In 2022, research into the experiences of trans and non-binary users of maternity services in England was commissioned by the Health and Wellbeing Alliance (jointly managed by the Department of Health and Social Care, Public Health England, NHS England and NHS Improvement (LGBT Foundation, 2022). The research was conducted and published by the LGBT Foundation, and culminated in a report entitled ‘improving trans and non-binary experiences of maternity services’ (ITEMS) (LGBT Foundation, 2022). The ITEMS report argued that trans and non-binary maternity service users have ‘poor experiences’ and ‘poorer outcomes’ compared to other maternity service users, and that trans and non-binary ‘patients and their babies [are] being put at risk’ in the NHS (LGBT Foundation, 2022).
The report made a range of recommendations, which included changes to use of language, display of trans-inclusive communications in maternity settings, staff use of pronoun badges, the provision of personalised and trauma-informed care to trans and non-binary service users, staff training, and changes to IT systems and demographic monitoring that would require staff to ask all service users about their gender identity. NHS England (2022) announced plans to spend £100 000 on staff training based on the report's recommendations. These plans were subsequently withdrawn, following a petition by clinicians expressing concerns that the report's conclusions and recommendations were unsupported by its data (With Woman 2023a, b).
This article argues that there are substantial weaknesses with the framing, data collection and interpretation of findings in the ITEMS study. Therefore, the claims and recommendations made in the report should be viewed with caution. Furthermore, some of the recommendations include policy changes that would impact all maternity service users (not only the minority who identify as trans and non-binary). The authors of this article feel that caution should be taken, and wider impact assessed, before the experiences of a small minority are used to inform policy for all service users. They recommend that the NHS should not make policy on the basis of the ITEMS report. Further, they feel that NHS England and NHS Improvement should review the report and consider whether there is a need for more rigorous research into the experience of trans and non-binary maternity service users, and that any consideration of policy change on the basis of such research should balance the needs and interests of different groups of service users.
Critique of the study and report
This article presents a critique of the ITEMS study, based on its report (LGBT Foundation, 2022). The authors of this article argue that the study framing lacks clarity and balance, that there is a lack of engagement with relevant literature, that the methodology is substantially flawed, leading to potentially invalid findings, and that some of the recommendations are not based on the findings of the study.
One-sided, uncritical framing and lack of conceptual clarity
Concepts relating to gender and gender identity are unstable, controversial and lack universal consensus (Jones and MacKenzie, 2020; Joel and Fine, 2022; Sullivan, 2023). With reference to maternity care, Gribble et al (2022) argued that sex-based (rather than gender identity-based) language is important for communicating clearly. However, the ITEMS report uses gender-identity based language without acknowledgement that it is contested. For example, the study sets out to ‘allow comparison of maternity care between trans and non-binary birth parents and cis women’, with ‘cis’ defined as ‘someone whose gender identity matches with their gender assigned at birth’. ‘Gender identity’ is a contested term, and not everybody agrees that it is universally experienced (Stock, 2022; McGrath, 2023). The authors of this article suggest that a report into maternity service provision should be reflexive about language use and make an evidence-based case for the choice of concepts on which the study is based.
Lack of thorough engagement with literature
The ITEMS report begins with a literature review, which the authors state was the basis for elements of their methodology and some of their recommendations. The authors do not provide any information about the methods used for the review, the search terms or the inclusion criteria. However, there are substantial gaps in the cited literature. The authors did not engage with key literature that evidences the complexities and differences in individual preferences and strategies of trans and non-binary people navigating reproductive services (for example Hoffkling et al, 2017; Klein and Golub, 2020; Agénor et al, 2021). It also does not distinguish between international and UK research contexts, compromising transferability to the NHS.
Lack of methodological rigour
The report's research design included a survey and individual qualitative interviews. However, the authors of this article suggest that the design of the survey lacked rigour and transparency. This included major weaknesses with sampling strategy and survey design, as well as misrepresentative analysis and presentation of findings, which are discussed in this article. The authors believe that this means that the findings of the report may be unreliable and invalid.
