Setting

The study was conducted in three maternity services in London, England: University College London Hospital (UCLH), Homerton Hospital (HH), and St Thomas’ Hospital (STH). In England women are encouraged to have their first antenatal appointment with a midwife, known as the ‘booking’ appointment, before the eleventh completed week of pregnancy. At ULCH and HH, the LMUP was included in the booking appointment whereas at STH the LMUP was available for self-completion by women as part of a pre-existing pre-booking workflow in the maternity app ‘Badgernotes’.

Implementation

We sought approval from service leads and developed a technical specification for the changes required to the maternity information system (‘Medway’, then ‘EPIC’ at UCLH, a local build of ‘Cerner’ at HH, and ‘Badgernet’ at STH) which went through internal approval processes, commissioning, building and testing in 2017–2019 at UCLH, 2018–2019 at STH, and 2019–2020 at HH. In April 2019, UCLH changed to EPIC; however, problems with the EPIC rollout and the COVID-19 pandemic halted the pilot until April 2021. While the LMUP was available in Badgernet from early 2019, STH did not implement it until 2023. At HH the LMUP went live in August 2020 and was mandated, as is most of the booking appointment; at UCLH the LMUP was optional. UCLH and STH included the full wording of the LMUP whereas HH created a summarized version (Supplementary file S2).

At UCLH, JH met with midwives, facilitated by the Lead Midwife for Antenatal Care, and a midwife was employed 0.2WTE from May 2021 to support the implementation, including raising awareness via small group and 1:1 discussions, and evaluation at all sites. A video, covering the rationale and frequently asked questions, was sent to UCLH midwives and was included in the new staff pack, and posters were displayed in staff areas (UCLH) and via staff Instagram (HH) to maintain awareness. Specific training was not provided at HH or STH as locally it was felt that adaptation to using the LMUP would not be complex, especially as midwives at STH already asked whether the pregnancy was planned (yes/no), but staff were emailed about its introduction.

We worked with NHS Digital to create a space in the national Maternity Services Dataset²¹, and guidance and SNOMED codes for the LMUP²²; LMUP data were successfully submitted and received by the Office for Health Improvement and Disparities from 2023.

Evaluation

Analysis of LMUP data

Uptake and acceptability were measured using anonymized data from EPIC (UCLH) and Cerner (HH) from two months (August and September 2022) and from Badgernet (STH) for July–October 2023. This gave over 1000 women per site, considered an excellent sample size for evaluating a measure in new contexts²³. Completion of the LMUP was used as a proxy for acceptability for UCLH and STH data where it was not mandatory. We calculated the Cronbach’s alpha (reliability) and confirmed unidimensionality with a Principal Components Analysis (structural validity). Data were analyzed using chi-squared tests to ascertain whether certain groups of women were less likely to complete the LMUP or certain questions, or whether midwives had concerns with any questions. For construct validity we tested our hypotheses, based on previous research^13-15,17, that unplanned pregnancies would be more common in those aged <20 years, or who were unmarried/not in a relationship, or were parity three or higher. We used Kruskall-Wallis tests as LMUP score is non-parametric and because we were looking at differences across more than two groups.

Focus groups with midwives

The study was advertised to midwives via email. Focus groups were conducted with midwives who carry out booking appointments, according to a semi-structured topic guide (Supplementary file S3), to explore if/how they use the LMUP, what they liked or disliked about it, and how rollout of the LMUP could be improved. Given midwives at STH were not involved in asking the LMUP, we only conducted one focus group discussion (FGD) to explore their thoughts on the inclusion of the LMUP in the pre-booking workflow.

A mixture of eighteen community, clinic, and continuity midwives took part in three in-person FGDs at UCLH in February–June 2022. At HUH, six midwives took part in July–November 2022. The STH FGD with four midwives took place in May 2023. All FGDs and midwife interviews were facilitated by JH and either BS or CS, and took 40–55 min.

One-to-one interviews with women

The study was advertised to women through posters and flyers in booking packs. At STH, this was supplemented with targeted email invitations due to the low uptake of the pre-booking workflow (unrelated to the LMUP). We conducted one-to-one in-depth interviews with pregnant women who had had their booking appointment within the last four weeks (at point of contact) and had completed the LMUP questions (see Supplementary file S4 for topic guide). We planned to conduct up to 25 interviews.

