A systematic review of interventions to increase the use of smoking cessation services for women who smoke during pregnancy

Background Although many pregnant women accept referrals to stop‐smoking support, the uptake of appointments often remains low. Aim The aim was to review the success of interventions to increase the uptake of external stop‐smoking appointments following health professional referrals in pregnancy. Materials and Methods Embase, PubMed, Cochrane Central Register of Controlled Trials, Scopus and CINAHL were searched in February 2023 for studies with interventions to increase the uptake rates of external stop‐smoking appointments among pregnant women who smoke. Eligible studies included randomised, controlled, cluster‐randomised, quasi‐randomised, before‐and‐after, interrupted time series, case–control and cohort studies. Cochrane tools assessing for bias and Preferred Reporting Items for Systematic Reviews and Meta‐Analyses (PRISMA) guidelines were followed. Results Two before‐and‐after studies were included, including a combined total of 1996 women who smoked during pregnancy. Both studies had a serious risk of bias, and meta‐analysis was not possible due to heterogeneity. One study testing carbon monoxide monitors and opt‐out referrals showed increased uptake of external stop‐smoking appointments, health professional referrals and smoking cessation rates compared to self‐identified smoking status and opt‐in referrals. Results were limited in the second study, which used carbon monoxide monitors, urinary cotinine levels and self‐disclosed methods to identify the smoking status with opt‐out referrals. Only post‐intervention data were available on the uptake of appointments to external stop‐smoking services. The number of health professional referrals increased, but change in smoking cessation rates was less clear. Conclusions There is insufficient evidence to inform practice regarding strategies to increase the uptake of external stop‐smoking appointments by women during pregnancy.


BACKGROUND
Maternal smoking is one of the most important modifiable risk factors for reducing adverse pregnancy outcomes, 1 including low birth weight, preterm birth and stillbirth. 2 These adverse outcomes have a profound and long-lasting impact on parents, families and care providers, with stillbirth rates showing little improvement globally over the past two decades. 35][6] Although maternal smoking in Australia had shown a downward trend from 15% in 2009 to 9% in 2020, 7 women from low socio-economic backgrounds, young women 8 and Aboriginal and Torres Strait Islander women have disproportionately higher rates of smoking in pregnancy. 9Cochrane systematic review of interventions to support pregnant women to stop smoking showed that feedback (use of carbon monoxide monitors or measures of urine cotinine levels), incentives and psychosocial counselling appear to be effective in increasing smoking cessation rates. 6,10Nicotine replacement therapy (NRT) may also be effective in improving the smoking cessation rate. 11This suggests that women will have a better chance of quitting when engaging with a smoking cessation program that provides these interventions.Health professionals can follow the 5A's pathway, which prompts them to Ask, Advise, Assess, Assist and Arrange smoking cessation support for women. 12This internationally recognised framework promotes smoking cessation by giving women smoking cessation advice and referring them to a locally appropriate stop-smoking service.
Many pregnant women accept a referral to stop-smoking support; however, the uptake of appointments (and thus smoking quit rates) often remains low. 13A study from the UK shows a gap in accepting referrals, with only 14% of referred smokers attending one or more appointments with stopsmoking services. 14In 2019, the Stillbirth Centre of Research Excellence (Stillbirth CRE) introduced the Safer Baby Bundle (SBB), which implements strategies to prevent stillbirths and other adverse perinatal outcomes across Australia. 15The SBB recommends carbon monoxide monitoring and opt-out referrals to appropriate local stop-smoking services.Preliminary surveys completed by postnatal women before the implementation of the SBB showed that only 11% of referred smokers attended one or more appointments with stop-smoking support (unpublished data).
Carbon monoxide monitors and opt-out referrals may increase identification and referral rates; however, women who are not ready to stop smoking may agree to a referral yet not follow with an external appointment. 13Smoking cessation support during pregnancy is identified as a key stillbirth and preterm birth prevention strategy. 2,15,16Many families could avoid the tragedy of stillbirth by ensuring pregnant women receive high-quality smoking cessation support structured to meet their needs.This systematic review aims to determine the efficacy of interventions to increase the uptake of external stop-smoking support appointments following a health professional referral for women who smoke during pregnancy.

MATERIALS AND METHODS
The methodology for this review complies with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines. 17The study protocol was registered through PROSPERO (CRD42021237279) and accepted on 17 March 2021.

