Increasing accurate self-report in surveys of pregnancy alcohol use

Increasing accurate self-report in surveys of pregnancy alcohol use

Midwifery 31 (2015) e23–e28 Contents lists available at ScienceDirect Midwifery journal homepage: www.elsevier.com/midw Increasing accurate self-re...

358KB Sizes 0 Downloads 15 Views

Midwifery 31 (2015) e23–e28

Contents lists available at ScienceDirect

Midwifery journal homepage: www.elsevier.com/midw

Increasing accurate self-report in surveys of pregnancy alcohol use Evelyne Muggli, MPH (Senior Research Officer) a,b,n, Brendan Cook, MPH (Research Assistant)a,c, Colleen O’Leary, PhD (Post Doctoral Research Fellow)d, Della Forster, PhD (Professor)e,f, Jane Halliday, PhD (Professor)a,b a

Murdoch Childrens Research Institute, Parkville 3052, Vic., Australia Department of Paediatrics, The University of Melbourne, 3010 Vic., Australia c Central Australian Aboriginal Congress, Alice Springs, 0870 NT, Australia d Telethon Kids Institute, Perth 6845, WA, Australia e La Trobe University, Melbourne 3000, Vic., Australia f Midwifery and Maternity Services Research, The Royal Women's Hospital, Parkville, 3052 Vic., Australia b

art ic l e i nf o

a b s t r a c t

Article history: Received 14 July 2014 Received in revised form 1 October 2014 Accepted 3 November 2014

Background: pregnancy alcohol research relies on self-reports of alcohol consumption. Reporting bias may contribute to ambiguous and conflicting findings on fetal effects of low to moderate pregnancy alcohol exposure. Objective: this study aimed to identify the determinants which would enable women to provide accurate data in surveys of alcohol use in pregnancy. Design and participants: six focus groups were held with a total of 26 pregnant women and new mothers. Participants reviewed a set of alcohol survey questions followed by a guided discussion. Transcripts were analysed using inductive content analysis. Setting: public hospital antenatal clinics and Mother & Child Health Centres, Melbourne, Victoria, Australia. Findings: women's emotional responses were generally favourable, although the potential for anxiety and fear of judgement was acknowledged. Barriers to accurate self-report were recall, complexity and use of subjective language. Facilitators were appropriate drink choices, occasional drinking options and contextualising of questions. Confidentiality and survey method, including a preference for methods other than face-to face, were also important factors. Key conclusions and implications for practice: questions embedded in clear context may reduce anxiety around questions about alcohol use in pregnancy. Methods using shorter recall periods, a list of drinks choices, measures of special occasion drinking and minimising complex and subjective language will increase accurate self-report. A setting perceived as confidential and anonymous may reduce a desire to provide socially acceptable answers. & 2014 Elsevier Ltd. All rights reserved.

Keywords: Antenatal alcohol exposure Pregnancy Ethanol Questionnaires Qualitative Research

Introduction The evidence is clear that heavy and chronic alcohol consumption in pregnancy is associated with adverse effects on the developing fetus and child, but the relationship between outcomes and low to moderate levels of pregnancy alcohol exposure (PAE), including occasional binge drinking remains ambiguous, even after systematic review and meta-analysis of studies investigating such an exposure (Makarechian et al., 1998; Polygenis et al., 1998; Armstrong and Abel, 2000; Testa et al., 2003, Henderson et al., n Correspondence to: Murdoch Childrens Research Institute, Royal Children's Hospital, 50 Flemington Road, Parkville, Vic. 3052, Australia. E-mail address: [email protected] (E. Muggli, MPH (Senior Research Officer).)

http://dx.doi.org/10.1016/j.midw.2014.11.003 0266-6138/& 2014 Elsevier Ltd. All rights reserved.