Sampling and eligibility
The ITEMS report makes direct comparisons between its findings and the findings of the national maternity survey (NMS) (Care Quality Commission, 2020), and on this basis claims that trans and non-binary maternity service users have poorer experiences than the wider population of maternity service users. However, sampling and participant recruitment differed substantially between the two studies.
The NMS in 2019 used a self-selecting sample from their target population, which was all those who gave birth in any one of 121 NHS trusts during February 2019. Overall, 37% of this target population completed the survey (n=17 151). In contrast, the ITEMS survey was publicised via the research group's personal and professional networks and to the general public via social media. Responses were collected during a 5-month period from November 2020 to March 2021. No specific eligibility criteria are recorded in the report, and there is no indication that respondents were required to have given birth in England, or under NHS care.
The report claims that the survey achieved a ‘completion rate’ of 41%, but it is not clear what this claim is in relation to. They may be referring to the proportion of respondents that completed the whole survey (as opposed to part of it). However, they also report that there were 121 respondents in total and that none of the respondents completed the entire survey. Furthermore, each question relating to experiences of maternity care was answered by only 52–62 respondents.
The authors then compare their completion rate to that of the NMS (Care Quality Commission, 2020), which they claimed was 36.5%. However, the report of the 2019 survey does not include information on a completion rate; their adjusted response rate was 37%. It is possible that the ITEMS report authors were comparing their own (undefined) completion rate with the survey response rate, which was calculated in relation to the number of maternity service users in the country. In contrast, as the ITEMS authors identify, there are no data on numbers of trans and non-binary users of maternity services in England, so it is not possible to identify the response rate in relation to this population. Whether the 41% completion rate refers to the number of respondents that completed the whole survey, or is a response rate in relation to the target population, it does not appear to be supported by the data provided.
Unlike for the 2019 NMS, there was no requirement for ITEMS respondents to have given birth in England. The ITEMS report does not include any information about where respondents gave birth, and there is no basis on which to assume that their experiences took place in England or the UK.
It is also important to note that there is a substantial difference in the time period under investigation between the NMS in 2019 and the ITEMS report. The report's respondents gave birth over a 30-year period, with 45% giving birth before 2015. NHS maternity services have changed substantially since some participants gave birth in the early 1990s. Previous NMS results show initial improvements in women's experience of maternity services between 2013 and 2019 (Care Quality Commission, 2015; 2018; 2020), and a subsequent decline (Care Quality Commission, 2023). The ITEMS report is therefore comparing the experiences of (any) maternity service users across an extended time span with those who used the NHS service in England during 1 month, at its highest rate of maternal satisfaction in at least a decade.
Furthermore, nearly half of the participants in the ITEMS study gave birth before the introduction of contemporary models of practice implemented after the ‘Better Births’ report (NHS England, 2017) and the introduction of the NHS long term plan, which aimed to increase personalised care (Winfield and Booker, 2021) and patient access to digital records (Kulakiewicz et al, 2023). The number of midwifery-led units has also increased substantially (McCourt et al, 2014). All of these changes may have impacted the experience of maternity service users, meaning that the ITEMS survey cannot be directly compared to the NMS of 2019 without controlling for year of use of service.
In addition, the ITEMS sample was demographically different from the national maternity survey sample. ITEMS respondents were less likely to be white than NMS respondents (62% ITEMS vs 84% NMS). Additionally, 29% of ITEMS respondents reported that they were disabled. NMS respondents were not asked if they were disabled, but 15% of respondents reported a long-term health condition. ITEMS respondents were thus more likely to be in groups already known to have disproportionately negative experiences of maternity care (Malouf et al, 2017; Higginbottom et al, 2019; Peter and Wheeler, 2022). These differences were not controlled for in the ITEMS analysis, and the authors of this article suggest that attribution of disparities in experience to trans and non-binary identity may be invalid.
