Acceptability and feasibility of a planned preconception weight loss intervention in women with long-acting reversible contraception: the Plan-it mixed-methods study

Prevalence and risk

More than 50% of women who gave birth in the NHS between 1 April 2016 and 31 March 2017 had a body mass index (BMI) of ≥ 25 kg/m² (which would be classified as being in the overweight range),1 including 26.7% who had a BMI in the obese category (BMI of ≥ 30 kg/m²). Obesity levels are expected to continue to rise, with the majority of the UK population expected to be obese by 2050. 2 Obesity places women at a greater risk of experiencing complications during the antenatal, intrapartum and post-partum periods. Maternal risks associated with obesity in pregnancy include backaches, leg pain, increased fatigue, gestational diabetes, miscarriage, pre-eclampsia, thromboembolism, slow labour progress, high caesarean section rates, post-partum haemorrhage, hypertension and maternal death. 3,4

There is also a generational dimension, as described in the Foresight report,2 with an increased risk of adverse effects on the child due to biological, social and environmental factors, including child obesity,4 developing insulin resistance5 and, for girls, having polycystic ovary syndrome (PCOS). 6 With the increasing evidence of the influence of the periconceptional period on fetal growth and long-term effects on risk factors for non-communicable diseases, it is essential that services incorporate interventions not only to reduce obesity levels of women in the preconception period and maintain weight loss during pregnancy and post partum, but also to prevent this passage of risk to the next generation. 7

The Southampton Women’s Survey reported that, in their community sample, almost half of the women recruited gained excessive weight in pregnancy according to the 2009 Institute of Medicine (IOM) guidelines,8 and there are health risks associated with gestational weight gain (GWG), irrespective of pre-pregnancy weight status. 9 However, women with overweight or obesity are at greater risk of excessive GWG in early pregnancy. The development of interventions targeting women in the preconception period, with the aim of reducing GWG, may potentially reduce the risks associated with early GWG. 10

There are clear health gains from a reduction in BMI pre pregnancy. In a population-based study in Canada11 comprising 226,958 women (64% with normal weight, 20% with overweight and 12% with obesity) with singleton pregnancies, a 10% lower preconception BMI was associated with a clinically meaningful risk reduction in pre-eclampsia, gestational diabetes, preterm delivery, macrosomia and stillbirth.

Excessive GWG is associated with weight retention in the longer term,12 and women in the post-partum period describe barriers to losing post-partum weight, including depression and a lack of weight management support. 13 The cumulative effects of excessive GWG without post-partum weight loss in multiple pregnancies may, therefore, contribute to obesity. Systematic reviews of weight loss interventions (WLIs) in the post-partum period suggest that WLIs with diet and physical activity components combined or that use diet alone, but not physical activity alone, can contribute to post-partum weight loss. 14 Furthermore, post-partum interventions that have an information or communication technology component may also be effective in increasing weight loss post partum. 15 However, further research and larger, high-quality trials with longer follow-up are required.

Guidelines for health-care practitioners providing care for women before conception and during pregnancy

The risks associated with obesity in pregnancy are reflected in the National Institute for Health and Care Excellence (NICE) guidelines for health-care practitioner consultations with all women before and during a pregnancy. 16 These guidelines make lifestyle recommendations, such as stopping smoking, taking folic acid and avoiding alcohol, and they include weight management. Specific recommendations are made for consultations with women with a BMI of ≥ 30 kg/m². Women in this group planning a pregnancy should receive information regarding the risks of obesity during pregnancy and childbirth, and advice and support to help them achieve a healthy weight prior to pregnancy, with the suggestion that losing 5–10% of their body weight would make a significant difference. Once pregnant, women with a BMI of ≥ 30 kg/m² should be offered tailored information about their diet and exercise. Women should be told about the risks that a raised BMI poses to the pregnancy but also should be told not to diet in pregnancy and that the risks will be managed by the health-care practitioners caring for them. The monitoring of GWG during pregnancy is not advised unless there is a clinical need. These recommendations are supported by the Centre for Maternal and Child Enquiries/Royal College of Obstetricians and Gynaecologists (RCOG)17 and the RCOG’s Care of Women with Obesity in Pregnancy guidelines. 18

NICE guidance also recommends that health-care practitioners be equipped with behaviour change knowledge, skills and competencies and receive communication skills training to be able to discuss weight sensitively with women. However, unlike the IOM guidance in the USA,19 there are no UK-specific guidelines for recommended optimal weight gain in pregnancy. The FIGO (International Federation of Gynaecology and Obstetrics) guidelines stress that ‘management of obesity in pregnancy should be considered in the context of a life course approach, linking with preconception and post-partum and interconception services to prevent excess weight gain before and during pregnancy’. 20