Interviews and FGDs were conducted either in person at the Trust or online and were recorded using Zoom. All participants provided written informed consent. Field notes were made during interviews, and reflective notes afterwards, that were referred to during analysis. Video files were deleted once the transcription had been completed, checked and anonymized. Participants were asked if they wanted a copy of the transcript and none returned any comments. Transcribed interviews and FGDs were uploaded to Nvivo for analysis.

Recruitment took place at UCLH between February and May 2022. In all, 157 women expressed an interest, 31 were eligible, 20 were invited, and 13 were interviewed. At HH, recruitment took place July–October 2022; 24 women expressed an interest, 12 were eligible, 8 were invited, and 8 were interviewed. The most common reason for ineligibility was that their booking appointment was more than four weeks ago. We selected women from those who replied, to ensure we had a range of LMUP scores, ages, ethnicities and obstetric experiences, maximizing the information power of each interview²⁴. At STH, low completion of the pre-booking workflow necessitated a more active recruitment strategy; however, limitations in the search function of the electronic maternity system meant that not all the potentially eligible women were identified. Consequently only 39 women were contacted within the study. Four were ineligible due to non-completion of the pre-booking workflow, four consented to interview and the remainder did not reply. Interviews were conducted by JH, BS or CS on Zoom, and lasted 30–40 minutes.

Analysis

A thematic Framework Analysis^25-27 aligned with the qualitive description approach²⁸ was conducted to explore women’s and midwives’ thoughts about the LMUP; differences of opinion between women with unplanned compared to planned pregnancies were investigated.

JH developed a framework with high-level themes at the individual level of midwife and woman, organizational level and external factors, based on other research^29,30, and sub-themes based on the data, and discussed with BS and CS. The transcripts were indexed by CS, BS and JH.

Participants are referred to with an identifier comprising their location, whether or not they were a midwife (MW), and a number. For example, HHMW24 refers to the setting of Homerton, MW means midwife, and 24 is the participant’s assigned number.

Patient and public involvement

The patient and public involvement group aligned to the P3 Study (Pregnancy Planning, Preparation and Prevention), of which this research was part, were involved in discussions about the design and conduct of this research, reviewed participant documentation and topic guides, and discussed findings.

Details of ethical approval

Ethical approval was granted by the South West, Cornwall and Plymouth Research Ethics Committee (REC reference: 21/SW/0174) for UCLH on 5 January 2022. The amendment to add HH was approved on 25 April 2022. The amendment to add STH was approved on 13 December 2022, and to send emails to potential participants was approved on 15 August 2023.

Analysis of anonymous antenatal data

Data on 1221 women were extracted from UCLH, 1110 from HH, and 1129 at STH. Population characteristics are shown in Table 1. The populations varied, reflecting the hospital’s catchment areas, with the highest average age at UCLH, greater ethnic diversity at STH, and more deprived populations at HH. Consequently, there was variation in the levels of pregnancy planning, with the highest proportion of planned pregnancies (LMUP score >9) at UCLH (84.2%) (Table 2). The LMUP performed as expected in terms of its psychometric properties at UCLH and STH; at HH there was a possible issue with question 1, but otherwise performed well (Supplementary file S5).

During the implementation at UCLH, completion of the LMUP increased from 51% in May 2021 to 85% in January 2022, largely due to the work of the LMUP champion. Completion of the LMUP questions by site is shown in Table 2; uptake shows clear differences according to the workflow. Partial completion was uncommon, especially at STH, and for each individual question there were <5% missing data at all sites, demonstrating acceptability. At UCLH, women with IVF pregnancies (n=124 vs n=1049) were less likely to be asked the LMUP (p<0.001) but there were no significant differences in completion by age, marital status or ethnicity. However, question five was more likely to be missing in women who were recorded as unmarried (n=69 vs n=175, p=0.015). Conversely, at STH, women aged <20 years (n=28 vs n=1832, p<0.001), who were not White or Asian (n=454 vs n=1402, p<0.001) or who were not in a relationship (n=79 vs n=1694, p<0.001) were less likely to complete the LMUP. The relationship with fertility treatment was more complex at STH, with the main difference being that women who conceived with fertility treatment were more likely to partially complete the LMUP (n=4/1048 vs n=9/147, p<0.001).