Types of studies
All relevant randomised controlled trials, cluster-randomised trials, quasi-randomised controlled trials, controlled before-and-after, interrupted time series, case-control and cohort studies were included.

Population
Pregnant women accessing antenatal care at any stage in pregnancy have been identified as smokers.Smoking is defined as all products that contain nicotine -cigarettes, cigars, pipes, ecigarettes containing nicotine, tobacco that is chewed, sucked, dissolved or inhaled and snuff or snus.

Intervention
Interventions to increase appointment uptake at external stop-smoking support following health professional referral were reviewed in this study.External stop-smoking support is defined as a service not connected to or in the same organisation as the referring health professional.

Comparator
Health professionals opt-in referral to external smoking cessation support, for women identified as smoking while attending routine pregnancy care.

Primary outcome
• Change in the uptake rate of appointments with external stopsmoking support following health professional referral, defined as attending one or more appointments after a referral

Secondary outcomes
• Change in the rate of health professional referrals to external smoking cessation support, defined as being referred to stopsmoking support, as reported by the study investigators • Change in the rates of smoking cessation, defined as smoking cessation during pregnancy and up to six weeks postpartum

Data collection and analysis
The electronic database search was performed using EndNote x9 and Covidence software. 18Initial search results were combined from the five databases, duplicates were removed and all the remaining articles were uploaded to Covidence.Two authors (author one and author two) independently screened all abstracts and/or titles.Full-text articles were obtained for those studies where a decision could not be made based on the title and abstract.Two reviewers (author one and author two) undertook data extraction and quality assessment.Any uncertainties or conflicts were discussed and resolved with a third (author three) and fourth reviewer (author four).

Quality assessment
The Cochrane Risk of Bias tools 19,20

RESULTS
The initial search yielded 1717 potentially eligible studies.After duplicates were removed, 756 unique articles were transferred to Covidence and assessed against eligibility.Thirty-three studies were reviewed in full text as potentially eligible for the review, with 31 studies excluded for the reasons described in Figure 1.
Only two studies met the criteria for inclusion in the review.Metaanalysis could not be performed due to the lack of data and heterogeneity in the interventions and the study sites.

Characteristics of included studies
The review included two before-and-after studies, 13,21 both conducted in the UK, with a combined total of 1996 pregnant women identified as smoking.The study by Campbell et al. 21 (n = 780) as smokers.Both studies were implemented in areas of socio-economic deprivation with a high smoking population (see Table 1).Participants' ethnicity and maternal age were not recorded in either study.

Interventions
Campbell et al. used carbon monoxide monitors to identify smokers and opt-out referrals to external stop-smoking services at the routine 12-week ultrasound scan and found that carbon monoxide monitoring identified more pregnant women who smoke when compared to women self-reporting smoking status.Pre-intervention, 23% (n = 536) of women self-reported smoking at the booking appointment compared to 30% (n = 680) post-intervention.Bauld et al. identified smokers using several approaches, including self-reported, carbon monoxide and urinary cotinine methods, followed by an opt-out referral to external stop-smoking services.This was implemented at the eight-week midwife appointment or the 12-week ultrasound scan.

Risk of bias
Both studies showed a serious risk of bias in one or more areas (see Table 2 • Classification of intervention (as not all eligible women were referred).
• Missing data, due to the lack of quality in pre-intervention data.
• Outcomes measurement, with some women not receiving a referral from the health professional, as they were already in touch with stop-smoking support but were included in the uptake numbers in the study.
A summary of the bias assessment is presented in Table 2.

Change in the uptake of external stop-smoking support appointments
Campbell et al. reported that 2.5% (95% confidence interval (CI): 1.9-3.2%,P < 0.001, n = 57) of pregnant women attended one or more appointments with stop-smoking support pre-intervention, compared to 5.3% (95% CI: 4.4-6.3%,P < 0.001, n = 121) postintervention, showing an increase of 112%.Bauld et al. reported that 3.5% of pregnant women (n = 129) attended one or more appointments with stop-smoking support post-intervention (see Table 3).Due to the lack of pre-implementation data, it was not possible to report uptake rates.to stop-smoking services post-intervention compared to the service's quarterly data from the previous year (see Table 3).