Interventions in pregnancy

There has been a considerable focus, particularly in the last decade, on weight management interventions in pregnancy. A meta-analysis21 of individual participant data from 36 randomised controlled trials (RCTs), including 12,000 participants, demonstrated that weight management interventions targeting diet and/or weight management in women across all BMI categories have not been associated with beneficial maternal or child outcomes, with the exception of caesarean section rates. However, weight management interventions in pregnancy were associated with a small but significant reduction in GWG (–0.7 kg),21 A review of reviews by Farpour-Lambert et al. 22 demonstrated moderate- to high-quality evidence of significant reduction in several outcomes in women across all BMI categories receiving a weight management intervention targeting diet and/or physical activity. These included reductions in GWG (–1.8 to 0.7 kg), excessive GWG, caesarean section rates [relative risk (RR) 0.91–0.95] and neonatal respiratory distress syndrome rates in the child (RR 0.56). However, in women with overweight/obesity, the evidence demonstrating beneficial outcomes of weight management interventions was of low to moderate quality. There was also heterogeneity among studies; study methods, intervention settings, target population and main intervention components varied between studies and interventions.

A systematic review by Agha et al. 3 across 14 studies in pregnancy and one in the preconception period demonstrated that the techniques most commonly used in interventions effective at reducing GWG included diet and physical activity counselling, motivational counselling, weight monitoring and feedback at appointments with health-care practitioners. In a meta-analysis of 89 RCTs, the authors concluded that although there was no optimal intensity, frequency, duration and delivery method that predicted success, interventions containing a group component or a healthy eating component, no matter the intensity, were more likely to be effective. 23

The timing of the initiation of a weight management intervention is important. Excessive weight gain in the first trimester is predictive of excessive weight gain throughout pregnancy24 and gestational diabetes mellitus. 25 This means that starting an intervention in pregnancy may well be too late, as the first point of contact with a health-care practitioner in pregnancy is usually towards the end of the first trimester. Interventions most effective at reducing GWG are those initiated in early pregnancy rather than those delivered later in pregnancy,3 and in Farpour-Lambert et al. ’s22 meta-analyses investigating the effects of lifestyle interventions on GWG or post-partum weight retention, the authors concluded that weight loss prior to pregnancy is probably required to achieve both GWG goals and optimal pregnancy outcomes.

Given the limited success of interventions for obesity in pregnancy, the issue of timing and the key impact of health behaviours in the very early days of conception, attention has turned to the preconception phase as critical to improving maternal and child health. However, owing to the paucity of WLIs in the preconception stage, evidence of their overall effectiveness or of what intervention components may be beneficial is limited.

The preconception period

The World Health Organization defines preconception care (PCC) as:

. . . the provision of biomedical, behavioural and social health interventions to women and couples before conception occurs. Its ultimate aim is to improve maternal and child health, in both the short and long term.

World Health Organization. 26 Reproduced with permission under the CC BY-NC-SA 3.0 IGO licence

As highlighted in the 2018 series on preconception health in The Lancet27–29 and by Stephenson et al. ,30 preconception is an ‘underappreciated period in the life course when health, behavioural, and environmental exposures can have far-reaching consequences, not only for pregnancy outcomes but also for health across generations’. 30 One of the complexities, critical to the clarity of focus of any programme or intervention, is defining the meaning of preconception. Stephenson et al. ,27 suggested that there were three perspectives to consider in a discussion of preconception: (1) biological perspective (the days or weeks before the embryo develops), (2) individual perspective (the weeks or months before conception for individuals who have a conscious intent to conceive) and (3) public health perspective, which includes the months or years before conception. Hill et al. 31 recognised the value of this classification but felt that it lacked enough specificity to identify the population to be defined and targeted in interventions. They proposed four perspectives, each based on a particular combination of defining attributes: reproductive age, man/woman, not pregnant, sexually active, intent to conceive. Some differences from Stephenson’s classification are that the ‘public health’ group is subdivided into those of reproductive age (public health) or younger (life course) and that the individual perspective is subdivided into those with intent and those without intent (potential preconception). Many preconception interventions target the ‘intentional preconception’ group, but this excludes those who are sexually active, who may or may not be using contraception, but have not made a conscious decision to conceive.