It is not clear that all of the ITEMS participants were trans or non-binary, or what precisely the authors mean by trans and non-binary maternity service users. There was no requirement that respondents must have identified as trans and non-binary at the point of giving birth, nor when completing the survey. Two of four interview participants reported that they concealed their trans and non-binary identity from healthcare professionals, which perhaps makes it particularly uncertain that negative experiences were associated with trans and non-binary status. Only 70% of the ITEMS survey respondents declared their gender identity when completing the survey, which means that there is no way to know the gender identities of 30% of the respondents.
While the report defines the terms ‘trans’ and ‘non-binary’ in its glossary, not all those who gave a gender identity identified themselves by either of these terms. Respondents identified variously as man, woman, non-binary, agender, gender queer, genderfluid, bigender, transmasculine, demi-boy and ‘in another way’. Most of these terms are not defined in the report, and all appear to be included in the analysis as representative of trans and non-binary experiences (including those identifying as ‘woman’).
None of the terms appear to relate to whether people have undergone physical procedures in relation to gender reassignment. The needs of medically transitioned maternity service users are likely to be different from other trans and non-binary service users, but are mentioned only briefly in the ITEMS report. For example, elective double mastectomy can impact the ability to produce breast milk (Gribble et al, 2023), and long-term testosterone can cause vaginal and uterine atrophy (Grynberg et al, 2010), which can impact birth and postnatal health (Indig et al, 2023). Hoffkling et al (2017) found that previous medical interventions made a difference to the needs of trans maternity service users, so failure to ask respondents about their history of physical transition may mask important information relevant to the needs of trans and non-binary maternity service users. Lack of clarity about the target population of the study means that there is no way to identify who precisely may require particular types of care.
Lack of internal reliability
The survey instrument for the ITEMS study appears to have asked imprecise questions, making the responses difficult to interpret and threatening internal reliability. The authors of this article were not able to find a publicly available record of the full questionnaire, but some of the questions are documented in the report.
For example, respondents were asked ‘did you get support from NHS or private midwives during your pregnancy/pregnancies?’. Almost a third (30%) of respondents answered ‘no’ to this question. The ITEMS research team interpreted this as meaning that 30% of their respondents gave birth ‘without ever accessing perinatal care’ (LGBT Foundation, 2022). However, the report states that, elsewhere in the survey, respondents were asked whether they received antenatal, labour and birth, and postnatal care, with 82%, 79% and 75% responding affirmatively, respectively. If the former question is intended to determine whether respondents received any perinatal care at all—as it is being interpreted by the report authors—this disparity between the answer to this question and answers to questions about receipt of care indicates a lack of internal reliability. It cannot be the case that 30% of respondents received no antenatal, birth and labour or postnatal care, and that 82%, 79% and 75% respectively received this care.
The authors of this article suggest that there are at least five ways to interpret the question ‘did you get support from NHS or private midwives during your pregnancy/pregnancies?’ and at least four reasons for a respondent to answer ‘no’ that do not mean a respondent received no perinatal care. For example, respondents could have been supported by nurses or doctors during their pregnancy, may have given birth before the arrival of a midwife or their arrival to hospital (born before arrival) (Birthrights, 2023), may have interpreted the term ‘supported’ to mean whether they felt supported, rather than whether they were under clinical care, or may have interpreted ‘during your pregnancy’ to mean antenatal care only, not birth, labour or postnatal care.
This lack of clarity, and questionable internal reliability, undermines the credibility of the ITEMS report headline claim that ‘30% of trans and non-binary respondents did not access NHS or private support during their pregnancy or pregnancies. This is sometimes called freebirthing’ (LGBT Foundation, 2022).
