Stepped-wedge cluster randomised trial of a smoking cessation counselling training programme for midwives treating women with functional health illiteracy and low socioeconomic status (PROMISE): a study protocol

We have selected three regions for implementing the PROMISE protocol. Eligible regions (1) had at least 250 smoking pregnant women with low SES in the previous year; (2) were willing and capable to implement PROMISE within its obstetric partnership; (3) had a well-organised level of cooperation between midwives, OB-GYNs, post-natal care nurses, and youth health care; and (4) were not participating in any other tobacco-related research projects.

After receiving informed consent from midwives and OB-GYNs, we will randomly assign their practices to one of four steps. We will follow a training and cross-over procedure in accordance to our stepped-wedge study design. At baseline, all clusters of practices will use the existing V-MIS protocol for supporting women to quit smoking (i.e. control group). Two months after baseline, we will train the first cluster of practices (i.e. step 1) and will provide them with the PROMISE materials. From that moment onwards, this first cluster will cross over to the experimental condition and will continue to use the new PROMISE protocol. We will train clusters 2, 3, and 4 subsequently, each at a 2-month interval. After training, these clusters also cross over to the experimental condition. This way, we will train all practices in 4 steps over a 10-month period and all practices will be using the PROMISE protocol from month 8 onwards. By gradually implementing the PROMISE protocol among clusters of practices, we will replicate a real-life implementation of protocols that require face-to-face training. The order at which practices will receive the face-to-face training will be randomised at the practice level and stratified by region. We had no inclusion or exclusion criteria for the midwives and OB-GYNs, other than being employed at one of the practices enrolled in our study. At the start of the study (January 2017), participating midwives will complete a baseline questionnaire (T0) about their level of support to smoking pregnant women. They will receive a similar questionnaire at each step of the study (T1–T4) and will receive a final questionnaire 2 months after the final step (T5). These questionnaires include our main outcome variables. Figure 1 provides a schematic overview of the training steps. Care professionals will receive no incentive for filling out the questionnaires, but those who complete the training will receive continuing education credits.

Post-natal care nurses and youth health care professionals will be offered a face-to face training too. Contrary to the midwife training, their training will only focus on preventing relapse after pregnancy. As the performance of post-natal care nurses and youth health care professionals will not affect the main outcome of this study, the order at which post-natal care nurses and youth health care professionals receive training will not be randomised stratified by region. Instead, their training will be provided at a first come, first served basis. Post-natal care nurses and youth health care professionals will receive a baseline questionnaire before the training and one follow-up questionnaire 12 months after training. These questionnaires will be used for evaluating the materials and improving future materials for these professionals. Participants will receive no incentive for filling out the questionnaire, but those who complete the training will receive continuing education credits.

Pregnant women will be recruited through the midwifery practices and hospitals. Eligible women (1) are 18 years or older, (2) have been 0–3 months pregnant, (3) are current smokers or have quit smoking within the 6 months before pregnancy, and (4) have been registered in a practice or hospital that takes part in PROMISE. We will have no exclusion criteria. Midwives and OB-GYNs will provide recruitment materials to potential participants and inform them about the study. After providing written informed consent through their midwife of OB-GYNs, participants will receive a baseline questionnaire after 8 weeks of pregnancy and follow-up questionnaires at 2 weeks after delivery (T1) and at 6 months after delivery (T2). In those questionnaires, we assess smoking status and ask about the support they received from their midwife or OB-GYNs, post-natal care nurse professional, and youth health care professional. Women will receive a gift voucher of 10 euros for each questionnaire they complete.

This study was conducted from January 2017 to December 2019. All professionals have participated from January 2017 onwards, and the first training session has begun in March 2017. Pregnant women were able to enrol until March 2018. Among care professionals, we collected data from January 2017 to December 2017 and among pregnant women from January 2017 to December 2019.

Midwives assess their client’s smoking status, number of cigarettes per day, and whether the client is motivated to quit within 30 days. Smoking status is registered in the online system.

If a client is not motivated to quit smoking, midwives use motivational interviewing techniques to strengthen their client’s intrinsic motivation to quit smoking. They also discuss short-term advantages of quitting and disadvantages of smoking.

If a client is motivated to quit, midwives discuss barriers to successful quitting. Together, the client and midwife discuss how to mobilise social support for quitting.

Together, the midwife and client set a quit date. Midwives register the quit date in the client’s records and discuss the quit attempt during the next session.

