looking to use e-cigarettes as a smoking cessation strategy during pregnancy.

The results of this study could serve as a point of reference for future research, as well as future practice, by bringing to light knowledge and attitudes midwives have regarding e-cigarette use in pregnancy, and thus may be able to suggest some changes to improve current practices.

Key points

• This research sought to explore midwives’ knowledge, attitudes and practice in relation to e-cigarette use as a smoking cessation strategy in pregnancy.
• Further investment into activities to alleviate uncertainty and influence attitudes may be required to enable midwives to feel more comfortable with current recommended practice.
• The results could serve as a point of reference for future research as well as future practice by bringing to light midwives’ knowledge and attitudes regarding e-cigarette use in pregnancy, and thus may be able to suggest changes to improve current practices.

CPD reflective questions

• Do you think that guidelines for midwives need to be updated to reflect the knowledge and attitudes shown in this research?
• How did you come to this conclusion?
• How did the COVID-19 pandemic affect the data?
• What key findings can influence practice-based learning and resources?