2007a; Henderson et al., 2007b; Dolan et al., 2010; Patra et al., 2011; Flak et al., 2014). Variable research findings following low and moderate PAE and occasional binge drinking are ongoing, with some studies reporting effects on child development (Sayal, 2007; Alvik et al., 2011; Larkby et al., 2011; Alvik et al., 2013; Nykjaer et al., 2014) and others showing no differences between alcohol exposure groups (Bakker et al., 2010; Robinson et al., 2010; Bay et al., 2012; Kesmodel et al., 2012; Underbjerg et al., 2012; Humphriss et al., 2013; Kelly et al., 2013; Skogerbo et al., 2013). Conflicting results have been described as relating to inadequate adjustment of contextual factors and/or a lack of accurate data on alcohol intake (Testa et al., 2003; Henderson et al., 2007a). A number of validated instruments exist to screen pregnant women in order to determine their use of alcohol. The focus of these instruments is identifying ‘at risk’ pregnancies and women

e24

E. Muggli, MPH (Senior Research Officer) et al. / Midwifery 31 (2015) e23–e28

with ‘problem drinking’ (Burns et al., 2010). Validated instruments to measure exact PAE, in particular at low and moderate levels, are not available and there are currently no objective measures such as biomarkers representing ‘whole of pregnancy’ exposure. Consequently, most studies to date have relied on self-reports of PAE. There is, however, little research into the actual willingness and ability to truthfully report alcohol intake and to identify the type of survey questions that support accurate disclosure of alcohol use by pregnant women. If the ambiguous and often conflicting results on fetal effects of low and moderate PAE are to be resolved, it will be particularly important to have accurate ascertainment of these particular exposures to ensure any purported outcomes are not influenced by reporting bias. Confidence in having high quality exposure data will lead to more meaningful and useful outcome data for advising women and developing policy related to all levels of alcohol consumption. Aim The aim was to identify the determinants which would enable women to provide accurate data in surveys of alcohol use in pregnancy. In doing so, we would understand the barriers and facilitators that impact on accurate reporting of alcohol use, and be able to obtain a comprehensive and measurable picture of a woman's past and current alcohol intake, even at the lowest level. This research was an important aspect of the pilot work leading to a major longitudinal cohort study of fetal effects of low to moderate alcohol exposure.

Methods To meet our aim, we designed a study to explore women's views about a set of 22 questions related to past and current alcohol consumption. A focus group design was chosen in preference to individual interviews as the diversity of group dynamics within focus groups has been shown to ‘stimulate participants to analyse their opinions more deeply’ (Liamputtong and Ezzy, 2005). Ethical approval Prior to commencement of the study, ethics approval was granted by the Barwon Health (#09/83), Royal Children's Hospital (#29107) and Royal Women's Hospital Human Research Ethics Committees (09/44); the Eastern Health Research and Ethics Committee (#E26-0910); and the Victorian State Government's Department of Education and Early Childhood Development Research Committee (Project Alcohol use in pregnancy: Can we get the questions right?). The Plain Language Statement contained information about the effects of alcohol use in pregnancy on the baby and details of an independent help line should women experience any anxiety or unforeseen concerns about their own alcohol consumption during pregnancy as a result of taking part. Participants Focus group participants were women 18 years and older who spoke English and resided in Victoria, Australia. The first category of participants were pregnant women recruited from public hospital antenatal clinics, with a singleton pregnancy and no known pregnancy complications at the time of recruitment (e.g. threatened miscarriage or fetal anomaly). The other category comprised new mothers with a singleton infant no more than 12 weeks old, recruited from maternal and child health (MCH) centres. This latter group were included to investigate potential recall or other issues affecting accuracy of reporting given the

