The conversation matters: a qualitative study exploring the implementation of alcohol screening and brief interventions in antenatal care in Scotland

Abstract

Background

Alcohol screening and brief intervention (SBI) in antenatal care is internationally recommended to prevent harm caused by alcohol exposure during pregnancy. There is, however, limited understanding of how SBI is implemented within antenatal care; particularly the approach taken by midwives. This study aimed to explore the implementation of a national antenatal SBI programme in Scotland.

Methods

Qualitative interviews were conducted with antenatal SBI implementation leaders (N = 8) in eight Scottish health boards. Interviews were analysed thematically and using the ‘practical, robust implementation and sustainability model’ (PRISM) to understand differences in implementation across health boards and perceived setting-specific barriers and challenges.

Results

In several health boards, where reported maternal alcohol use was lower than expected, implementation leaders sought to optimize enquires about women’s alcohol use to facilitate honest disclosure. Strategies focused on having positive conversations, exploring pre-pregnancy drinking habits, and building a trusting relationship between pregnant women and midwives. Women’s responses were encouraging and disclosure rates appeared improved, though with some unexpected variation over time. Adapting the intervention to the local context was also considered important.

Conclusions

This is the first study to explore implementation leaders’ experiences of antenatal SBI delivery and identify possible changes in disclosure rates arising from the approach taken. In contrast with current antenatal alcohol screening recommendations, a conversational approach was advocated to enhance the accuracy and honesty of reporting. This may enable provision of support to more women to prevent Fetal Alcohol Spectrum Disorders (FASD) and will therefore be of international interest.

Background

Alcohol use in pregnancy can cause harm to the developing fetus, including growth restrictions, low birth weight, pre-term birth, and fetal alcohol spectrum disorders (FASD) [1,2,3,4]. Recent estimates suggest that 41.3% of women in the UK consume alcohol at some point during pregnancy, among the highest in the World Health Organization (WHO) European Region. Furthermore, the UK also has a high estimated Foetal Alcohol Syndrome (FAS) prevalence, which is the most severe form of FASD [5]. Identifying women who drink during pregnancy, and providing information and effective support, is therefore of public health importance.

International guidelines recommend that health professionals screen all pregnant women for alcohol use and provide an intervention to those who drink, supporting behaviour change [6]. Screening and brief intervention (SBI) consists of a short conversation focused on identifying problem drinking, motivating and facilitating reduction in drinking or abstinence to reduce the risk of harm [7]. SBI is typically intended to be delivered by a generalist health professional to patients who are not seeking treatment for alcohol problems [8]. The evidence for SBI effectiveness in reducing drinking among adults in primary care settings is relatively strong [9, 10], though not unproblematic [11] and may be less applicable to women [9]. In antenatal care, systematic reviews have cautiously supported SBI efficacy for reducing alcohol consumption during pregnancy, although few relevant studies have been published. High risk of bias and complexity of interventions contribute to important uncertainties regarding efficacy in this setting [12, 13]. Despite this evidence gap, WHO guidelines note that SBI benefits are likely to outweigh any potential adverse effects [6], justifying implementation in antenatal care.

In Scotland, clinical guidelines have highlighted antenatal care as an important setting for SBI delivery since 2003 [14]. In 2008, this was formalised as a national programme by setting a target for “alcohol brief intervention” (ABI) delivery in three priority settings, including antenatal care. ‘ABIs’ included screening, and the term can be considered synonymous with SBI [15]. A national training programme, practitioner materials, and significant funding including for specialist alcohol services accompanied the target. Official drinking guidelines for pregnant women accompanied the national SBI programme. At the time the national target was first introduced, the advice was not to “drink more than 1–2 units of alcohol once or twice per week” and not to “get drunk” [16]. This changed to an abstinence-focused message in 2010 [17]. The 14 health boards (regional health providers in the UK National Health Service) are obliged to report to the Scottish Government quarterly on their progress in delivering SBI. Within this programme, all midwives are expected to be trained, and should screen all pregnant women.

