Understanding delayed access to antenatal care: a qualitative study Rosalind Haddrill

4.3.4 Collection of demographic data

Prior to the interview women were asked to complete a demographic questionnaire, or it was completed for them if they had difficulties with reading and/or writing, and consent was obtained to access the women’s medical records if necessary (though this was not required). This data collection was designed to reflect the socio-demographic factors associated with late booking in the literature (see chapter one pages 11-15) and to provide contextual information for the interviews (Bryman 2008). The questionnaire is presented in appendix 4.12, and a summary of the results are presented in table 5.1.1, in chapter five.

Table 4.3.4: Demographic data collected from the women interviewed (factors associated with late booking in parentheses):

Home address (living in areas with high deprivation indices) and length of time at address (temporary accommodation/recent arrival)

Age (older or young women)

Ethnicity (not ‘White British’ ethnicity)

First language (English not first language)

Parity (high and low parity) and age when first pregnant (young age)

Educational level of mother and baby’s father (low educational attainment)

Relationship status and support from baby’s father (lack of support, single parenthood)

Occupation and employment status of woman and partner (unemployment, low income)

4.3.5 The interviews

I carried out 24 of the 27 individual semi-structured interviews with the women in the study, between May 2007 and July 2008. The interview guide for these took the research question, “what perceptions and beliefs do late booking women express about the delayed initiation of antenatal care?” as its starting point, focusing on issues relating to late booking women’s experience, knowledge and beliefs. The flexible guide was designed to act as a prompt and a guide for discussion, without trying to limit participants’ responses. It was used to make interviewing more systematic and comprehensive, to provide a framework for the semi-structured interview process and to ensure some consistency in the data collected. However, there was a recognition that the standardisation of questions did not mean the standardisation of meanings for participants (Murphy et al 1998). The guide allowed for different interview patterns, individual interpretation and a degree of participant control. Its flexibility also allowed for modification during the research process (Bryman 2008, Holloway and Wheeler 2010).
The guide for the individual interviews was based upon four key areas relating to the initiation of and attendance for antenatal care. Its design reflected theoretical sensitivity, using professional experience and an exploration of existing socio-demographic studies of antenatal care attendance. It was developed through discussion with a group of academics and health and social care practitioners involved in the care of pregnant women. This included an academic social scientist, a GP, an obstetrician and a midwife. The initial interview guide (appendix 4.11) contained questions grouped around four main themes:

1) Reasons why the woman had presented late for antenatal booking;

2) The woman’s understanding of early antenatal booking and the health checks associated with this, and their importance;

3) Factors that might have prompted or helped the woman attend earlier for their antenatal booking visit and;

4) The woman’s plans for future care attendance and utilisation.
Again, reflecting the iterative approach of the research, the interview guide was designed to avoid prescription and preconceptions about why women might book late, which might limit the interview process. Questions were designed to be as open-ended and non-directive as possible, whilst still guiding; clear and at an appropriate level of understanding for the participant to enable them to tell their story. The interview guide consisted of a preliminary statement about the purpose of the study which was read to all participants, and simple questions, avoiding ambiguity, and using relevant language. It included some prompts and/or exploring questions to enable elaboration and allow different perspectives to emerge (Bryman 2008, Holloway and Wheeler 2010). Women were encouraged to consider barriers to their care because, as Melnyk (1988) identifies, “the concept of barriers is of particular interest because it helps to define the interface between the consumer and the system” (p196). There was no explicit consideration of possible solutions to late booking, though this was raised by some women.
The interview explored some general topics but was otherwise based around the way the participant framed and structured their responses. As women were interviewed other relevant themes began to emerge and the interview format was modified to incorporate these. Through the process of constant comparison the interviews were recorded and listened to afterwards to re-familiarise myself with the content, which in turn informed subsequent interviews. This revision, as research progresses, is a common part of iterative qualitative approaches. As Green and Thorogood (2014) argue, flexibility is essential when framing questions, using what works well in early interviews, what resonates with participants’ experiences, and removing what works less well. This was fundamental to not simply reflecting my own framing of the world but trying to reach the women’s, respecting the emic perspective by constantly returning to the data, rather than imposing preconceived ideas (Hall et al 2012).
In particular, modification of the interview guide resulted from some women’s general defensiveness with some of the questions, and a reluctance to talk openly about their late booking, which limited the initial interviews. The guide was modified, both in terms of the language used and the questions posed, to avoid the professional vocabulary that, on reflection, was evident, and to encourage women to tell the story of their pregnancy experience so far. In particular women were asked about discovering the pregnancy and their first thoughts and feelings. They were also encouraged to talk about how they made their initial access to antenatal care, and the influence of family and peers on the woman, her pregnancy and this initial access. Broad enquiry questions such as “tell me about your experience of finding out you were pregnant” were eventually used to start the interview, rather than questions about women’s late initiation of care. Patton (2002) suggests starting with simple, noncontroversial questions such as this, which encourage description, then moving on to opinions and feelings afterwards. He discusses how in qualitative interviewing the interviewee needs to become actively involved in providing descriptive information as soon as possible instead of becoming conditioned to providing short-answer, routine responses.
The interview guide was designed to create order and flow to questions, but with some flexibility. Charmaz (2002) distinguishes three types of questions which follow the flow and phases of an interview: initial open ended questions, intermediate questions and ending questions. These were used to encourage the women to elaborate on their experiences and their attitudes and those of the people close to them, but also to steer the conversation and to change subject. The flexible approach aimed to be responsive to participants, using the full range of questions, varying not only the content but the emphasis and order of the questions. This flexibility was also reflected in the timing and location of the interviews and the discrete use of taping. The guide was used less as the research progressed, as my confidence and skills, and familiarity with the interview process, grew. Some questions, for example relating to attendance for care in the future, were removed as they failed to elicit responses other than the expected ‘of course I will attend’.