Focus groups with midwives

The 18 midwives included seven clinic and 11 community or continuity midwives, ranging from newly qualified midwives to those with more than a decade’s experience. We did not collect data on their sociodemographic characteristics.

One-to-one interviews with women

The characteristics of women are shown in Table 3. Most women had had at least one previous pregnancy (n=17, gravida 1–7), and some had experienced miscarriage, termination or IVF. Several had pre-existing medical conditions, including diabetes, cardiovascular disease and substance abuse. We interviewed women with a good spread of pregnancy intentions (LMUP score 2–12).

Findings

The three themes and related sub-themes are shown in Figure 1. Given the overlap identified during analysis between what the women and midwives had said, some codes included midwives and women’s opinions together. Illustrative quotes are provided in Table 4.

Figure 1

Themes and sub-themes identified within the overall framework of individual, organizational and external factors, from pregnant women and midwives interviewed in 2022 at University College London Hospital and Homerton Hospital, and in 2023 in St Thomas’ Hospital

Individual level

LMUP is acceptable

Across all settings, women found the LMUP easy to understand, and midwives found it easy to use. Women thought the questions fitted well within the booking appointment, whether in person or by self-completion in Badgernotes. Midwives agreed, saying that the addition of the questions had not made much difference to the booking appointment and that these questions were not as personal as others, such as those on domestic violence.

Benefits

Rapport

The main benefit of asking the questions, according to midwives and women at UCLH and HH, was that the ensuing discussion can help build rapport and allow the midwife to get to know the woman, and her situation, better. This was one reason some of the women at STH thought that the questions would be best asked by a midwife rather than in the pre-booking, or at the very least should be discussed in the booking appointment.

Value of discussion

Midwives and women agreed that the questions could be used to open up valuable discussions not only about the circumstances of the pregnancy, but also about support, emotional wellbeing, relationships, continuing the pregnancy, safeguarding or general health. Women said that they thought these conversations would be especially valuable for women with unplanned pregnancies, with one woman saying that the questions led to a discussion that helped her process her unplanned pregnancy.

Personalization of care

Midwives thought that the questions were a good prompt for finding out health information about the woman, which can be used to identify any changes to care that need to be made or to trigger a conversation about health behaviors. It could be used to help identify women with other vulnerabilities, such as domestic violence or sexual abuse, or to prioritize discussions about postnatal contraception.

Barriers

Perception of judgement

The main concern of midwives was the potential to make women feel judged or guilty, especially women with unplanned pregnancies and particularly with question 6 about taking action to prepare for the pregnancy. This was one reason why self-completion was suggested. By contrast, none of the women interviewed found the questions concerning, upsetting or judgmental. A few women, mainly those with planned pregnancies, thought that women in different circumstances might find the questions more sensitive. However, LMUP score was not associated with how women felt; women with unplanned pregnancies did not find the questions any less acceptable than women with planned pregnancies. Neither did women report feeling offended or guilty by question 6.

Relevance of LMUP

Some women were unsure about the relevance of the LMUP questions in general, suggesting that if you were pregnant, it did not matter how you got there. However, others could see why the questions were being asked, with some labeling them ‘important’, particularly if women had other challenges.

Organizational level

Prioritization and awareness

At UCLH the main reason midwives did not ask the LMUP questions was time constraints; it was seen as a low priority due to the lack of associated action and poor knowledge about the link between unplanned pregnancy and adverse outcomes. At HH, midwives were unsure why the questions had been introduced and were unaware of the impact that unplanned pregnancies can have. Midwives generally agreed that if there was clear action triggered by the LMUP score and they better understood why the LMUP was being asked and the impact that unplanned pregnancies can have, then they would be more likely to ask the questions as intended.