Healthy diet/nutrition and weight management have been identified as the top two priorities in international preconception research. 32 There are significant gaps in preconception nutritional intake internationally. 33 In the UK it is estimated that 77% of women aged 18–25 years have dietary intakes below Reference Nutrient Intake daily recommendations for iodine and that 96% of women of reproductive age have intakes of iron and folate below the daily recommendations for pregnancy. 27 The situation is not greatly improved in women who are planning a pregnancy: the results of the Southampton Women’s Survey showed that only a very small proportion of women planning a pregnancy followed the recommendations for nutrition and lifestyle in the preconception period. For example, only 2.9% of women who became pregnant [95% confidence interval (CI) 1.2% to 6.0%] complied with recommended levels of folic acid supplementation and drank four or fewer units of alcohol a week, compared with 0.66% (95% CI 0.52% to 0.82%; p < 0.001) of those women who did not become pregnant. 34

The lack of engagement with the recommendations may well be due to a lack of knowledge relating to preconception health or, potentially, beliefs relating to the behaviour, as there are indications that smoking, a more commonly known risk behaviour, does reduce before pregnancy. 35 Preconception education and counselling can improve maternal knowledge, self-efficacy and risk behaviour, but the impacts on anthropometric measures and pregnancy outcomes are less clear. 36,37 There may also be demographic differences underlying differential behavioural response to information; for example, younger preconceptional women and women with children were less likely to engage in preconception health behaviours,38 and women with higher levels of education were more likely to engage. 39,40

In England, the Preconception Partnership of key stakeholders is working to operationalise policy to ensure that PCC becomes a more integrated part of routine practice. It has proposed an annual report identifying progress on a set of core metrics for agencies, reflecting planning and preparation for pregnancy at an individual and public health level. 30 This would include improving the health environment, putting the importance of preconception health on the school curriculum and embedding it in policy initiatives. On an individual level, the goals include normalising conversations about pregnancy intentions in health care, increasing support for behaviour change to improve preconception health and training health-care providers to have conversations about preconception health.

One of the issues that is often cited as the reason for the lack of preconception health care is the number of pregnancies that are planned, often assumed to be small. In a cross-sectional survey35 of 1288 women conducted in a London maternity centre, 73% had clearly planned their pregnancy and 24% were ambivalent, with only 3% of pregnancies unplanned. These figures would suggest that planning is more common than is often assumed. There are also different types of ‘planning’, for example women with high levels of pregnancy planning who take up interventions but also those who plan but describe themselves as having poor awareness of preconception actions. 41 However, if the interventions suggested by the Preconception Partnership were put in place and preconception health conversations were more integrated into routine care, the level of positive intention and planning could become less of a barrier.

Preconception and obesity

Until very recently, few studies have targeted obesity in the preconception period; interventions in the preconception period have reported programmes for very specific populations, such as gastric bypass patients,42 those with diabetes43 or couples seeking in vitro fertilisation (IVF) treatment,44 or have been small scale. 45 Two Cochrane reviews46,47 investigating the effectiveness of preconception and antenatal health programmes and interventions in improving pregnancy outcomes or reducing weight in obese women identified no eligible studies in the preconception phase. In a systematic review48 of the impact of preconception lifestyle interventions on live birth, birthweight and pregnancy rate, six of the included studies reported participants’ weight and BMI, five of which were in the context of assisted reproduction. The Strong Healthy Women (SHW) intervention49 (a six-session preconception health course with a social cognitive approach50 to behaviour change in preconceptional and interconceptional women living in low-income rural communities) was excluded from the reviews as it was not possible to identify outcomes by BMI category (see Studies in progress).

A more recent scoping review51 of RCTs of behavioural interventions supporting women of childbearing age in the prevention and treatment of overweight and obesity identified only one intervention in the preconception period, the Prepare study protocol. 52 The Prepare trial is a pre-pregnancy intervention in the USA for women with a BMI of ≥ 27 kg/m² identified through routine data. The intervention is based on social cognitive theory, and the intervention group received 6 months of weekly weight management calls from a health coach followed by monthly calls for 1 year. The primary outcome was GWG but with prenatal weight loss as a secondary outcome. In the results, now published,53 the women receiving the intervention lost an average of 3.5% of their weight before becoming pregnant (control arm 0.5%; p < 0.001). However, by the end of the third trimester, the GWG was the same for both groups, leading the authors to suggest that an intensive weight management approach is needed beyond preconception, and throughout pregnancy, to manage GWG.

