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

4.3.2 Recruitment

Women were approached opportunistically to participate in this study (see figure 4.3.1) between September 2006 and August 2008 (interviews took place between October 2006 and July 2008). When I joined the study in May 2007, three women had already been recruited and interviewed by the previous research midwife. Late booking was defined as more than 19 completed weeks gestation when attending for the first hospital antenatal appointment. This was chosen as, although the end of the first trimester is widely considered the limit of early booking, from a clinical viewpoint 20 weeks is the gestation by which all antenatal screening tests should have been completed and is a widely accepted definition of late booking (NCCWCH 2008, 2010, EURO-PERISTAT 2013). As discussed in chapter two, different definitions of late will inevitably influence findings. Women booking at 13 weeks gestation are likely to demonstrate differing pregnancy experiences and attitudes towards their care than women booking at a more advanced gestation. The 20 week gestation was therefore also chosen to maximise the number of women making a deliberate choice to book late, rather than women whose discovery of pregnancy had been slightly delayed. Attendance at the first hospital booking appointment, rather than the initial referral appointment with the woman’s community midwife or GP, was chosen to examine whether external (such as administrative) as well as personal factors were influential in the delay.
Women were identified initially from the antenatal referrals faxed or posted to the hospital by community midwives or GPs. Women with an estimated gestation of approximately 20 weeks or more when attending for their first appointment were identified. Once this was confirmed by ultrasound scan at their hospital appointment I then spoke to them or arranged for the staff in the antenatal clinic to do so. As the study progressed and the process of theoretical sampling was developed, women were identified and approached following liason with midwifery staff working in antenatal clinic, particularly from specialist clinics such as the teenage clinic and substance misuse clinic.
Any woman who was interested in participating was then given an information sheet to read about the study and a consent form. Because initial attempts to recruit women found that they failed to send back consent forms, I obtained initial consent to participate and consent to contact the woman by telephone first before leaving the appointment. The women were then given a minimum of 48 hours to consider, before being contacted by telephone. Once contacted, if they were happy to participate and be interviewed, they were then asked for written consent, face to face, before the start of the interview, which was usually one to four weeks after the initial contact. Women who required the use of interpreters for their antenatal appointments were not recruited as there was no funding for this, however a few women whose first language was not English were recruited, where they were able to understand the project fully and give written informed consent to participate. One woman gave a joint interview with her husband. No financial inducements were made to participants, although reimbursements for any travel expenses incurred were offered.
Table 4.3.3 details the outcomes of the recruitment process for the women. If, after several attempts, a woman could not be contacted by telephone, a card was sent to their home address inviting them to contact me to arrange a time to meet. Many women (n = 39) were un-contactable after initially agreeing to be part of the study, and none of these returned the postage-paid cards that were sent out. Of the women that were contacted and agreed to participate, ten subsequently did not attend for interview and could not be contacted, and four declined to be interviewed. Recruitment stopped when, in discussion with the supervisory panel for the study, it was felt that theoretical saturation had been achieved. This resulted in a sample of 27 women.
Table 4.3.3: outcomes of the recruitment process for the study

Women recruited at the Jessop Wing, September 2006 – August 2008 Women attending for first antenatal (booking) appointment with gestation (confirmed by USS) > 19^+6/40. Potential candidates identified from referral letters by research midwife and through discussion with antenatal clinic midwives and specialist midwifery teams.
Eligible women given information pack and asked for initial consent for research midwife to contact (n = 83)	Not meeting entry criteria (n = 3)
Contacted by telephone and/or post by research midwife (n = 80)	Unable to contact by telephone/post (n = 39)
Contacted by phone and interview arranged in location of woman’s choice: at home, community children’s centre or next antenatal hospital appointment (n = 41)	Did not attend for interview and unable to contact by telephone/post (n = 10) Declined to be interviewed (n = 4)
Interviewed at home, children’s centre or hospital (n = 27)

4.3.3 The interview setting

It is recognised that in qualitative research, data collection approaches need to be sensitive to the social context in which data is collected. If not they risk being rigid, unnatural or unsympathetic, which could reduce the authenticity of the data (Topping 2010). Initially women were asked to come to the Jessop Wing specifically to be interviewed for the study. However, this proved to be ineffective as most women failed to turn up and this, in combination with women’s reluctance to participate, meant that initial recruitment was extremely poor. I recognised that to try to improve participation and eliminate barriers to attendance, I needed to understand the women’s lives better (again, a consideration of the emic perspective) and they needed to have some control over the interview process. As a result, interviews were subsequently conducted at the woman’s convenience, in her choice of location, either at the hospital to coincide with any further antenatal appointments, or at home or at another location of her choice, for example a local children’s centre. All women were interviewed antenatally, apart from one who was interviewed 48 hours after the birth on the postnatal ward.
It was quickly apparent that the setting for the interview affected its content and quality. Some authors have commented that it is usually preferable to interview people in their own homes or in a neutral location, reflecting the naturalistic approach of qualitative research. The rapport and trust between interviewer and participant is influenced not only by the language and body language of the interviewer but also the setting for the interview (Liamputtong and Ezzy 2005, Britten 2006). Therefore an interview in a private space that the participant has some ownership over is more likely to result in a relaxed atmosphere and a better rapport, and (by implication) richer interview data (Holloway and Wheeler 2010). This was certainly the case during the study, with the length and depth of the interviews varying considerably, depending both on women’s reticence to talk and the location. The interviews ranged in length from 16 to 63 minutes (mean = 32 minutes); women at home in general talked more openly and for longer.
It has been argued that a location related to a particular organisation, of which the interviewer is perceived to be a part (the NHS or hospital in this case), could be threatening. However, I wanted the participants to have some control over the process and particularly the choice of location. There was a trade-off between convenient settings for interviews (hospital) and more productive and trusting ones (home) (Green and Thorogood 2014). Some women did not wish to be interviewed at home, which seemed to be perceived as inconvenient or intrusive. This was reflected by the four women who were unavailable when I arrived at their homes to interview them, and other women who were interviewed but chose to do this at the Jessop Wing, prior to or following their next hospital visit. This resulted in convenient but somewhat time-limited conversations.