While setting a quit date, the midwife and client discuss whether the client needs additional support. Midwives may provide self-help materials or refer the client to a specific type of behavioural support (i.e. telephone counselling, personal coaching, or group counselling). The use of nicotine replacement therapy is not encouraged in the Netherlands among pregnant women and may only be used among specific groups of heavy smokers [24].

If a client has quit smoking, midwives continue to discuss smoking (cessation) in the following visits in order to prevent relapse. They discuss past moments of craving and ways to deal with them. If a client has relapsed, midwives try to motivate their client to quit smoking again.

Midwives could help prevent relapse after delivery by discussing this beforehand with their clients. In the last stages of pregnancy, the client and the professional discuss which barriers and difficult situations the client may encounter. The midwives stimulate their client to develop strategies for overcoming these barriers. They also discuss the dangers of second-hand smoke to the baby.

We added the use of carbon monoxide (CO) measurements to steps 1 and 6 of the V-MIS protocol. A carbon monoxide meter measures the CO level in someone’s expired air and can be used to determine smoking status during pregnancy [37]. While CO levels continue to be elevated for at least 24 h after smoking a cigarette, CO meters are able to determine smoking status even if someone abstained from smoking a few hours prior to measurement. CO measures offer a non-invasive way to show pregnant women how much CO potentially reaches her unborn child. CO measurements are mandatory in smoking cessation care for pregnant women in some countries [38], but they are rarely used in Dutch midwifery practices. This study is the first one to explore the use of CO measurements among pregnant women in the Netherlands.

We will train midwives to use the CO measurement as a conversation starter and to avoid using the measurement outcome in a disapproving or criticising way. They will use it to stimulate smoking cessation (step 1) or staying smoke free (step 6). We will provide specific PROMISE score forms illustrating how CO and nicotine from cigarettes reach the unborn baby. Midwives could use this to communicate the health effect of smoking to women with functional health illiteracy. In line with previous research [37], midwives will be trained to use a cut-off point of 4 part per million to identify smokers among their clients.

We developed storyboard leaflets to help midwives enhance motivation among women with functional health illiteracy (steps 2 and 3), to guide these women towards smoking cessation (step 5), and to prevent relapse after quitting (step 6). Storyboard leaflets are visual stories, accompanied by a small text. By using simple texts and illustrations, information can be made accessible and comprehensible to both people with either poor or good literacy skills [39]. Storyboard leaflets stimulate doctor-client communication and have proven useful in other health-related domains, such as diabetes and obesity [40, 41]. They may be used for counselling for both people with poor literacy skills and for people with good literacy skills [40].

In an exploratory study, we interviewed 23 pregnant women (of whom most had a low SES) and asked them to help developing the story line for the leaflets. Their feedback made the texts and illustrations highly suitable for women with poor functional health literacy skills. Based on the target group’s feedback, we developed three versions of the same storyboard: one for pregnant women with a partner, one for pregnant women without a partner, and one for women who recently gave birth (see Fig. 3 for example images). Midwives will use the first two versions; post-natal care nurses and youth health care professionals will use the third one.

In the original V-MIS protocol, midwives were only able to refer clients to intensive smoking cessation counselling in step 5. In the PROMISE protocol, they will be able to do so from step 2 onwards. By allowing them to refer women earlier on in the protocol, they are able to help clients even if they lack time to counsel them themselves. As midwives are the main care provider for pregnant women, we will stress that midwives should only refer clients if they see no other option. To stimulate the use of the referrals in such cases, we developed an online list of local referral options for each of the three regions (a social map).

Since most women who quit smoking during pregnancy relapse after delivery [9, 20, 32], relapse prevention is extremely important. Unexpected stress factors may arise after delivery and could increase the chance of relapse [42]. While midwives and OB-GYNs visit their clients 1–3 times after delivery, relapse prevention has been a small part of the original V-MIS protocol. Post-natal care nurses spend more time with clients after delivery (8 days, 3–10 h per day), and mothers visit the youth health care clinic about 15 times in the first 4 years of their children’s life. Both post-natal care nurses and youth health care therefore play an important role in guiding new mothers and preventing relapse after childbirth.

Although both types of professionals do not influence the main study outcome, we felt a strong moral obligation to help these professionals too and to allow continued smoking cessation support after delivery. This is why post-natal care nurses and youth health care professionals will be offered both a face-to-face training and a counselling manual too. Contrary to the training and manual for midwives, their training and manual will be focussed at preventing relapse after delivery (instead of stimulating smoking cessation). We also developed specific storyboard leaflets about remaining smoke free after delivery. Post-natal care nurses and youth health care professionals could use this to support young mothers in remaining abstinent.