retrospective nature of their considerations. Eligible women were approached by a member of the research team (BC) while they were attending either an antenatal clinic or a MCH centre. Women who were interested completed a form, indicating their availability to attend specified focus group dates and times and providing basic demographics, i.e. age group, parity, gestational age and level of education. Data collection Survey questions about a woman's history of alcohol use and her alcohol use in pregnancy were developed following a comprehensive review of the literature. Questions aimed to (a) estimate a pregnant woman's pattern of lifetime alcohol drinking; (b) identify changes in drinking behaviour since pregnancy recognition; and (c) accurately measure pregnancy alcohol exposure by dose and pattern for distinct time periods in pregnancy (e.g. from becoming pregnant but before pregnancy recognition; and from pregnancy recognition until 13 weeks (first trimester)) and (d) estimate family history of drinking and partner drinking. A drinks guide was designed to allow women to choose pictures of 33 different types of drinks without requiring them to calculate the number of ‘standard drinks’. There were pictures of beer and cider of varying serving sizes and strengths, two types of wine, plus port, sherry and champagne, again of varying serving sizes. There were also eight different types of spirits in the form of shots and mixed drinks. The guide was designed with a key based on the Australian Standard Drink, to convert drink choices into grams of alcohol consumed. Questions accompanying the drinks guide enabled tabulation of frequency and quantity for specific time periods, e.g. ‘Using the code(s) provided in the pictorial Drinks Guide provided, please complete the table below, including; (a) what type of drink(s) you typically drank in this period, (b) how often you usually drank this type of alcohol (for each), and (c) how many of these would you usually drink on each occasion (for each type of drink).’ Responses would allow for computation of detailed exposure patterns by means of an algorithm (O’Leary et al., 2010). After introducing the purpose of the study and obtaining written consent, the focus group moderator (BC) reiterated that the aim of the research was not to collect data on individual alcohol consumption, but to gather opinions about the alcohol questions. An open-ended ice breaker question was used to initiate discussion around women's own experiences of, and attitudes to being asked about alcohol use in pregnancy. Participants were then given the opportunity to read the questions, after which the focus group moderator followed a schedule to guide discussion, informed by the methodology of Krueger (1998a). The schedule included six key areas covering first impressions of the alcohol questions, whether participants felt they would respond honestly or could answer accurately, and specific feedback on the layout and wording of the questions and drinks guide. Participants' actual alcohol use was not collected by the researchers. The same focus group moderator and co-moderator (note-taker) (EM) conducted all focus groups. All discussions were audiotaped and transcribed verbatim by the focus group moderator. Participant names were replaced with pseudonyms; non-verbal communication, such as the degree of spontaneity and participant involvement, was transferred to the transcript from field notes where appropriate. Each focus group was reviewed for patterns; the discussions by the moderators and new insights and/or issues were progressively incorporated until only redundant information was found. Data analysis Basic demographic data were collected for each participant and described using frequencies and percentages. All transcripts were analysed using inductive content analysis (Elo and Kyngas, 2008).

E. Muggli, MPH (Senior Research Officer) et al. / Midwifery 31 (2015) e23–e28

Content analysis is particularly useful to systematically and objectively identify specific messages in any type of social communication. It establishes the existence and frequency of concepts through inclusion or exclusion of content according to consistently applied criteria relevant to the research aims. Analysis involved BC and EM repeatedly reading the transcripts, coding and annotating the text in the margins with headings which represented manageable content categories. A process of selective reduction then produced an agreed analysis matrix, which consisted of hierarchical flow charts and diagrams to pictorially represent each heading and any possible connections between them. Data were then abstracted into this matrix in a dynamic process by further reviewing and refining headings with similar responses. These formulated categories became the final framework used to report results. Each category was named using a term that was ‘contentcharacteristic’ (Elo and Kyngas, 2008).

Findings Twenty-six women attended one of six focus groups in November–December, 2009, with three being the smallest number of women per focus group and seven the highest (Table 1). Seventy-four per cent of participants were aged less than 36 years, 77% were primigravida, and 46% were new mothers with a healthy infant aged between 10 and 12 weeks. Most of the participants were university educated (73%), and almost a quarter (23%) completed their education at between nine and 12 years of schooling (i.e. high school level) (Table 1). The three main categories of our analysis matrix were (1) emotional responses to being asked about alcohol in pregnancy; (2) barriers to, and (3) facilitators of, providing accurate data in surveys of alcohol use in pregnancy. Between three and five subcategories remained necessary to describe each categorical concept in order to increase the understanding of it and to generate knowledge around the topic. The conceptual framework for understanding the factors which increase accurate self-reporting of pregnancy alcohol use is shown in Fig. 1 and discussed in detail below. Emotional responses Participant's emotional responses to being asked about their alcohol consumption during pregnancy consistently belonged to Table 1 Demographics of focus group participants.