One potential reason for the lack of evidence in relation to preconception and obesity is that weight and weight management often do not feature in preconception health programmes that have been evaluated. In a systematic review of studies identifying factors related to preconception health behaviours,38 weight status was measured in only 3 of the 24 included studies and, therefore, was not identified in the review as one of the six categories of factors. Similarly, in a scoping review of preconception health interventions, only 4 of the 29 included programmes identified weight/obesity as a risk factor to be addressed. 37 More recently, in a published trial protocol concerning the uptake of PCC in the Netherlands,54 the primary outcome was change in lifestyle behaviours (e.g. folic acid use, smoking and alcohol use) and secondary outcomes were pregnancy outcomes (e.g. miscarriage, preterm birth, gestational diabetes) and the uptake of PCC, but weight management was not included. Overall, the quantity and quality of the evidence in this field is very limited. 36

Barriers to engaging women in preconception weight loss interventions

With the rising rates of women of childbearing age who are obese,55 the development of effective preconception WLIs for women with overweight/obesity may provide an important step in reducing the health risks to mother and child. As with pregnancy,56 the preconception period may be considered a ‘teachable moment’, during which efforts may be made to positively influence women’s diet and health behaviours. However, there are barriers to overcome. As already highlighted, preconception is a difficult time to identify. ‘Planning’ a pregnancy is often declared only once the pregnancy has been confirmed; prior to this, it is something regarded by many as deeply personal and essentially private. 57 Furthermore, women may not be aware of the importance of preconception health, and they may not perceive there to be risks or that the risks are relevant to them. 45,57 Specific issues for women with overweight/obesity and who are planning a pregnancy include poor uptake of health activities, inaccurate self-categorisation of weight, unsuccessful weight loss attempts and inadequate advice regarding pre-pregnancy weight loss. 58 Often, the complex lifestyle changes required for weight loss before pregnancy are challenging to achieve, but women who recognise that they have knowledge gaps about the impact of obesity in pregnancy are keen to receive information about antenatal, intrapartum and post-partum risks. 43

The Health Survey for England (HSE)59 reports that obesity rates for women are higher in areas of highest Index of Multiple Deprivation (39%) than in the least deprived areas (22%). Evidence of obesity rates in ethnic minority groups is scarce, with the exception of the 2004 HSE. 60 These data suggest that obesity rates vary among ethnic groups, with higher rates in black and Pakistani groups and lower rates in Chinese groups than in the general population. 61 Health risks such as diabetes are observed in ethnic groups at lower rates of overweight/obesity than the white European population,62 and, therefore, reducing obesity levels in ethnic minority groups could be of increased relevance. Uptake and retention of programmes may vary in women of different ethnic backgrounds. In a process evaluation of a postnatal WLI in an ethnically diverse population, non-white British/Irish women were less likely to attend a commercial weight loss management group than white British/Irish women. The women were more likely to describe barriers to attending the sessions, such as time, access and child-care needs, than white British/Irish women. 63 Furthermore, women from ethnically diverse communities reported that they had modest or poor awareness of preconception health issues and that there was little culture of preconception preparation. 64 Therefore, cultural differences would need to be considered and addressed when targeting women from ethnic minority groups and when designing an effective intervention for ethnically diverse populations.

Health-care practitioners also experience barriers to raising weight management during preconception and pregnancy-related consultations, including lack of skills, time, financial reimbursement, sensitivity of topic and confidence in the available interventions. 65–67 Interviews with UK health-care practitioners showed that there was a low awareness of preconception health issues and confusion about responsibility for delivery of PCC. 35 Taking a broader view of health-care provision, these barriers resonate with those found in a systematic review of the barriers to and enablers of health-care practitioners in delivering behaviour change interventions (diet, physical activity, alcohol reduction, smoking cessation and weight management). 68 Four themes emerged as both barriers and enablers: (1) perceptions of the knowledge or skills needed to support patients’ behaviour change, (2) perceptions of their professional role, (3) beliefs about resources needed and (4) practitioners’ own health behaviour. Cross-disciplinary barriers included a perceived lack of time, negative attitudes towards patients and perceptions of patients’ motivation, including a lack of prioritisation of health behaviour change. 68 Training, context and attitudes towards the intervention were the enablers identified, and any programme of preconception health will need to use these to address individual and systemic barriers if provision is going to change.

Studies in progress

Several trials are in development or pre-reporting that are taking a range of approaches to weight management in the preconception period in different cultural contexts.