DISCUSSION
To the best of our knowledge, this is the first systematic review to investigate the efficacy of interventions to increase the uptake of external stop-smoking support appointments following a health professional referral of women who smoke during pregnancy.
However, small participant numbers, heterogeneity of outcome measures and low study quality limit review findings.
A combined total of 1996 pregnant women who smoke were included in this review.The results from Campbell et al. are encouraging, indicating that carbon monoxide monitoring may benefit the identification of smokers and that opt-out referrals increase the number of women referred to external stop-smoking support.However, whether this intervention increased external stop-smoking support appointment uptake and smoking cessation rates is unclear.In addition, Bauld et al. reported that some women were already in contact with external stop-smoking support services and did not receive a referral but were included in appointment uptake numbers.The study did not account for this; therefore, the appointment uptake of women referred by a health professional and smoking cessation rates could be much lower.
Unfortunately, no data were reported on women's experiences of the intervention or attendance at external stop-smoking support, so no recommendations could be made.
Both studies show that only a small proportion of pregnant women who smoke were referred to external stop-smoking support services, and even fewer attended their appointment with external stop-smoking services.These low rates of stop-smoking support uptake following referral have also been reported in the  2018 evaluation of the Saving Babies Lives Care Bundle from the UK 22 and the pre-implementation data from the SBB in Australia (unpublished data).This may be due to referrals accepted by women who are not ready to make the change. 13A lack of routinely collected data on referral and uptake of smoking cessation services globally and in Australia prevents robust analysis.Further effective monitoring of service delivery and process outcomes are needed to inform the development of strategies to increase smoking cessation support uptake.
Both of the included studies found that more pregnant women who smoke could be identified when carbon monoxide monitors and urinary cotinine tests were implemented to determine their smoking status.An increase in the identification of smokers using carbon monoxide monitors in pregnancy has also been reported in other studies. 21,23Carbon monoxide monitors can remove the fear of disclosing, with some studies showing that carbon monoxide testing is acceptable for many women as part of their routine pregnancy care. 24,25Health professionals may also find testing helpful when educating women about smoking during pregnancy, as this can eliminate the awkward conversations asking women about their smoking status, 26 and visual feedback on carbon monoxide levels can provide a powerful motivator for quitting. 27wever, some midwives reported feeling concerned that testing may affect their relationship with women, 27,28 as screening such as this makes some women think they cannot be trusted to disclose that they smoke. 29e opt-out approach increases the number of referrals in both studies compared to the opt-in referrals.Bauld et al. reported   an increase in referrals compared to the previous year, with Campbell et al. reporting double the number of women referred.
This increase in referral rates has also been described in other studies. 23,30Implementing the opt-out referrals at the 12-week ultrasound scan may have contributed to the increased referral rates in both studies included in this review.However, the motivating factor of seeing the baby on the scan was not accounted for in either analysis.Therefore, we could not identify whether the ultrasound scan was the motivating factor associated with increased referral and uptake rates, as seen in a 2009 US study by Stotts et al., 31 which found that smoking cessation increased in light smokers when motivational interviewing was provided at a second-or third-trimester ultrasound scan.
Socio-economic factors may have influenced the uptake of referral and quitting in the included studies.In several countries, including USA, 32 Britain, Finland, 33 China, Ghana, India and South Africa, 34 lower socio-economic factors have been linked to greater use of tobacco in members of the general public.Low socioeconomic status has also been associated with increased tobacco use and the reduced likelihood of quitting smoking in pregnancy. 35is could have reduced the potential number of women who stopped smoking from the studies in this review.Interventions that are personalised and sensitive to the needs of women who face barriers posed by complex socio-economic disadvantages may improve outcomes. 36Financial incentives offered to lowincome pregnant women have shown improvements in engagement and treatment, 37 and The Breathe Study showed favourable results with a clinic-based specialist service, offering a face-toface appointment with a specialist smoking cessation midwife. 38wever, further research is needed on strategies to help women uptake external appointments with stop-smoking services.
In among Indigenous mothers. 7A 2016 systematic review estimated that around 87% of pregnant women who smoked tried to quit and were unsuccessful, or did not attempt to stop, 40 showing the need for improvements in current cessation supports.Offering referrals to external stop-smoking services at all antenatal appointments could improve the rate of appointment uptake further.Campbell et al. 30 identified that a significant number of women who initially opted out of the referral would have liked to receive further stop-smoking service referrals at subsequent antenatal appointments.