Maternal age 18–25 26–30 31–35 36–40 41–45

FG1

FG2

FG3

1

1

1 1

1 1

2

1 1

Primigravida/para Yes No

3

Gestation 14–26 weeks 27–40 weeks New mother

1 2

Education High school University Trade

1 2

1 2

2 1 1

Number of participants

3

3

4

3

1 3

1 2

3 1

FG4

FG5

FG6

Total

%

1 2 1

3 2

2 4 1

3 7 9 6 1

12 27 35 23 4

4

5

7

20 6

77 23

5 9 12

19 35 46 23 73 4

4 5

7

4

5

2 5

6 19 1

4

5

7

26

e25

four distinct categories (awareness, acceptance, anxiety and fear of judgement). First, answering the questions increased participants' awareness of their own drinking levels by compelling them to reflect in detail on their drinking habits. The questions included a section on family history, and women often expressed a degree of surprise on how this section affected the way they regarded the drinking habits of family members. Women used words such as ‘surprising’, ‘a bit shocked’ and ‘reasonably confronting’ on reflection. Second, after their initial reaction to the actual questions, participants then tended to voice a general view that routine maternity care involved comprehensive health and lifestyle assessments and therefore most women would accept that questions about alcohol use will be asked. Many participants said that they themselves would have ‘no issues’ with answering such questions, with strong non-verbal assent from other members in their groups. Further, an overriding feeling of altruism persisted and participants cited a vested interest in the topic of alcohol use in pregnancy, wishing to provide information that would assist pregnant women in the future. The third category of emotional response was anxiety. Participants repeatedly spoke about the potential for detailed questions on alcohol consumption, particularly those relating to the time of conception, to cause a level of anxiety for some women. They would be suspicious that the reason for such detailed questioning was suggestive of an underlying potential for harm having been done. Fourth, a fear of being judged based on the amount of alcohol consumed was a topic that many women raised. Participants said this could raise sensitivities and they put forward many examples of situations in which a woman may underreport her actual levels of drinking, for example if she was a medium to heavy drinker and may have had difficulties reducing her intake.

Barriers to providing accurate data Problems with recall, complexity of the study instrument and use of subjective language were all identified as barriers to provision of accurate or honest answers to our questions. Recall was a particular issue for questions eliciting lifetime history of drinking. Further, when considering questions about drinking patterns over specific time periods (i.e. at different ages) many participants felt they were unable to accurately describe their typical drinking habits, as they may have changed significantly within any given time period. New mothers found it difficult to remember even their antenatal drinking patterns, commenting on their ongoing ‘pregnancy brain’ or describing life before the baby's birth as ‘a little bit hazy’. Family history of drinking often presented a dilemma because participants felt unable to categorise some family members. Reasons for this were either estrangement or because participants did not know or were not able to remember whether the current drinking behaviour of some elderly relatives reflected the person's lifetime drinking patters. Given that our questions were designed to accurately measure alcohol consumption using dose and pattern for distinct pregnancy time periods, questions were presented in several tables. This level of complexity was taxing for some participants with the potential for inadvertent mistakes or to prevent participants from continuing with the questionnaire. Similarly, terminology used in the questions to describe drinking frequency or which attempted to assess drinking behaviour e.g. ‘moderately’, or ‘intoxicated’, proved to be confusing and were a topic of discussion. These terms were considered to be open for individual interpretation and participants asked for better definitions.

e26

E. Muggli, MPH (Senior Research Officer) et al. / Midwifery 31 (2015) e23–e28

Fig. 1. Conceptual framework for understanding the determinants of accurate self-report in surveys of alcohol use in pregnancy.

