Brief intervention to reduce risky drinking in pregnancy: study protocol for a randomized controlled trial

This project focuses on the reduction of risky alcohol consumption during pregnancy. Risky drinking refers to a level of consumption that increases the likelihood of health problems (hazardous drinking), or that has already caused such problems (harmful drinking), and a pattern of high-intensity drinking associated with intoxication or drunkenness (binge drinking) [1]. Around a third of women drink more than the medically recommended levels [2]. Although older and more affluent women tend to drink more frequently than younger and less affluent women, the latter tend to binge drink (around a fifth of women aged 16 to 44 years binge drink [2]). Binge drinking is strongly associated with sexual risk-taking and unintended pregnancy [3]. Around a third of pregnancies are reported as unintended, but this can be as high as 70% for single women [4] and 82% for those living in impoverished, urban communities [5]. Thus a large number of pregnancies, particularly those in women from lower socio-economic positions, are likely to be significantly related to alcohol [6].

Although most women report that they abstained or reduced alcohol consumption during pregnancy, 8% of recently pregnant women report exceeding the medically recommended guidelines for low-risk drinking in pregnancy [7]. Of the estimated 723,000 births registered in England and Wales in 2010 [8], about 58,000 may have been significantly exposed to alcohol. The teratogenic effects of heavy alcohol consumption are well known [9]. Experts agree that there is a dose-dependent effect of alcohol on fetal and child development [10]. In addition, the timing of alcohol exposure is crucial [9]. Both the first and third trimester of pregnancy in humans are vulnerable periods for central nervous system development [9, 11, 12]. Thus, although intervention before or very early in pregnancy gives the best opportunity of reducing harm to the developing fetus, reducing alcohol consumption at any stage in pregnancy will have beneficial effects. The likely under-reporting of fetal alcohol effects in the admissions data of UK hospitals makes it unclear whether these data are similar to the higher rates in deprived communities reported from the USA [13‐15].

Although recent UK government strategy aims to raise awareness of health risks around alcohol use during pregnancy [16], there is currently no formal program of screening or alcohol intervention during antenatal care in England. Although most women recall alcohol being mentioned by midwives, this seems to have little effect on their drinking behavior, in part because women perceive that they often receive conflicting or unclear messages from health professionals [7]. There is therefore a need for clear, consistent, and effective alcohol advice in antenatal care. A large and robust evidence base supports screening and brief alcohol intervention in other populations. In this secondary preventive strategy, based on social cognitive and social learning theories, risky drinkers receive short advice or counseling, focusing on both personal and contextual factors to promote reduced drinking [17, 18]. A recent Cochrane Collaboration meta-analysis identified 29 randomized controlled trials (RCTs) in primary care, and found consistent positive effects of brief interventions compared with control conditions in terms of weekly reductions in alcohol consumption [19]. However, the authors noted that there was a lack of data relating specifically to women. A more recent Cochrane Collaboration review of pre-pregnancy health-promotion interventions identified just one alcohol-specific trial [20], which was based in the USA [21]. Although positive effects were reported for brief intervention, these disappeared in the sensitivity analysis [21]. Thus, brief intervention research focused on antenatal care is urgently needed.

Five brief intervention trials have been conducted with pregnant women in the USA [22‐26]. Two trials were small (pilot) studies [23, 25], with assessment and/or brief intervention procedures that were much longer in duration (1 to 2 hours) than the time available in antenatal care in the UK [23‐25]. Four trials were based in obstetric settings [22‐25], and three used specialist practitioners to deliver brief intervention [22, 24, 25]. Two trials included a single session of brief intervention, one of 10 to 15 minutes [26] and the other of 25 minutes in duration [23]. The former reported significant effects of brief intervention [26], whereas the latter reported reduced drinking in both the control and intervention groups [23]. Finally, one study found that women in antenatal care receiving brief interventions were significantly less likely than women receiving no treatment to have an infant with low birth weight or to suffer preterm labor problems, while another study found positive effects of brief intervention on prenatal drinking in women with higher initial rates of consumption [22, 23]. Most recently, a trial of brief interventions in antenatal clinics in South Africa found that women who received a brief intervention had significantly greater reductions in scores on the Alcohol Use Disorders Identification Test (AUDIT) screening tool [27] by their third trimester, compared with women receiving assessment only [28]. Taken together, this evidence suggests that brief intervention is a promising approach for alcohol risk reduction in antenatal care. However, it is necessary to establish if evidence from the USA and elsewhere is applicable to a UK context with marked cultural and health-system differences.

New guidance from the Medical Research Council on developing and evaluating complex interventions [29] is intended to help institute appropriate methods in such research and to enhance the usefulness of its evidence. The guidance points out the potential for evaluations of complex interventions such as brief interventions to be undermined by problems that could be highlighted in a pilot study. Piloting is therefore recommended, but should not be used to test a hypothesis [30]. The pilot study should establish the acceptability of procedures, identify likely recruitment and retention rates, and inform a sample-size calculation for a definitive trial [29]. Because brief interventions to reduce alcohol consumption have not been tested previously in UK antenatal care, a pilot study for such research is proposed on this basis.

Findings from this pilot study will indicate whether and how a definitive trial can establish the effectiveness and cost-effectiveness of screening and brief intervention to reduce heavy drinking among pregnant women in the UK. The outcomes will include the protocol for such a trial, with a sample-size calculation, which can usefully extend the evidence base in this field at an international level.

Screening with brief interventions in healthcare settings is increasingly recognized as an effective approach to improving public health, and is prioritized in the UK government’s recent alcohol strategy [16]. However, knowledge of how effects vary across models and modes of delivery, populations, and settings is limited [51]. Several features of this research are noteworthy in this respect. Alcohol screening and brief intervention involves communicating relatively long-term risk to an individual’s own health, and advising that individual on how to bring their drinking in line with recommended guidelines to reduce that risk [18]. By contrast, brief intervention against heavy drinking in pregnant women seeks to achieve change to a lower level of consumption than the universal guidelines to reduce short-term risk to the pregnancy; even a change during the limited period of pregnancy and early infancy can be considered a successful outcome in this respect. Furthermore, alcohol use is a particularly stigmatized issue to discuss with pregnant women [52‐54]. For this reason, compared with the wider population, pregnant women may be more anxious for guidance [55], or more reluctant to engage in talk about alcohol, and these attitudes may affect their willingness to participate in a trial. Finally, much of the evidence base for brief interventions comes from primary healthcare settings. Community midwives have a distinct role and relationship from other health professionals in that they are not treating diseases or disorders, but facilitating the processes of pregnancy and childbirth. However, there are considerable constraints on their time with each patient. Findings from this study will contribute to the broader understanding of screening and brief intervention and related research by indicating how the trial of an intervention focused on pregnancy is likely to be received by women and midwives in the novel healthcare setting of community antenatal care.