To further stimulate cooperation between midwives, OB-GYNs, post-natal care nurses, and youth health care, we developed a new transferring document. Health professionals could use this form to exchange information about the smoking status and smoking cessation.

To stimulate the correct use of the PROMISE protocol, all midwives, OB-GYNs, post-natal care nurses, and youth health care professionals will be offered a comprehensive face-to-face training. Although some midwives and OB-GYNs might have received a training in the V-MIS protocol years before this study, the materials used in that training were outdated. That is why we developed a new training with newer materials.

The training for midwives and OB-GYNs will consist of four parts. First, they receive background information about smoking cessation and the PROMISE protocol. Next, they will receive background information about people with functional health illiteracy. This section includes ways of recognising whether a client has poor literacy skills and techniques for communicating with them. One such technique is the teach-back method, in which a health professional checks whether a client understands the professional’s recommendations by asking the client to summarise the professional’s recommendations in their own words [43]. Third, we will teach motivational interviewing techniques to strengthen communicating with low SES and functionally illiterate women and their partners [44]. Participants will practice example conversations and discuss past challenging conversations with clients. Finally, participants will be trained in using the carbon monoxide meter.

The training for post-natal care nurses and youth health care professionals will only include the background information and the motivational interviewing techniques. Both post-natal care nurses and youth health care professionals will not use the carbon monoxide meter. By leaving out information about low health literacy, we avoid making the post-natal care nurses uncomfortable, as they often have poor literacy skills themselves.

All training participants will receive a comprehensive counselling manual. This manual discusses the V-MIS counselling protocol and the elements we added. It provides information about people with poor functional health literacy skills and ways of using the carbon monoxide meter. It includes a large number of example conversations about common barriers, difficult situations, withdrawal symptoms, and ways of mobilising social support. All example conversations are based on motivational interviewing techniques. We have developed three manuals: one for midwives and OB-GYNs, one for post-natal care nurses, and one for youth health care.

Professionals and pregnant women will be asked to fill out brief questionnaires. Midwives and OB-GYNs will fill out six questionnaires: one at baseline, four after each step (i.e. at each 2-month interval of the stepped-wedge design), and a sixth one 2 months after the last training. Professionals in both post-natal care and youth health care will fill out two questionnaires: one before they are trained with PROMISE and one at the end of 2018 (approximately 12 months after training). Pregnant women will fill out three questionnaires, containing questions about the smoking cessation care they receive during and after pregnancy. These questionnaires could be filled out online, via telephone, or on paper. An overview of the data collection among midwives can be found in Table 1 and Fig. 4.

Our primary outcome measure is change in percentage of smoking pregnant women with whom midwives (or OB-GYNs) discussed specific quit-smoking strategies. We use the self-reported data from the questionnaires filled out by the midwives and OB-GYNs for this. We will calculate this outcome measure by dividing the number of women with whom the professional discussed specific smoking cessation strategies by the total number of smoking women seen by the professional. This will result in a percentage at each time point for each individual professional. We expected that this percentage will increase over time in the total sample, as more professionals will use the PROMISE protocol and will be better able to support smoking clients.

We will use a Generalised Linear Mixed Modelling analysis in SPSS 25+ or R 3.0+ to analyse sample-wide changes in our primary outcome over time. In stepped-wedge designed studies, data should be interpreted at population level [34]. This means that not individual changes, but sample-wide changes over time are used to test the effectiveness of the new protocol (similar to a repeated cross-sectional design). In our study, all care professionals fill out a questionnaire at each of the six time points. In this questionnaire, they indicate the number of smoking women they have seen in the past month and the number whom they provided detailed smoking cessation counselling.

We calculate the percentage of smoking women who received detailed counselling for each care professional at each of the six time points. As each care professional will attend one of four training sessions, all professionals switch from the usual care condition (control) to the experimental condition at one of those time points.

Using a Generalised Linear Mixed Modelling (GLMM) analysis, we will analyse whether professionals provided detailed counselling to a higher portion of their smoking clients after receiving the training than before. We will include the factors ‘cluster’ (i.e. practices) and ‘individual’ (i.e. midwife/OB-GYN) as random effects, and ‘step’ will be included as fixed (time) effect. The exact link function (e.g. nominal, binominal) for the model will be based on the outcome variable. We will conduct an intention-to-treat analysis to control for lost to follow-up. Only members of the research team will have access to the research data.

As we do not anticipate any adverse effects in any of the two intervention arms, we will not conduct any interim analyses to determine if the trial needs to be stopped.