Facilitators in providing accurate data The coding process revealed five categories which facilitated accurate data provision. These related to clear guides in serving size and type of drinks (drink choices), ability to record occasional drinking, increased understanding of the purpose of the questions, high level of confidentiality and convenience in administration. Many discussions arose around the topic of drink size (e.g. ‘I tend to think of a glass as a drink’). The concept of a ‘standard drink’ was perceived as too difficult to translate into drinking behaviour. Participants, however, felt reasonably comfortable in choosing a number of drinks from a drinks guide, providing it contained enough choices on sizes and drink types. Participants were concerned that, in responding to questions about typical individual drinking behaviour, they would not report occasional high-intake events and asked for an item to include such occasions, e.g. did you ever drink more, for example at a special occasion or during difficult times?. Putting alcohol use questions into context by providing an introductory explanation as to why participants are being asked to complete each section was viewed as important. Where this was not done, participants failed to understand the reasoning behind some questions and were put off by their level of detail and seemingly repetitive nature. Participants needed more assistance to better understand their purpose, particularly where questions were unexpected to them. Participants strongly felt that they would not favour face-toface interviews, or even telephone interviews, and this would result in under-reporting of pregnancy alcohol exposure in order ‘to present themselves in a better light’. Anonymity, with a choice of printed or web-based questionnaire, and also independence from health care providers (fear of judgement) were seen as important factors to encourage honest responses. Most participants also agreed that they would be more likely to complete the questionnaire in clinic while waiting for appointments, rather than at home where there are many competing demands.

Discussion This study used focus groups to explore how best to ask questions on alcohol exposure during pregnancy. Participants were asked to review detailed alcohol consumption questions to increase understanding of factors which help or hinder provision of honest and accurate responses in surveys of alcohol use in pregnancy. The conceptual framework developed in our analysis describes eight topics to consider when formulating such questions. Analysis also revealed a number of emotional responses to answering the questions, displaying sensitivities for careful

consideration by researchers and clinicians when asking women about alcohol use in pregnancy. For example, while women are prepared to talk about lifestyle issues such as smoking and alcohol consumption with their maternity clinicians (Jones et al., 2011), very detailed questions on patterns of drinking can cause a degree of concern. The issue of anxiety arising from discussing alcohol use in pregnancy has received little attention in the research setting. Anxiety undoubtedly exists for women who may have consumed alcohol before pregnancy recognition and are aware or become aware of policies and guidelines, including those published by the Australian National Health and Medical Research Council, recommending that not drinking alcohol while pregnant is the safest option (National Medical Health and Research Council, 2009). Women are likely to worry about the potential harm done to their unborn child and some may experience considerable anxiety associated with unintended exposure to alcohol and the uncertainty about how this will affect their child. Being asked about alcohol consumption in this situation may heighten existing worries. It was therefore not surprising to find in our study that participants talked about the potential to raise anxiety in some women along with an increased awareness of the level of personal drinking. Further, the reaction to such questions may produce emotions such a guilt and fear of judgement, which may result in inaccurate self-reports. This is in line with findings from another Australian study on pregnant women's attitudes to being asked about alcohol use in a clinical setting where participants felt that, while they may see no issues for themselves, not everyone may be truthful, especially if women experience guilt around their own alcohol consumption. (Jones et al., 2011) One of the barriers to accurately assessing alcohol use was remembering and estimating typical consumption over a longer period of time. This was a particular issue when women were asked to reflect on their lifetime drinking history. Recall bias is a well-demonstrated measurement error in retrospective studies and recent recall methods have been shown to encourage fuller reporting of alcohol consumption in population surveys (Stockwell et al., 2004). The use of a comprehensive drinks guide to link the type of drinks with questions to specify drinking patterns and timing of exposure was well received and reported as easy to use. Given that previous research has shown participants find it difficult to translate standard drinks into real life situations and consequently have a tendency to underestimate their actual drinking, (DevosComby and Lange, 2008; Kaskutas and Kerr, 2008; Kerr and Stockwell, 2012) a drinks guide is an important means to enable provision of accurate information. Further, we identified that a question about special occasion drinking encouraged women to disclose infrequent events where their alcohol intake was higher