Misleading claims
On the basis that 30% of their respondents answered ‘no’ to ‘did you get support from NHS or private midwives during your pregnancy/pregnancies?’, the ITEMS report authors make the questionable claim that 30% of respondents free birthed. The Royal College of Midwives (2022), maternity rights charity Birthrights (2023), and widely cited scholars Feely and Thompson (2016) all define free birthing as choosing to give birth without the presence of a healthcare professional. Birthrights explicitly distinguishes between freebirthing as an active choice, and born before arrival. Where unassisted birth means deciding to give birth at home or somewhere else without the help of a healthcare professional, such as a midwife, it does not mean giving birth at home before the midwife you planned had time to arrive. This is called ‘born before arrival’ (Birthrights, 2023). Based on this definition of freebirth, a cohort study estimated that less than 0.05% women freebirth in the UK (Loughney et al, 2006).
It is not clear that ITEMS participants were asked whether they free birthed, much less whether they were provided with a definition of the term, and there are several other possible interpretations of the responses given. Despite this, the ITEMS report goes on to make a range of claims about the specific experiences of respondents who allegedly free birthed, without explanation of the empirical basis of those claims. The report claims that 30% of those who free birthed agreed that they would not consider accessing maternity services, almost 40% said they would have been uncomfortable accessing maternity services, and only 20% of those who free birthed reported being confident accessing maternity services if they felt that they needed to.
The report contains no detail of the basis of these assertions. For example, it is not clear that any respondent identified themselves as having free birthed, or that these respondents were asked particular sets of questions. Overall, the claim that 30% of participants free birthed (some 600 times more than the UK national freebirth rate) is not a valid interpretation of the data presented in the report, and any particular claims about the experiences of those who free birthed are thus also invalid.
Problems with sampling, instrument design and interpretation of data all raise concerns about the reliability and validity of the ITEMS report findings. In particular, the claims that ITEMS respondents had relatively poor experiences compared to national maternity survey 2019 respondents, and the claim that 30% of ITEMS respondents free birthed are used by the authors of the report as evidence that trans and non-binary maternity service user and their babies ‘are being put at risk’. The authors of this suggest that the study's data do not support this claim.
Recommendations that are not supported by findings
The ITEMS report includes recommendations that are not related to the claimed findings of the study. One recommendation is that the experiences of trans and non-binary maternity service users would be improved by the use of visible markers such as ‘posters, badges, including name badges with pronouns, and lanyards’ to ‘communicate that they are welcome’. However, there is no justification for these recommendations on the basis of the research findings. The survey report does not mention any investigation of people's experiences with interventions such as these, and one of the four interviewees reported that ‘a rainbow poster’ was inadequate, that they wanted personalised care. This policy recommendation therefore appears to be unsupported by the findings of the study.
The importance of a balanced approach to policymaking
Overall, the ITEMS report contains some substantial flaws relating to its framing, methodology and recommendations. The authors of this article therefore suggest that the NHS should exercise caution before making use of these recommendations to inform policy. However, the ITEMS authors do make some recommendations that are, to some extent, grounded in the reported findings.
For example, the report recommends the use of ‘inclusive language for every service user’. The authors do not specify precisely what they mean by ‘inclusive language’, but they include an example of an interview participant who said that receiving a letter referring to ‘pregnant women’ made them ‘worry that the maternity service would not be able to accommodate them’. This appears, therefore, to be a recommendation for a move from sex-based to gender-based referents in maternity services, and it appears that at least one ITEMS participant would have found such a change beneficial. However, whether the NHS should change its approach overall should be considered by balancing the interests of this person (and any others with similar experiences) with impact assessments relating to the wider population of maternity service users.