Strong Healthy Women was originally a six-session intervention delivered to 692 preconceptional and interconceptional women across 12 weeks in low-income and rural communities. They focused on managing stress, physical activity, nutrition (including folic acid supplementation), preventing gynaecologic infection, tobacco exposure and alcohol use (weight was not included as a focus). Their intervention adopted a social cognitive approach to behaviour change,50,69 identifying self-efficacy, motivation and intention to change as important determinants of behaviour change. Participants receiving the intervention had improved pre–post intervention outcomes relating to nutrition, physical activity and stress management compared to the control group. 70,71 At the 12-month follow-up, the intervention group had reduced weight (mean group difference of 4.33 lb), reduced BMI (mean difference of 0.75 kg/m²) and lower GWG in those who became pregnant (mean difference of 17.95 lb). 49 Despite a small sample size, these were promising results. However, the face-to-face nature of intervention delivery was considered resource-intensive and expensive by the research team, and the time investment required for participation was burdensome for some women. The study team have therefore explored a ‘smart’ version, using qualitative methods to determine which components women would be happy to receive digitally and which to retain face to face. 72

The INTER-ACT study in Belgium73 is an interpregnancy study designed to reduce pregnancy complications in women whose pregnancy weight gain exceeds the IOM-recommended levels. The lifestyle intervention incorporates an e-health application (app) and coaching delivered partially post partum and then during the following pregnancy. The women’s experience of using the app suggests that combining a personalised app with coaching is a positively received intervention.

The Jom Mama trial in Malaysia74 is a population-based pre-pregnancy intervention that recruits as couples get married. It is designed to enhance healthy dietary choices, increase exercise and manage stress. It is grounded in the theory of triadic influence and uses behaviour change counselling combined with WhatsApp chat support and an e-health habit formation app.

TOP Mums75 is a preconception intervention designed to reduce perinatal morbidity for women with a BMI of > 25 kg/m² planning to conceive in the next year and is due to report imminently. It is a 26-week programme incorporating motivational interviewing, personalised goals and a health intervention, ‘Smarter Pregnancy’. BMI is the primary outcome from baseline to 6 weeks post partum, with GWG one of the secondary outcomes.

Other studies that will involve measurement of pre-pregnancy weight, but do not evaluate population based pre-pregnancy interventions, evaluate the quality of the preconception diet (Healthy for my Baby);76 the relationship between maternal weight and baby weight;77 reducing pre-pregnancy and pregnancy weight gain in the context of depression;78 and reducing the reoccurrence of gestational diabetes. 79

Users of long-acting reversible contraceptives

The complexities of identifying the preconception population mean that it is difficult to engage with people at the right time to have them consider taking part in a weight reduction programme to improve their preconception health. Women who use long-acting reversible contraception (LARCs) and who require removal of the device to become pregnant represent a unique group in which there is an opportunity for intervention. However, at this point in their reproductive decision-making, it may be difficult to ask women to delay conception through continued use of LARC and engage in weight loss programmes, raising pragmatic and ethical issues for both an intervention and any research study designed to establish effectiveness. A small feasibility study of an intensive WLI offered to women with a BMI of 30 kg/m² or more attending for LARC removal80 demonstrated that some women were willing to consider delaying LARC removal for 6 months to participate. This small evidence base suggests that there may be an interest in weight loss and a willingness to delay LARC removal in relevant populations. However, with high rates of non-participation and attrition from the programme, it has not yet been established what, if anything, the nature of an acceptable intervention would be.

To answer this question, a mixed-methods approach is required, incorporating the use of routine data, qualitative data collection and analysis, and giving a central role to stakeholders. This will lead to a better understanding of the LARC pathway from an individual and population perspective and its interface with weight management. LARC users’ and health-care practitioners’ experiences of LARC services, decision-making and management of weight around pregnancy will be explored alongside their views of the ethical and methodological issues associated with the timing of informed consent and a potential preconception WLI. Health-care practitioners’ LARC practice and consultation patterns regarding LARC use and removal will be identified utilising data sets, collected routinely across the four UK nations, to compare the population across the different health-care settings as well as over time, taking into account factors such as the impact of different general practice incentives on activity and recording. 81 All of this information will be critical to consider when developing a future intervention.

Preconception weight loss as a complex intervention

A preconception weight loss programme would be described as a complex intervention in the terms of the Medical Research Council framework for the development and evaluation of complex interventions,82,83 which involve multiple interacting components in a complex context and require behaviour change on the part of practitioners and participants. The Medical Research Council framework identifies four phases in the development and evaluation of complex interventions: development, feasibility, evaluation and implementation. The Plan-it study falls into the development phase, exploring an idea for a potential intervention that could be delivered in the NHS. There are, clearly, many generic weight loss programmes that women could access in the preconception phase. The question here is whether or not it is acceptable and feasible to develop a specific preconception weight loss programme that includes a delay to a LARC removal as the entry point to the programme.

As described by O’Cathain et al. ,83 the development phase is a dynamic, iterative process that includes seeking feedback from key stakeholders on intervention ideas, reviewing evidence, drawing on theories, exploring context and undertaking some primary data collection. Through this process, the strengths of the idea and the problems can be identified and used to further refine possible intervention designs. Once an idea for an intervention has been developed that is acceptable and feasible in its formulation, it moves to the feasibility phase, when it is tested in practice, followed by an evaluation, potentially by a RCT, to establish its efficacy.