E. Muggli, MPH (Senior Research Officer) et al. / Midwifery 31 (2015) e23–e28

than normal. This is in line with previous findings of respondents tending to report their median intake rather than the mean and that when asked to record typical individual alcohol intake, highintake occasions are excluded as they do not fall into a usual pattern (Stockwell et al., 2004). Providing context to questions about alcohol intake was identified as encouraging honest reporting. Strategies to contextualise alcohol questions have in the past included, placing them in the broader context of health behaviours and health screening in clinical settings (Russell et al., 1994), asking about items other than alcohol first (Harford, 1994; Budd et al., 2000), and asking about family history and personal drinking history before asking about current drinking patterns (Alvik et al., 2005). Confidentiality and convenience in the way questions are presented as well as the setting in which they are administered was also deemed to encourage honest and accurate reporting. Further, participants' preference was for a questionnaire method other than face-to-face, indicating a wish for increased privacy. Our study is not the first to identify convenience as a facilitator in pregnancy alcohol-use research (Streissguth and Gunta, 1992). Eligible individuals are more likely to participate if it is easy to do so and the time and effort required is perceived to be minimal. There are some limitations to be considered when interpreting our findings in the context of other situations. While the ‘purposeful’ sampling strategy (Krueger, 1998b) ensured the focus groups were composed of participants with something in common that was relevant to the study, e.g. being pregnant or recently given birth, selection bias may have resulted in the sample not being a true representation of the wider population of pregnant women, (e.g. 77% were primiparous compared with 43% in the Victorian population; Consultative Council on Obstetric and Paediatric Mortality and Morbidity, 2012). All new mothers in the focus groups had healthy infants, which may have affected their recall. In addition, response bias could have influenced what women said if they wanted to be viewed favourably by others in the group, perhaps not expressing their true opinions for fear of judgment. Steps were taken to limit response bias with researchers emphasising before commencing each focus group that they were not using any data on the participants' own drinking habits and the purpose was solely to express views about our survey questions. Although the study sample was small and self-selected, no new concepts arose before the last focus group was concluded, giving us confidence that we sufficiently explored the special knowledge, opinions and experiences of our groups. Together, they have provided an in-depth explanation of women's barriers and enablers to providing accurate data about alcohol use in pregnancy.

Conclusion Our findings suggest that while women's emotional responses to being asked about their alcohol consumption in the research setting were generally favourable, questions that provide a clear context may reduce anxiety arising from being asked questions about alcohol intake. Methods using shorter recall periods, supplying a comprehensive and relevant list of drinks choices and minimising complex and subjective language, are likely to increase survey participants' ability to provide accurate alcohol consumption data. A setting that is perceived as being confidential, and using methods other than face-to face interviews, may reduce women's desire to provide socially acceptable answers, thus increasing participants' willingness to be truthful. All these considerations will assist researchers in eliciting accurate information on alcohol exposure in pregnancy, particularly for lower levels of drinking, where women may have vested interests in the research outcomes obtained.

e27

Conflict of interest statement None declared.