The ambitious scale, resourcing, and profile of the national SBI programme was unprecedented in the UK, comparing only with a few initiatives internationally. Evidence from the Swedish ‘Risk Drinking Project’ show that educational efforts led to improved midwife knowledge and competence in identifying pregnant women defined as having at-risk consumption patterns. However, the project had a uniform approach to addressing alcohol across primary care, occupational health care, and child and maternity care, and few details about views on implementation or adaptations have been published [18].

To date, the Scottish national SBI programme, which could provide a useful model in other jurisdictions, has not been extensively evaluated. Evaluations to date have focused on SBI delivery in primary care [19] and youth settings [20]. The aim of this paper is to explore implementation leaders’ experiences of incorporating SBI into routine practice in antenatal care under the Scottish national SBI programme. Specifically, our research question was how local health boards adapted, implemented and experienced the national SBI programme in antenatal care.

Methods

Study design and sample

A qualitative design was chosen to explore local health professionals’ experiences of implementing SBI in antenatal care. Participants took part in in-depth interviews, conducted by telephone to accommodate their clinical commitments, there being no good evidence of the superiority of face-to-face interviews [21]. This paper draws on secondary analysis of data from a wider study of the national SBI programme led by NF. The methods have been published in full in line with COREQ and RATS guidelines [22, 23], where they are described in detail [24].

Fourteen key people who worked as local implementation leaders were purposively recruited for the original study [25]. Sampling included leaders with experience from high-performing as well as low-performing health boards, defined as above or below the median of SBIs delivered in antenatal care. Of those sampled, one senior midwife who had initially agreed to take part was not contactable and did participate in the original study. Eight of the original 14 participants were responsible for implementation in the antenatal setting and are included in this paper: 6 of these were also responsible for SBI delivery in other settings. Participants worked as specialist midwives/nurses, Alcohol and Drug Partnership (ADP) coordinators, and a senior public health doctor with both clinical and strategic experience (Table 1). The original study obtained ethical approval from the Ethics Committee of London School of Hygiene and Tropical Medicine.

Table 1 Interviewee characteristics

Data collection

Identified individuals were contacted via email and invited to take part in a telephone interview. NF conducted the interviews between September and November 2013. The interviews were semi-structured and used a pre-circulated topic guide [21]. Participants were encouraged to speak freely about the SBI implementation experiences in their health board. Interviews were audio recorded and complemented by notes taken at the time of the interview. Participants verified interview notes and transcripts through member checking, with opportunity to add or clarify the interview. In accordance with ethical approval, and to reduce burden on participants from having to separately return written consent forms, all interviewees provided audio-recorded, fully-informed, formal, verbal consent and were reassured of confidentiality.

Data analysis

All interview transcripts from the full study were reviewed for data relating to antenatal care, resulting in a final dataset of implementation leaders from eight of the eleven Scottish mainland health boards. These interviews lasted an average of 74 min each. These transcripts were subject to a detailed analysis. Both authors read the eight relevant interview transcripts several times to gain familiarity with the data. LS undertook initial inductive coding, which was discussed, and the codes were organised thematically, by NF and LS [26], the practical robust implementation and sustainability model (PRISM) was also used to organise the findings. The PRISM analysis focused on conceptualising implementation in relation to the recipients, intervention, organizational factors and external context [27].

Results

Approaches and strategies taken to implement SBI in antenatal care varied between health boards. Table 2 outlines implementation by the four PRISM areas (recipients, program, infrastructure and sustainability, and external environment). Local structures and factors, midwives’ attitudes towards women's drinking habits in and outwith pregnancy caseloads, and time available for training were important factors for implementation. The following sections describe implementation in relation to integration into routine practice, perspectives on screening, contextual factors, and perceived outcomes. Two case studies are included to illustrate the reported impact of using different approaches to asking pregnant women about alcohol consumption.

Table 2 Findings organised by PRISM domains

Integrating SBI into routine practice

Participants noted the importance of senior management support in the implementation process, but this was not always available. For example, in Health Board B, there was “no buy-in from senior people in antenatal”. On the other hand, strong support from Head of Midwifery was instrumental in progressing the programme in several health boards.