Analysis of policy documents

A review of policies, best practice guidelines and other clinical/advisory documents pertaining to the use (and particularly the removal) of LARCs in the UK was conducted to understand how services are expected to approach discussions of weight loss with women with overweight/obesity, how LARC treatment pathways currently operate, guidance on health behaviours prior to conception and the practical/ethical challenges to successful service delivery in current service structures, including equity of access to interventions (for detailed methodologies, see Chapter 4).

Engagement with LARC users

Qualitative surveys using closed and open-text questions were utilised to understand women’s experiences of discussing weight with health-care practitioners; how, where and when women access preconception WLIs; women’s knowledge of the risks of overweight/obesity in pregnancy; the barriers to and facilitators of the introduction of a WLI at LARC removal appointments; and the preferred components of a potential intervention. Detailed methodologies are described in Chapter 4.

Aims of the chapter

The Plan-it routine data WP used anonymised routinely collected data from SHCs in England, Wales and Scotland and the CPRD linked to the Pregnancy Register to determine the most appropriate LARC removal settings, the annual numbers of potential participants available to be recruited and an indicative time frame for recruitment.

The objectives were to set up access to anonymised data from multiple health settings to:

• understand the pattern of LARC use (removal/insertion/in situ) to identify opportunities to intervene

• identify women, who request LARC removal and subsequently become pregnant, who would be eligible for recruitment to a WLI study

• report the annual number of women in the UK requesting removal of LARC without replacing it with an alternative prescribed contraception

• identify events in GP and hospital records to explore time from LARC removal to conception or appointments relating to difficulties conceiving (if possible).

The outcomes identified were:

• rates of women in the UK who request LARC removal and subsequently have a pregnancy from routine data

• identification of opportunities to intervene in the preconception pathway.

Study design/setting

Study population

Statistical methods

For the individual anonymised data from CPRD, rates of fitting and removal of LARC in general practice were calculated and reported over time (either quarterly or annually, depending on numbers). Rates were calculated as the number of LARCs fitted and removed as a proportion of all women of reproductive age.

Trends in rates were examined by country, LARC type and, where available, attendance type (pre-booked vs. walk-in consultations). Similarly, we examined trends using the Welsh SHC data. This enabled a comparison of the case mix of women recorded for a LARC removal in general practice and those who visit a SHC by age, ethnicity, BMI and deprivation. The data from SHCs in Scotland and England were in aggregate format at NHS Sexual and Reproductive Service provider level and reported rates of fitting and removal annually. Trends in rates by age group at time of fitting or removal, change of contraception method from/to a LARC and, if recorded, BMI category and deprivation quintile were explored. These rates were compared with those arising from the SHC data in Wales. The quality of recording of BMI in all data sets was explored. Previous work in the CPRD shows that completeness of BMI has increased over time (to around 77%) and was higher in female individuals, especially those of reproductive age. 92

Analysing these data sets identified variation in the numbers, pattern and duration of use of LARCs in the different health settings, geographical areas (rural/urban) and demographic groups. It was possible to consider what opportunities (e.g. consultation types, frequency of consultations) are available to intervene in the different service delivery designs across the UK. The routine data analysis determined the most appropriate LARC removal settings, the annual numbers of potential participants available to recruit and an indicative time frame for recruitment.

For patients attending their GP for the fitting or removal of their LARC, the data allowed exploration of the duration of LARC use prior to removal and the changes to contraception use over time; through linking to the Pregnancy Register algorithm, a pregnancy episode related to the women in the cohort (estimated start of pregnancy) was flagged. Accessing these data will enable a broader understanding of the population for whom this intervention will be targeted and, potentially, identify those who had a LARC removed for the purpose of planning a pregnancy. An examination of how time to conception may differ between BMI and age categories was also possible. For women whose pregnancy was identified following a LARC removal, the natural distribution of the time between the two events (LARC end and pregnancy start) was examined to assess whether or not a rule could be applied to indicate that the pregnancy and LARC removal were associated.