Acknowledgements This work was supported by grants from VicHealth and the Australian Commonwealth Department of Health and Ageing. References Alvik, A., Aalen, O.O., Lindemann, R., 2013. Early fetal binge alcohol exposure predicts high behavioral symptom scores in 5.5-year-old children. Alcohol.: Clin. Exp. Res. 37, 1954–1962. Alvik, A., Haldorsen, T., Lindemann, R., 2005. Consistency of reported alcohol use by pregnant women: anonymous versus confidential questionnaires with item nonresponse differences. Alcohol.: Clin. Exp. Res. 29, 1444–1449. Alvik, A., Torgersen, A.M., Aalen, O.O., et al., 2011. Binge alcohol exposure once a week in early pregnancy predicts temperament and sleeping problems in the infant. Early Hum. Dev. 87, 827–833. Armstrong, E.M., Abel, E.L., 2000. Fetal alcohol syndrome: the origins of a moral panic. Alcohol Alcohol. 35, 276–282. Bakker, R., Pluimgraaff, L.E., Steegers, E.A., et al., 2010. Associations of light and moderate maternal alcohol consumption with fetal growth characteristics in different periods of pregnancy: the Generation R Study. Int. J. Epidemiol. 39, 777–789. Bay, B., Stovring, H., Wimberley, T., et al., 2012. Low to moderate alcohol intake during pregnancy and risk of psychomotor deficits. Alcohol.: Clin. Exp. Res. 36, 807–814. Budd, K., Ross-Alaolmolki, K., Zeller, R., 2000. Two prenatal alcohol use screening instruments compared with a physiologic measure. J. Obstet. Gynecol. Neonatal Nurs. 29, 129–136. Burns, E., Gray, R., Smith, L.A., 2010. Brief screening questionnaires to identify problem drinking during pregnancy: a systematic review. Addiction 105, 601–614. Consultative Council on Obstetric and Paediatric Mortality and Morbidity, 2012. Annual Report for the Year 2009. State Government of Victoria, Melbourne. Devos-Comby, L., Lange, J.E., 2008. “My drink is larger than yours”? A literature review of self-defined drink sizes and standard drinks. Curr. Drug Abuse Rev. 1, 162–176. Dolan, G.P., Stone, D.H., Briggs, A.H., 2010. A systematic review of continuous performance task research in children prenatally exposed to alcohol. Alcohol Alcohol. 45, 30–38. Elo, S., Kyngas, H., 2008. The qualitative content analysis process. J. Adv. Nurs. 62, 107–115. Flak, A.L., Su, S., Bertrand, J., et al., 2014. The association of mild, moderate, and binge prenatal alcohol exposure and child neuropsychological outcomes: a meta-analysis. Alcohol.: Clin. Exp. Res. 38, 214–226. Harford, T.C., 1994. The effects of order of questions on reported alcohol consumption. Addiction 89, 421–424. Henderson, J., Gray, R., Brocklehurst, P., 2007a. Systematic review of effects of lowmoderate prenatal alcohol exposure on pregnancy outcome. BJOG: Int. J. Obstet. Gynaecol. 114, 243–252. Henderson, J., Kesmodel, U., Gray, R., 2007b. Systematic review of the fetal effects of prenatal binge-drinking. J. Epidemiol. Commun. Health 61, 1069–1073. Humphriss, R., Hall, A., May, M., et al., 2013. Prenatal alcohol exposure and childhood balance ability: findings from a UK birth cohort study. BMJ Open 3, e002718. Jones, S.C., Eval, M., Telenta, J., et al., 2011. Midwives and pregnant women talk about alcohol: what advice do we give and what do they receive? Midwifery 27, 489–496. Kaskutas, L.A., Kerr, W.C., 2008. Accuracy of photographs to capture respondentdefined drink size. J. Stud. Alcohol Drugs 69, 605–610. Kelly, Y., Iacovou, M., Quigley, M., et al., 2013. Light drinking versus abstinence in pregnancy – behavioural and cognitive outcomes in 7-year-old children: a longitudinal cohort study. BJOG: An Int. J. Obstet. Gynaecol. 120, 1340–1347. Kerr, W.C., Stockwell, T., 2012. Understanding standard drinks and drinking guidelines. Drug Alcohol Rev. 31, 200–205. Kesmodel, U., Bertrand, J., Stovring, H., et al., 2012. The effect of different alcohol drinking patterns in early to mid pregnancy on the child's intelligence, attention, and executive function. BJOG: An Int. J. Obstet. Gynaecol. 119, 1180–1190. Krueger, R.A., 1998a. Focus Group Kit 3: Developing Questions for Focus Groups. SAGE Publications, Thousand Oaks, CA. Krueger, R.A., 1998b. Focus Group Kit 6: Analysing & Reporting Focus Group Results. SAGE Publications, Thousand Oaks, CA. Larkby, C.A., Goldschmidt, L., Hanusa, B.H., et al., 2011. Prenatal alcohol exposure is associated with conduct disorder in adolescence: findings from a birth cohort. J. Am. Acad. Child Adolesc. Psychiatry 50, 262–271. Liamputtong, P., Ezzy, D., 2005. Qualitative Research Methods. Oxford Univeristy Press, Melbourne.