It is true that I was starting from a lower base in terms of relationships. I didn’t have a strong link into antenatal settings [ … ] but there were good links with Head of Midwifery and it was made clear to midwives they had to do it. We needed to use that strategy to influence delivery in maternity so it was more of a top-down approach (Health Board C)

In order to report to Scottish Government, implementation included a focus on recording, integrating screening questions into existing electronic patient record systems, or (in one case) development of a new paper-based system. This influenced the construction of protocols for delivering SBI and referral to specialist services (see Table 2). More fundamentally, however, participants had to determine the intervention target group following the screening process. This was clear in Health Board F, where it was evident that midwives own attitudes did not match up with guidance at the time in a pamphlet given to women, where guidance was to limit intake to one to two units once or twice per week.

A lot of time (was) spent debating about exactly which women would we actually be delivering a BI [brief intervention] to, would it be women who were drinking above the 14 units limit [ … ] Was it women drinking more than the ‘Ready Steady Baby’ limits? Or was it actually what midwives felt strongly professionally, which was women who were actually drinking any alcohol at all in pregnancy? (Health Board F)

Two participants described how midwives’ attitudes towards pregnant women’s alcohol use influenced how SBI was implemented and designed locally. Whilst midwives were in agreement about total abstinence during pregnancy, the national guidance and training materials did not include a clear abstinence message, in contrast to the abstinence-focused approach taken in most of the local areas.

The national packs were useful after we had done the training for trainers but on the back of the work that was done, the initial antenatal packs said that women didn’t need an ABI [alcohol brief intervention] if drinking small amounts, but now anybody drinking in-pregnancy gets an ABI (Health Board D)

Screening in the antenatal setting

There was no consensus on the best way to identify pregnant women who were drinking alcohol. Perceived feasibility, including required time, influenced the screening tool used or approach taken. For example, the TWEAK test [28] was used in two health boards, but was regarded too time consuming in another health board.

People felt that TWEAK was too much and they were trying to incorporate it into the initial booking appointment where there are a lot of questions to work through in an hour. So we thought the simplest thing to do was to stick with questions that were already there in the SWHMR [Scottish Women’s Handheld Maternity Record] (Health Board C)

Several other health boards also decided to limit change in current practice by using questions from existing standard forms. In health boards where standardized tools were used, considerations to their application to the local context was considered important (see Table 2).

In several areas, reported alcohol use elicited through standard questions was lower than expected. Implementation leaders’ knowledge of local drinking culture led them to conclude that drinking levels being reported in pregnancy were not accurate.

When you look at the [local] culture of drinking and hazardous drinking among women and in the population in general, we don’t think that less than 1% of women are drinking in pregnancy (Health Board C)

In the two case study areas (Table 3), discrepancies led to consideration of how to approach screening. It was clear that implementation leaders felt that screening questions had to flexible and not simply asked verbatim of each woman. In Case Study 1, focusing on the context of alcohol consumption was considered an effective strategy to improve reporting levels, and influenced disclosure rates in some cases. This was seen as critical for offering help to women who might benefit from cutting down, reducing the risk to the fetus in the current, and potentially future, pregnancies. Ensuring that midwives and pregnant women were comfortable with alcohol questions was important and meant adapting questions to local (not formal) language. In both case studies, additional prompts and questions to encourage trust and overcome defensive responses were key. Emphasising pre-pregnancy drinking was a strategy to identify, and therefore provide effective support to those who might benefit, which was also used by other participants.

We had a lot of discussion about it being more important to ask about alcohol consumption before pregnancy, because pregnant women are less likely to disclose when they are drinking in pregnancy because they know they are not supposed to (Health Board H)

Case Study 2 indicated an increase in reported pre-pregnancy abstinence over time, felt likely due to a change in the accuracy and honesty of women’s reporting, rather than a genuine fall in consumption. One interpretation was that women were ‘coming prepared’ to answer the questions. Another was that a recent focus by midwives on asking about parenting capacity and home circumstances may have made women fearful about disclosing heavy drinking (see Case Study 2). In this case, midwives were encouraged to probe further if women reported no alcohol use pre-pregnancy, resulted in higher levels of disclosure.