Although the data between GP surgeries and SHC data could not be linked, the recording of a LARC removal in a SHC setting was explored in the GP notes. Previous work using an alternative primary care data source (The Health Improvement Network) identified that 24% of LARC-related records in primary care came from SHC letters. 93 The reporting of routine data was in accordance with RECORD (REporting of studies Conducted using Observational Routinely-collected Data). 94

Defining the study population in the Clinical Practice Research Datalink

The study population generation is reported in Figure 2. The study population was identified using the CPRD Patient file, which included just over 20 million patient records. Practices were excluded from the study population if the practice’s last contribution date was before 2009 or their up-to-standard date was after 2018. Practices with < 1 year between their last contribution date and their up-to-standard date (< 1 year of data available) and those not meeting the criteria in the specified years were also excluded. The following patient groups were excluded: male patients; patients of unknown or indeterminate gender; patients whose data were outside the up-to-standard date; patients whose year of birth was outside 1961–2002; patients who had died before 2009; patients registered with the practice after 2018; and patients whose current period of registration with the practice was after 2018.

FIGURE 2.

Participant flow diagram. Reproduced with permission from Channon et al. 91 This is an Open Access article distributed in accordance with the terms of the Creative Commons Attribution (CC BY 4.0) license, which permits others to distribute, remix, adapt and build upon this work, for commercial use, provided the original work is properly cited. See: https://creativecommons.org/licenses/by/4.0/. The figure includes minor additions and formatting changes to the original figure.

Reproduced with permission from Channon et al. 91 This is an Open Access article distributed in accordance with the terms of the Creative Commons Attribution (CC BY 4.0) license, which permits others to distribute, remix, adapt and build upon this work, for commercial use, provided the original work is properly cited. See: https://creativecommons.org/licenses/by/4.0/. The figure includes minor additions and formatting changes to the original figure.

This left 2,632,871 women at reproductive age in the study denominator. However, 2,314,831 were excluded as they were not in the predefined study population (i.e. with one LARC event), leaving 318,040 eligible women of reproductive age in the study population. The rate of women at reproductive age in the study denominator was constant over the time period 2009–18 (Table 3).

Objective 1: understand the pattern of LARC use (removal/insertion/in situ) to identify opportunities to intervene

Women attending general practice for a LARC-related consultation

To identify women who attended a GP surgery for a LARC event, a combination of Read and prescription codes were used. The flow chart (see Figure 2) shows that, from using these codes for LARC, 929,099 codes were present in the study population of 315,755 women. Table 4 shows these data broken down by LARC consultation type and type of LARC. Women in our study population could experience more than one code over the study period. For example, among the 127,909 women who had at least one LARC in situ code, 49,285 (39%) had more than one code recorded over the study period. Of those 269,999 women who had a LARC insertion, 182,090 (67%) had more than one LARC insertion code recorded over the study period. Of those 108,987 women who had a LARC removal, 12,615 (12%) had more than one LARC removal code recorded during the study period.

TABLE 4 - The LARC events by consultation and LARC type

n = 9062 codes were double counted owing to three codes [IUD – not otherwise specified (NOS), subcutaneous contraceptive NOS, and subcutaneous contraceptive] being coded as both LARC in situ and LARC insertion.

Reproduced with permission from Channon et al. 91 This is an Open Access article distributed in accordance with the terms of the Creative Commons Attribution (CC BY 4.0) license, which permits others to distribute, remix, adapt and build upon this work, for commercial use, provided the original work is properly cited. See: https://creativecommons.org/licenses/by/4.0/. The table includes minor additions and formatting changes to the original table.

	Unique number of events	Number of women with at least one event
LARC consultation typea	n = 938,161
In situ	222,555	127,909
Insertion	592,402	269,999
Removal	123,204	108,987
Type of LARC	n = 934,903
IUD	411,514	167,008
IU system	308,058	131,889
Implant	215,331	125,497

The number of women in the study population decreased over time due to the number of practices contributing to the CPRD declining; the rate of women per practice was constant (Figure 3 and Table 5). The proportion of women with one or more LARC consultations decreased over time, from 2.2% in 2009, which is approximately 42 LARC events per practice, to 1.3% in 2018, 24 LARC events per practice. By contrast, the Wales SHC data show a different picture (Figure 4); in 2018 the rate of women with SHC contact for LARC use out of contacts for any contraception was higher, at 24%.

FIGURE 3.

LARC users over time: CPRD (2009–18).

TABLE 5 - Number of LARC events and type of LARC between 2009 and 2018: CPRD

Year of first LARC consultation for women in the study population (n = 315,755). Percentage based on the number of women at the study denominator.

Percentage based on the number of consultations at the study denominator.