Strengths and limitations
The strengths of this review include the rigorous, pre-defined methods and a comprehensive search strategy to include all relevant studies.However, the small number and the low quality of the studies included in the review allow only limited conclusions on interventions to increase the uptake of external stop-smoking support.
Both studies found that most maternity services could integrate opt-out referral pathways into care.However, both studies were performed in areas with higher-than-average smoking rates in disadvantaged neighbourhoods.Therefore, these results may only be generalisable in regions that provide carbon monoxide monitoring and opt-out referrals at appointments offering ultrasound scans.

CONCLUSION
Due to the limited number of studies available on interventions to increase external stop-smoking support appointment uptake, there needs to be more evidence to inform future practice.There is some indication that carbon monoxide monitoring may benefit the identification of smokers; however, evidence on interventions that increase external stop-smoking support referrals, external stop-smoking support uptake and smoking cessation rates is less clear.In addition, no data were reported on women's experiences of the intervention and external stop-smoking services.Further high-quality research is needed on what will prepare women to make the change to enable the uptake of appointments with external stop-smoking services after referral.Research on how health professionals can support women to prepare to make the change is also needed.
were used by two reviewers (author one and author two) to assess the risk of bias.The risk of bias (RoB2) tool assessed the randomised studies, and the nonrandomised studies of interventions tool (ROBINS-I) assessed the non-randomised studies.Each domain was given an overall assessment of potential bias with scores of Low/Moderate/ Serious/Critical/No Information.Studies were not excluded on the grounds of risk of bias.
recruited 1216 participants in two sites, Kings Mill Hospital and Sherwood Women's Centre, in 2012/2013.The second study by Bauld et al. 13 recruited 780 participants in a hospital in Dudley over eight months and a community maternity setting in South Birmingham over six months in 2010/2011.Campbell et al. collected data from the same periods (May to October) one year apart.Bauld et al. used the data reported for the year before the intervention to compare to post-intervention.Campbell et al. recorded stopsmoking service referrals, uptake and four-week quit rates, while Bauld et al. recorded referrals and four-week quit rates.Postintervention, the external stop-smoking service uptake rates were recorded; however, no pre-implementation data on uptake rates were documented.No data were available on the secondary outcomes and women's experiences of interventions and attending smoking cessation support.Participant characteristics Across both studies, 6005 pregnant women accessed maternity care post-intervention, 2293 in the study by Campbell et al. and 3712 in the study by Bauld et al.Overall, 24% (n = 1460) of women were identified as smoking while pregnant, with Campbell et al. identifying 30% (n = 680) and Bauld et al. 22%

-
).The study byCampbell et al. showed a serious risk of bias in the measurement of outcomes.Pre-intervention referrals were completed in routine antenatal care, and post-intervention referrals were completed at the 12-week ultrasound scan after the woman had seen an image of her baby.This change could increase the proportion of women attending external stop-smoking service appointments and was not accounted for in the study.All other areas of bias were assessed as moderate.This study was not randomised, and there were no identified missing data; most smokers were enrolled into the study using carbon monoxide monitors, and data comparison aligned with the same period in the previous year.Bias was assessed to be serious in one area F I G U R E 1 Search results, study selection and inclusion process.SSS, stop-smoking service.No intervention to increase SSS uptake (n=8) -No SSS uptake data available (n=6) -No referral to SSS (n=5) -No comparison data (n=1) -Conference abstract only (n=6) and moderate in the other six, resulting in a judgement of serious risk of bias overall.The study by Bauld et al. showed a serious risk of bias in three areas and thus a serious risk of bias overall.The areas were as follows:

TABLE 1 Characteristics
of included studies Campbell et al. reported 2% (95% CI: 1.5-2.7%,P<0.001,n = 46) of pregnant women eligible to participate in the study had stopped smoking four weeks after contact with the stop-smoking service intervention, n = 97 of the women who were referred had stopped smoking four weeks after contact with the stop-smoking services compared to n = 86 of the women who were referred in 2010/2011 post-intervention (see Table3).No data were available on the secondary outcomes and women's experiences of interventions and of attending smoking cessation support.

TABLE 2
Risk of bias assessment

TABLE 3
Resultsshowing referral rates and uptake of external stop-smoking support 2293 †Includes women already in touch with the stop-smoking service.

Already in touch with stop- smoking services, n (%) Uptake of stop- smoking services, n (%) Four-week cessation rate of pregnant women eligible to participate in the study, n (%)
2018, a systematic review published in The Lancet estimated