Background

Pregnant smokers in the UK are routinely recommended nicotine patches (NPs), though their efficacy in this population is limited. Some pregnant smokers try spontaneously to reduce or stop smoking with the help of e-cigarettes (ECs), but safety and efficacy of such use are unknown. We compared NPs and ECs in a pragmatic multi-centre randomised controlled trial.

Setting

Twenty-three hospital sites across England, and one NHS Stop Smoking Service (SSS) in Scotland.

Participants

One thousand one hundred and forty pregnant daily smokers (12–24 weeks’ gestation) motivated to stop smoking, with no strong preference for using NPs or ECs.

Interventions

Participants in the EC arm were posted a refillable EC device with two 10 ml bottles of tobacco-flavoured e-liquid (18 mg nicotine). Participants in the NP arm were posted a 2-week supply of 15 mg/16-hour NPs. Further supplies of both products were provided for up to 8 weeks, with participants encouraged to source further supplies themselves as needed. Participants in both arms received support calls prior to their target quit date (TQD), on the date, and weekly for the next 4 weeks.

Main outcome measures

The primary outcome was biochemically validated prolonged abstinence from smoking at the end of pregnancy (EOP). Three pre-specified sensitivity analyses were conducted: (1) a per-protocol analysis that excluded participants who did not start product use or never established contact with the study team; (2) multiple imputation of missing data; and (3) an analysis that excluded abstainers who were regularly using non-allocated products. Participants lost to follow-up or not providing biochemical validation were included as non-abstainers. Secondary outcomes included self-reported abstinence at different time points, treatment adherence and safety outcomes.

Randomisation

Participants were randomised (1:1) via a pre-programmed list generated by an independent statistician, comprising random permuted blocks. Research midwives conducted the randomisation over the internet via the study database application.

Results

Validation of smoking status via postal saliva sampling kits proved problematic, with only 55% of self-reported abstainers providing useable samples. Due to this, validated prolonged abstinence rates were low (6.8% vs. 4.4% in the EC and NP arms, respectively). The quit rates in the two study arms were not significantly different [risk ratio (RR) = 1.55, 95% confidence interval (CI) 0.95 to 2.53; Bayes factor (BF) = 2.7]. Multiple imputation and per-protocol sensitivity analyses generated the same results, but when abstainers regularly using non-allocated products were excluded, the difference became significant (6.8% vs. 3.6%, RR = 1.93, 95% CI 1.14 to 3.26; BF = 10). About 30% of the sample did not set a TQD. The uptake of support calls was low in both study arms (median sessions = 1), as was the initial allocated product use, though this was higher in the EC arm (39.9% vs. 22.5% using their products at 4 weeks, RR = 1.78, 95% CI 1.48 to 2.13). At the EOP, 33.8% versus 5.6% were using their allocated product in the EC versus NP arm (RR = 6.01, 95% CI 4.21 to 8.58). Regular use of ECs in the NP arm was more common than regular use of nicotine replacement products in the EC arm (17.8% vs. 2.8%).

Rates of adverse birth outcomes were similar in the two study arms, apart from the EC arm having fewer infants with low birthweight than the NP arm (<2500 g) (9.6% vs. 14.8%, RR = 0.65, 95% CI 0.47 to 0.90; BF = 10.3).

Limitations

Return rates of posted saliva samples were low, resulting in low validated quit rates and reduced power to detect a difference between the two study arms. The 2019 outbreak of a lung disease in young vapers in the USA caused by illicit marijuana products but widely reported as due to nicotine vaping led some participants to stop using ECs and return to smoking. Treatment adherence was low, with a substantial proportion of participants not using the help on offer sufficiently enough to test its benefits. The finding that nicotine use does not affect birthweight only concerns nicotine use in late pregnancy, because all participants smoked in early pregnancy.

Conclusions

ECs were not significantly more effective than NPs in the primary analysis, but their effect appears to have been masked by EC use in the NP arm. When this was controlled for, ECs were almost twice as effective than NPs in all abstinence outcomes. ECs did not pose more risks to birth outcomes assessed in this study than NPs and may have reduced the incidence of low birthweight. In pregnant smokers seeking help, ECs are probably more effective than NPs and do not pose any additional risks to women or their infants.

Future work

If specialist SSSs add ECs to their offer to pregnant smokers, routine monitoring of birth outcomes in women using NPs and ECs would provide further important information.

Trial registration

This trial is registered as ISRCTN62025374 and Eudract 2017-001237-65.

Funding

This project was funded by the National Institute for Health and Care Research (NIHR) Health Technology Assemsent programme and will be published in full in Health Technology Assessment; Vol. 27, No. 13. See the NIHR Journals Library website for further project information.