4.3.5.1 My relationship with the women

As my experience as an interviewer increased, I recognised the importance of my social and language skills in optimising the encounters with the women. My goal was to allow the women to tell their stories by providing a facilitative audience that was appreciative of their position. The flow of the interviews was further encouraged with positive body language, careful listening and rephrasing, trying to avoid interrupting and leading questions, using silences and gentle prompts to encourage elaboration (Bryman 2008, Hall, McKenna and Griffiths 2012). My aim was to demonstrate empathy and encourage the women to be ‘not passive respondents but active participants in an important social encounter’ (Holloway and Wheeler 2010: 97).
Using the participant’s descriptive experiences as a starting point allowed me to get into their world and encourage ideas familiar and known to them, to understand their perspective, to focus on concrete rather than abstract or theoretical issues (Green and Thorogood 2014). It enabled me to create a context for the discussion, and establish some rapport and trust with the women, to try to move away from perceived judgement and assumptions, for example that late booking was a bad thing that they should be ashamed of (Patton 2002, Liamputtong and Ezzy 2005). It enabled us both to take a step back from the research question, to ask about experiences and feelings directly and indirectly related to the woman’s late booking. Encouraging the women to tell their stories using their own language was particularly important, as it allowed them to express ideas as they knew them, and tried to avoid favouring the articulate. This encouraged the participants to feel more comfortable and at home with their ideas, rather than struggling with ‘research language’ from the outset (Leininger 1985).
This focus on the women’s perspective required several approaches, and a delicately balanced role, which reflects previous consideration of reflexivity in chapter three and is discussed further later in this chapter. The first was to recognise my effect as an interviewer, to try to generate the right data. This was achieved by taking an empathetic approach and trying to establish a rapport with the participants. Recognising people as the subjects not objects of research, with different perspectives and perceptions, and attempting to achieve a relationship of equality and mutual respect was the starting point for this (Bowling 2009, Holloway and Wheeler 2010). Encouraging women to be honest and open, and to feel some ownership over the research, required a flexible approach to the interview process. Whilst previous evidence informed the overall structure of the interviews, the aim was to avoid superimposing preconceived ideas. As an interviewer, this necessitated stepping back in conversations, using minimal prompting to encourage stories to be told.
I approached the interviews being honest about my status as a midwife, albeit one who was not involved clinically in the women’s care. There was a constant consideration of this relationship, also the aims of the interview and how these could be achieved throughout the study, in terms of elements such as the content and timing of interviews, listening skills and prompting used, with modification as the study progressed. There was also a recognition that social differences in the interview relationship, and their impact on the data collected, needed to be acknowledged, documented and included in the analysis (Holloway and Wheeler 2010, Green and Thorogood 2014).
Despite my honesty and empathy, for some women there was a reticence associated with the relationship. However, for others there was an appreciation of my status, a recognition that I understood their situation and could provide information and reassurance. In part this reflected women’s lack of antenatal care up to that point in their pregnancies. Consequently, many interviews were preceded and followed by questions and discussion around routine aspects of pregnancy and antenatal care. Oakley (2004) also observed this questioning in her study of pregnant women. She acknowledges that it is impossible to ignore this relationship between interviewer and participant, and that it is a positive part of a more egalitarian interview process.

4.3.5.2 Recording and transcribing

The interviews were tape recorded to help preserve the participants’ words as accurately as possible. This also enabled me to have eye contact and pay full attention to what participants said, enabling the dynamics of the conversation to be maintained with minimal disruption and distraction (Patton 2002, Holloway and Wheeler 2010). As Lofland et al (2006) suggest “if conceivably possible, tape-record. Then you can interview… the process of note-taking in the interview decreases one’s interviewing capacity” (p106). Because of the risk of making people self-conscious or alarmed by recording them, all participants were asked for permission to record; none refused however the presence of the tape recorder may have limited responses. Notes were also taken by myself, after the interviews, to support the analytical process, and reflected ethnographic principles of naturalistic observation which were influential on the study.
The interviews were transcribed verbatim to keep the interviewers and interviewees words intact and provide the most reliable record of the discussions (Green and Thorogood 2014). This was an ongoing activity throughout the research process, enabling me to develop an awareness of emerging themes that influenced further interviews. It reflected the iterative and inductive approach of grounded theory, which was influential throughout the study (Bryman 2008). Transcription was a challenging and time consuming process, in part because the basic recording equipment used resulted in poor quality audio files, but also the consideration of what to include and exclude (Poland 2002). The audio files were converted into digital files using Audacity software to make the transcription process easier.
The interviews generated large amounts of text to be analysed. Approximately one third of the interviews were transcribed by myself, the remaining ones were transcribed by a professional transcribing service. All interview transcripts were reviewed by myself while listening to a digital version of the original recording. This was to ensure accuracy, particularly the inclusion of pauses, expressions of emotion and colloquialisms which were important to the meaning of the conversation. It also became an essential part of the analytical process. As Lofland et al (2006) argue

“it is in the process of transcribing that you truly ‘hear’ what the person has said and, as such, this is a period in which analytic insights are most likely to occur” (p 107).