Impact on care

Overall, most women felt that the LMUP questions had had little direct impact on their care, particularly those with the most planned pregnancies. When asked what services would be beneficial, women mentioned additional support, especially for those with unplanned pregnancies, in the form of counseling, group information sessions, face-toface appointments and signposting. Midwives at UCLH suggested that a specialized midwife providing support to women with low LMUP scores would be valuable, with midwives at HH suggesting women could be referred to their public health midwives. Other services mentioned were postnatal contraception, additional in-person antenatal appointments, early referral to their health visitor and additional postnatal care.

Presence of partner

All women said that their answers would not have changed if their partner was present. However, a few thought that for some other women, having their partner present might make it harder. Midwives were divided, with some finding it harder to ask or wondering if the woman might be less honest if their partner was present, while others said it was fine with the partners present, and that it can even be good as it involves the partner. At STH women are seen on their own for a few minutes at the start of booking, and both women and midwives suggested that this was a good opportunity for the LMUP questions.

LMUP being mandatory

Most HH midwives were happy with the LMUP questions being mandatory, like everything else in the booking. The only issue was regarding women who had experienced sexual abuse or rape, where an opt-out or free-text box was desired. Some UCLH midwives thought it should be made mandatory, but others preferred flexibility.

Pre-booking self-completion

Women and midwives at UCLH and HH independently suggested that women could complete the LMUP questions, and other information, before booking, saving time in the appointment. Some midwives thought women would be unlikely to do this, based on their previous experience. At STH, uptake of this workflow was low due to the number of steps women must take to download the app and log in. While the women we spoke to had successfully navigated this, most noted that it was a complex process. Once in the app, they found the whole questionnaire, LMUP included, straightforward. The benefits included being able to: complete it from the comfort of your own home at a convenient time; find out the answers (e.g. about a relative’s health) rather than being put on the spot; involving your partner should you wish; and relieving the burden on the midwife to free up time for more valuable conversations. Midwives at STH agreed that this content, and particularly family history, was well-suited to self-completion. On the rare occasion that the pre-booking information had been completed, it freed up time in the booking appointment as answers could be reviewed more quickly and the most important issues targeted.

External level

COVID-19

Due to COVID-19, booking appointments were conducted by phone for parts of 2020 and 2021. When asked about the effect of this, midwives agreed that telephone bookings were harder than in-person appointments and they might have been more likely to skim over the LMUP questions.

Service planning

Midwives thought that the information gained from the LMUP questions could be used to plan preconception or contraception services more effectively, by highlighting, for example, the ongoing low levels of uptake of folic acid before booking.

Suggestions

Introducing the LMUP

While most midwives start off the booking by alerting women to the fact that there are a lot of questions and everyone is asked the same, some thought that introducing the LMUP questions would help overcome concerns about sensitivity, judgement and relevance, and would allow women to answer more honestly. Some women at ULCH and HH also thought this would be beneficial. We developed, tested and agreed on the following introductory sentence in the July FGDs:

Women completing the questions in Badgernotes at STH generally did not feel the need for this.

Main findings

This is a novel study that evaluates the implementation of the LMUP in antenatal care in the UK. We have shown that it is feasible and acceptable to women and midwives in three large, busy London maternity services using different maternity information systems. Several advantages of asking the LMUP were highlighted by participants, including providing additional support and personalizing care, and concerns about judgment were not borne out; on the contrary women with unplanned pregnancies valued such discussions. Midwives highlighted an additional benefit that asking the LMUP questions can provide opportunities for women to disclose issues such as sexual abuse or substance misuse, though these were also reasons that sensitivity was needed. Disclosure can prompt discussions and referral to safeguarding, domestic abuse services, psychological support and other relevant support services and pathways. Women should be routinely asked about their mental health at each antenatal and postnatal visit. If a woman is flagged as having a lower LMUP score however, the importance of ensuring these enquiries are not missed could be emphasized, recognizing the higher prevalence of postnatal depression in this population.

Examination of the completion rate by question at UCLH showed that, where the LMUP was at least partially completed, question 6 (a list of actions taken prior to pregnancy) was the one that was most likely to be incomplete. This is in line with findings during the implementation about the challenges midwives experienced with this question. At UCLH each option (e.g. taking folic acid, achieving a healthy weight, stopping smoking) was listed separately with a yes/no tick box. Based on midwife feedback a ‘not applicable’ option was added. This was despite initial additional training and explanation that the focus of this question is on behavior change in preparation for pregnancy and not to assess the prevalence of behaviors. However, this did little to improve completion and we presented UCLH midwives with the summarized version used in HH and Sydney²⁹, which has now been implemented at UCLH.