Gribble et al (2022) have argued that failure to use sex-based language can have substantial negative impacts on women using maternity services, a position supported by the recent Women's Health Strategy (Department of Health and Social Care, 2022). Responses to the 2021 census suggest that only 0.5% of England's population self-identity as trans and non-binary, although the validity of this data has been called into question (Biggs, 2023). A population-level move from sex-based to gender-based referents may have a detrimental effect on clear communication, diminish accessibility of health communications and increase health inequalities for women with English as a second language, those with a learning disability, and those with low health literacy. If this were the case, it may contravene the ‘clear information principle’ of health communications (Department of Health and Social Care, 2022; NHS Digital, 2023), increasing the potential for unintended adverse health consequences, and excluding some groups of service users. Furthermore, ‘inclusive’ language can inadvertently include those that it should not include. For example, referring to ‘parents’ or ‘families’ instead of ‘mothers’ opens up the possibility that partners and family members be included as stakeholders in a pregnancy, because the centring of the pregnant woman becomes less clear (Munzer, 2021; Gribble et al, 2022).
In summary, what the ITEMS authors recommend as ‘inclusive’ language for trans and non-binary maternity service users may, in fact, be detrimental to many other maternity service users. This means that, before making any policy changes that affect services across the board, the NHS must consider a broad range of evidence, perspectives and potential impacts. The ITEMS report also recommends that staff training should be run by LGBT advocacy groups, but such groups may hold views that do not take into account the needs of other maternity service users. The authors of this article suggest that the NHS should be cautious about commissioning training run by organisations that represent only one very small group, and about making changes that affect the service overall, without an evidence-based assessment of impacts on diverse and varied cohorts of maternity service users.
One recommendation where the authors of this article agree with the ITEMS authors is on the need for personalised care. There is broad evidence that this is in the interests of all maternity service users (NHS England, 2016; Sandall et al, 2016) and this recommendation is also supported by other research into trans and non-binary maternity service users (Hoffkling et al, 2017). Were the NHS to assess the needs of trans and non-binary maternity service users in relation to the needs of service users overall, they might consider interventions known to increase personalised care, such as investing in staffing and continuity of care (Sandall et al, 2016; Sandall, 2017). Such an approach could help to meet the specific needs of trans and non-binary maternity service users without compromising the service's ability to meet need overall.
Conclusions
The ITEMS study was commissioned to generate knowledge about the needs of trans and non-binary maternity service users. This article argues that the framing of the study is one sided and lacks conceptual clarity. The report fails to engage with relevant literature, particularly literature that takes a different perspective from that of the authors. The survey study is methodologically flawed, with clear issues in relation to sampling strategy and internal reliability. The claimed findings in the report are not always consistent with the data, and recommendations are made that are not supported by the study's own findings. Despite these issues, the report was used as the basis for a planned intervention in the NHS, at significant cost. This was only stopped as a result of concerns raised by clinicians. This highlights why carefully considered, evidence-based policy planning is important.
The authors of this article suggest that the Health and Wellbeing Alliance, NHS England and NHS Improvement should review the ITEMS report and its data, and consider whether further research is needed to inform provision of evidence-based care for trans and non-binary maternity service users and their babies. The authors further suggest that any research into trans and non-binary maternity service users should consider whether the needs of this group might be consistent with the wider need for personalised continuity of care. If research into trans and non-binary maternity service users is to inform NHS policy, it would be appropriate for it to be peer reviewed. Any recommendations should be considered in relation to the needs of maternity services users overall, and should be impact assessed before implementation.
Key points
- The trans and non-binary experiences of maternity services report contains substantial conceptual and methodological errors, which make the findings unreliable and potentially invalid.
- The report makes misleading claims that do not appear to be grounded in the data, such as an unsubstantiated claim that 30% of respondents free birthed.
- The report makes recommendations that are not based on the reported findings, such as the introduction of pronoun badges for NHS staff.
- The report recommends changes to practice relating to all service users, such as changes to language use, despite the fact that trans and non-binary people make up a very small minority of service users.
- Despite these problems, the NHS announced £100 000 expenditure on the basis of the report's recommendations, without having conducted an assessment on how this would impact other service users.
- Evidence suggests that personalised care is beneficial to trans and non-binary maternity service users (as it is to all maternity service users), so it may be the case that improving personalised care overall could improve experiences for trans and non-binary service users.