Table 3 Case studies from local areas

Contextual factors affecting implementation

Wider maternal health and antenatal care policy agendas were important for success in implementing SBI. Several respondents mentioned that a focus on alcohol fitted with broader national efforts around early interventions for child wellbeing. This included the Getting It Right For Every Child (GIRFEC) agenda, aimed at improving health and wellbeing for children and young people in Scotland through timely support [29]. Participants however highlighted that the SBI programme did not necessarily align with GIRFEC or other relevant lifestyle and health agendas.

There were lots of different health improvement people going to the same target staff about different things to do with [how they address] lifestyle change etc. All these different approaches are being made to midwives and practice nurses or whatever separately – it’s not joined up (Health Board B)

Participants noted that this led to duplication of training, as addressing other lifestyle issues require similar skills.

There was a concern that there was not really a joined-up-ness about all of this. That people were being asked to be trained for talking about breastfeeding, looking for issues of domestic abuse, issues of smoking and behaviour change, and alcohol, but where was the joined up bit about it? Where could we capitalise on the shared skills, the crossover skills? (Health Board F)

A joined-up approach was pertinent considering that training midwives was a major task in many areas, requiring annual training of new trainee midwives and staff. Further, the number of health behaviours to cover in booking appointments was seen as increasing midwives’ workload and a burden for women. This appeared to create some resistance.

All these things were coming at the same time and setting an agenda that for midwives, and frankly for women coming for booking, was becoming too huge (Health Board F)

Perceived outcomes of the SBI programme

Participants perceived that introducing SBIs had positive outcomes, including consistency in asking all women about alcohol, increased FASD awareness among pregnant women, and reinforcing existing midwife practice through improved guidance on facilitating the conversation. Screening rates were however low in many health boards, meaning midwives delivered few SBIs. Even where there were higher screening rates, the reported prevalence of drinking in pregnancy was often low. Several participants reported that midwives believed they would already know of a woman’s drinking problem and in at least one area, the implementation programme failed to overcome this reservation.

Midwives were not particularly happy with it, their reservation was that if somebody had a significant problem they would already be known and if they didn’t have that level of problem but were drinking, they were unlikely to tell you, the others who were happy to talk about it had already reduced or stopped drinking anyway … we had to accept what they were saying. All we could do was offer more follow-up support and refresher training, which no-one accepted (Health Board B)

Discussion

This study explored the implementation of a national SBI programme in antenatal care in Scotland. We found number of barriers and facilitators to implementation, echoing previous research showing that open discussions are impeded by the topic’s sensitive nature [30], lack of an established relationship at booking [31], fear of judgement [32], and fear of child protection issues and involvement of social services [33]. Implementation leaders used several strategies to facilitate honest disclosures including positive conversations, exploring pre-pregnancy drinking habits, and building a trusting relationship between pregnant women and midwives. Women’s responses were encouraging and disclosure rates appeared improved, though with some unexpected variation over time. These findings can inform future SBI programmes.

The national SBI programme guidance suggested screening all pregnant women using a validated screening tool [34], which the WHO also recommends [6]. Formal screening instruments can facilitate discussion about alcohol [35]. For example, midwives in Sweden used the Alcohol Use Disorders Identification Test (AUDIT) as a “pedagogical tool” where conversations about current drinking were built on screening of pre-pregnancy consumption levels [18]. Our study found no universal adoption of a validated screening tool across health boards, as many adapted instruments to fit the local context. O’Brien [36] argued that universal application of SBI in antenatal care should be informed by evidence, but guidance should not specify a particular screening tool. Similarly, a recent literature review recommended development of national standards to facilitate SBI implementation, but made no recommendation on a specific screening tool best suited for maternal health services [37].

Several health boards emphasised the importance of a positive conversation and asking questions in a locally appropriate language: in a study from Norway, 61% of midwives reported that they would rather have a conversation with expectant parents without using a screening tool [38]. Furthermore an emphasis on building trust led to discussions of pre-pregnancy drinking behaviour, which was felt to be less stigmatised, and has been found to be an acceptable strategy [39]. Evidence that pre-pregnancy drinking levels predict continued alcohol use in pregnancy [40, 41] supports this approach.