Other observed variation in completion at UCLH, including women with IVF pregnancies being less likely to be asked the LMUP and question 5 (partner) being more likely to be missing in unmarried women fits with what midwives said and is likely due to midwives not asking rather than women refusing. Interpretation of the differences seen in LMUP completion by sociodemographics is complicated by the different workflows. At UCLH it was reassuring to see that midwives were no more or less likely to ask the LMUP based on age, ethnicity, or marital status. The differences seen in STH on these variables may represent the varied uptake of the pre-booking workflow by women with these characteristics rather than completion of the LMUP itself, but we were unable to investigate this due to the limitations of the data.

Interpretation

There is limited other evidence on the clinical implementation of the LMUP. An Australian evaluation of the use of the LMUP in antenatal care, which spoke only to midwives, found similar levels of general support and the same concerns regarding time constraints and the lack of associated action²⁹. Analysis of LMUP completion in those two hospitals in Sydney, where the LMUP was asked by midwives and was not mandatory, showed significant differences between the sites, with 32.0% uptake at the tertiary referral hospital and 96.3% at the secondary hospital³¹. Important differences were noted, including that leadership support was less strong at the tertiary hospital, which resonates with our findings of importance of local leadership and LMUP champion. In addition, implementation coincided with COVID-19 and while bookings continued in-person at the secondary hospital, the tertiary hospital switched to telephone appointments³¹; this was noted in our study to reduce the likelihood that the LMUP was asked. At the tertiary hospital, LMUP completion rates were lower in women born overseas, whose preferred language was not English, or who had lower socioeconomic status; these were not factors considered in our study³¹.

A study of the potential introduction of the LMUP into early pregnancy units described clinician’s thoughts on asking about pregnancy intendedness as ‘polarized’, with some considering it essential and others insensitive³⁰. Importantly, this study was hypothetical and, prior to being asked, none of the nurses interviewed knew of the LMUP. Surprisingly some thought that if the pregnancy was not continuing then it was not important to know, showing a lack of a holistic consideration of the woman’s needs; if that were an unplanned pregnancy a discussion of post-pregnancy contraception needs is indicated, in line with the Hatfield Vision recommendations³². Conversely women who were likely to try for another pregnancy would benefit from preconception advice.

It has been suggested that the LMUP is too complex to use, calculate or interpret in antenatal care³³. This opinion, not based on an implementation or evaluation, is not upheld by our findings. Maternity information systems can be programmed to calculate and display the score to the midwife for action, as is the case now in HH where the LMUP score can trigger referrals regarding postnatal contraception; pathways that are being investigated elsewhere.

Recommendations

Local implementation

National implementation

Future research

Strengths and limitations

A particular strength of this study is that we have incorporated the views of both those who collect (midwives) and provide (women) data at the booking appointment, and have triangulated this with data on completion for an in-depth evaluation. While our purposive sampling ensured as diverse a sample as possible, we were limited by who responded to advertisements. For example, no eligible women aged <26 years contacted the study. However, we were able to include women with unplanned pregnancies, the group midwives were most concerned about, and found that they valued the conversation the most. Evaluation in three sites has enabled a deeper learning of the barriers and facilitators to implementation by comparing the processes and findings. However, all three sites are in England where early antenatal care is led by midwives; consideration would need to be given to the transferability of these findings to settings where this is not the case. A limitation is that by only considering the booking appointment we are missing the women with the most unplanned pregnancies, who are seen in termination services or who may access antenatal care late or not at all, and those experiencing miscarriage, who may be seen in an early pregnancy unit. At STH it was not possible to identify women who had completed the pre-booking workflow but omitted the LMUP so we could not gain a complete picture of acceptability. The low level of completion mostly reflects the low uptake of the pre-booking workflow, which some midwives estimated to be as low as 10% in women attending antenatal clinics.