e28

E. Muggli, MPH (Senior Research Officer) et al. / Midwifery 31 (2015) e23–e28

Makarechian, N., Agro, K., Devlin, J., et al., 1998. Association between moderate alcohol consumption during pregnancy and spontaneous abortion, stillbirth and premature birth: a meta-analysis. Can. J. Clin. Pharmacol. 5, 169–175. National Medical Health and Research Council, 2009. Australian Guidelines to Reduce Health Risks from Drinking Alcohol; Guideline 4: Pregnancy and Breastfeeding. National Health and Medical Research Council, Canberra. Nykjaer, C., Alwan, N.A., Greenwood, D.C., et al., 2014. Maternal alcohol intake prior to and during pregnancy and risk of adverse birth outcomes: evidence from a British cohort. J. Epidemiol. Commun. Health 68, 542–549. O’leary, C.M., Bower, C., Zubrick, S.R., et al., 2010. A new method of prenatal alcohol classification accounting for dose, pattern and timing of exposure: improving our ability to examine fetal effects from low to moderate alcohol. J. Epidemiol. Commun. Health 64, 956–962. Patra, J., Bakker, R., Irving, H., et al., 2011. Dose-response relationship between alcohol consumption before and during pregnancy and the risks of low birthweight, preterm birth and small for gestational age (SGA) – a systematic review and meta-analyses. BJOG: An Int. J. Obstet. Gynaecol. 118, 1411–1421. Polygenis, D., Wharton, S., Malmberg, C., et al., 1998. Moderate alcohol consumption during pregnancy and the incidence of fetal malformations: a metaanalysis. Neurotoxicol. Teratol. 20, 61–67. Robinson, M., Oddy, W.H., Mclean, N.J., et al., 2010. Low-moderate prenatal alcohol exposure and risk to child behavioural development: a prospective cohort study. BJOG: An Int. J. Obstet. Gynaecol. 117, 1139–1150.

Russell, M., Martier, S.S., Sokol, R.J., et al., 1994. Screening for pregnancy riskdrinking. Alcohol.: Clin. Exp. Res. 18, 1156–1161. Sayal, K., 2007. Alcohol consumption in pregnancy as a risk factor for later mental health problems. Evid. Based Mental Health 10, 98–100. Skogerbo, A., Kesmodel, U.S., Denny, C.H., et al., 2013. The effects of low to moderate alcohol consumption and binge drinking in early pregnancy on behaviour in 5year-old children: a prospective cohort study on 1628 children. BJOG: An Int. J. Obstet. Gynaecol. 120, 1042–1050. Stockwell, T., Donath, S., Cooper-Stanbury, M., et al., 2004. Under-reporting of alcohol consumption in household surveys: a comparison of quantity-frequency, graduated-frequency and recent recall. Addiction 99, 1024–1033. Streissguth, A.P., Gunta, C.T., 1992. Subject recruitment and retentions for longitudinal research: Practical considerations for a nonintervention model. In: Kilbey, M.M., Ashgar, K. (Eds.), (National Institute on Drug Abuse Monograph No. 117). Methodological Issues in Epidemiological Prevention and Treatment Research on Drug-Exposed Women and Their Children. US Department of Health and Human Services, Rockville, pp. 137–154. Testa, M., Quigley, B.M., Eiden, R.D., 2003. The effects of prenatal alcohol exposure on infant mental development: a meta-analytical review. Alcohol Alcohol. 38, 295–304. Underbjerg, M., Kesmodel, U., Landro, N., et al., 2012. The effects of low to moderate alcohol consumption and binge drinking in early pregnancy on selective and sustained attention in 5-year-old children. BJOG: An Int. J. Obstet. Gynaecol. 119, 1211–1221.