Whilst screening adaptations appear to facilitate implementation, adaptation raises other questions: in one observational study in primary care, sensitivity was lost when health professionals made adaptations to prescribed screening tools [42]. Our case studies show that implementation leaders mandated a flexible approach, in order to build trust, and reported that it led to more frequent and/or more complete disclosures of alcohol consumption. The validity of informal adapted approaches merits further research, but it is also worth considering whether a more flexible approach may be valuable in other settings. McCambridge and Rollnick [43] argue for a more ‘patient-centred’ approach in primary care, to “encourage people with alcohol problems to tell us what their problems are, so that help can be provided to think these through”, and suggest this would distinguish face-to-face interventions from the simple, rigid, screening most commonly provided in electronic SBI.

One way of exploring the validity of flexible screening approaches is using biomarkers, which have been extensively studied in pregnancy, however evidence is insufficient to recommend routine use of currently available markers [44]. Recruitment bias and the lack of a gold standard reference test for in-pregnancy drinking impedes research in this area. With FASD being a leading preventable developmental deficit, innovative research and practice approaches are urgently needed to identify those who might benefit from support [45]. Combined self-report and biomarker methods have been utilised for identifying smoking in pregnancy [46]. For alcohol, cohort studies following up children’s outcomes after birth and through childhood, following biomarker testing combined with self-report screening in pregnancy, could provide further data to identify those most at risk in future [47]. Ideally, such screening would facilitate personalised feedback to women about the risk to their baby; a component of SBIs that appears to be important for changing behaviour [48].

This is the first indication that reporting rates of alcohol consumption in pregnant women may change over time, or be affected by other developments in the conversation before or after the alcohol questions are asked. Whilst Scotland-specific, the findings raise an important possibility that reduced disclosure of alcohol consumption may be an unintended consequence of a greater focus on child wellbeing/parenting readiness. This is likely to be relevant elsewhere. Current recommendations for implementing SBI by nurses and midwives tend to focus on structural and practical issues [37], with little discussion around the impact of contextual factors such as other policy agendas.

Our findings suggest that antenatal care may be a particularly sensitive ‘complex system’ in which interventions are influenced by policy agendas [49], and with feedback loops where over time women may be ‘prepared’ to answer in a certain way. Systems-informed evaluations of interventions in this setting that include consideration of unintended consequences are therefore vital [50]. Such evaluation should also consider whether an integrated approach to addressing alcohol and other public health topics in antenatal care could have helped, for example, by addressing cross-over skills, acknowledging the need to prioritise available time, ensuring that sensitive topics are not avoided, and addressing burden on staff. Finally, the drinking guidelines for pregnant women that existed at the time of implementation caused discomfort amongst midwives, who believed they should advise total abstinence. Several health boards therefore decided to offer SBI to any woman drinking in pregnancy rather than setting cut-off points for current drinking, an approach that was later reflected in the national programme. Implementation of a national programme therefore also needs to consider midwives’ own attitudes.

Strengths and limitations

This is the first study to explore in detail the experiences of SBI implementation leaders of a large-scale primary prevention programme to prevent harm caused by alcohol exposure during pregnancy in the UK. It adds to understanding of the detailed practical and ethical dilemmas involved in establishing alcohol SBI in the antenatal setting, and is likely to be relevant to other countries. Eight of the eleven mainland health boards in Scotland were included, providing an insight into the implementation process in a majority of areas. However, views in remaining health boards and island boards may differ, as may experiences in other countries, where local research would be valuable.

Conclusions

National resources, funding, and support from strategic, frontline and management staff were important for the implementation of SBI in antenatal care. A flexible, conversational approach to discussing alcohol with pregnant women was considered superior to formal tools, for identifying who might benefit from intervention. The approaches suggested could be implemented internationally and merit further study. Furthermore, national programmes should consider an integrated approach to health promotion in pregnancy in future, whilst recognising the potential for unintended consequences.