Year	Study denominator		Study population
	Women aged 16–48 years (n)	Consultations (n)	Year of first LARC consultationa		LARC consultation typeb						Type of LARC
					In situ		Insertion		Removal		IUD		IU system		Implant
			n	%	n	%	n	%	n	%	n	%	n	%	n	%
2009	1,413,791	2,253,647	61,440	4.3	31,074	1.4	65,523	2.9	11,293	0.5	58,232	2.6	27,390	1.2	21,814	1.0
2010	1,408,118	2,385,300	51,850	3.7	31,479	1.3	75,430	3.2	12,402	0.5	58,576	2.5	35,194	1.5	25,222	1.1
2011	1,380,332	2,474,914	43,456	3.1	30,540	1.2	72,290	2.9	13,891	0.6	56,355	2.3	33,977	1.4	26,207	1.1
2012	1,355,832	2,561,659	36,994	2.7	28,768	1.1	72,904	2.8	14,150	0.6	51,866	2.0	37,175	1.5	26,387	1.0
2013	1,330,823	2,485,337	32,975	2.5	25,384	1.0	71,152	2.9	14,735	0.6	46,513	1.9	38,469	1.5	26,058	1.0
2014	1,244,434	2,353,969	26,816	2.2	22,996	1.0	64,713	2.7	14,283	0.6	41,533	1.8	36,231	1.5	23,945	1.0
2015	1,108,991	2,149,417	21,344	1.9	18,558	0.9	55,091	2.6	13,410	0.6	33,673	1.6	32,266	1.5	20,780	1.0
2016	914,755	1,751,839	15,662	1.7	13,258	0.8	42,554	2.4	10,985	0.6	24,542	1.4	25,507	1.5	16,365	0.9
2017	812,739	1,479,553	13,268	1.6	10,972	0.7	38,001	2.6	9585	0.6	21,364	1.4	21,983	1.5	14,881	1.0
2018	741,453	1,304,630	11,950	1.6	9526	0.7	34,744	2.7	8470	0.6	18,860	1.4	19,866	1.5	13,672	1.0

FIGURE 4.

LARC users over time: Wales SHC data (2009–18). Note: data from 2012 to 2015 are unreliable due to under-reporting by some health boards.

To further understand LARC use in general practice and SHCs, patterns were examined by LARC consultation type (insertion, in situ, removal), type of LARC used (IUD, IU system, implant) and age group type to identify opportunities to intervene. SHC data were obtained from open-access sources, and extracts are presented in Appendix 4. LARC insertions are most frequently performed in general practice (see Figure 4) (see Appendix 3 for details of type of LARC use over time).

Objective 2: identify women requesting LARC removal who subsequently become pregnant who would be eligible to recruit to a weight loss intervention study

To identify eligible women to recruit to a WLI study, several stages need to be established (Figure 5):

• LARC use (objective 1)

• pregnancy within 1 year and 3 months

• codes to refute or confirm that the pregnancy was planned

• BMI status.

FIGURE 5.

Stages of identifying women for a WLI study.

LARC removal for the purposes of planning a pregnancy

In the 24,073 pregnancy events from 22,326 (7.0%) women in the study population, 11,381 (47.3%) conceptions occurred within the window of 1 year and 3 months (456 days), and 3090 (12.8%) occurred outside the window of 1 year and 3 months. With regard to the 411 (1.7%) pregnancy events with a removal code between pregnancy start and pregnancy end, we considered them as an unplanned pregnancy if the removal code was within 7 days after pregnancy start; otherwise, we would investigate other LARC code (if any) or exclude the pregnancy event and the removal code in between. For the 9191 (38.2%) pregnancy events with no removal code between start and end date, we included the event if the pregnancy start date was within the window of 1 year and 3 months after the last LARC in situ or LARC insertion date; otherwise, the pregnancy event would be excluded.

Among the 315,755 women in the study population with a LARC event, 15,297 women had at least one pregnancy, with 16,394 pregnancy events (Table 6). Of those, 4753 (29.0%) pregnancy events did not have a LARC removal code before pregnancy started and 299 (1.8%) had a removal code between pregnancy start and end. The number and percentage of the 16,394 pregnancy outcomes are described in Table 6. For the remaining 11,342 events, the median time to conception was 109 days (25th to 75th centiles = 47 to 220 days) (Figure 6).

TABLE 6 - Pregnancy outcomes (n = 16,394)

TOP, termination of pregnancy.

Pregnancy outcome	Frequency (n)	%
Live birth	8910	54.3
Stillbirth	33	0.2
Miscarriage	1700	10.4
TOP	241	1.5
Miscarriage or TOP or ectopic	1231	7.5
Ectopic	163	1.0
Molar	6	0.0
Blighted ovum	5	0.0
Unspecified loss	107	0.7
Delivery based on a third-trimester pregnancy record	278	1.7
Delivery based on a late-pregnancy record	46	0.3
Outcome unknown	3674	22.4
Total	16,394

FIGURE 6.

Conception time for events within 456 days (n = 11,342).

We have defined 479,044 different scenarios or pathways to a pregnancy event or not among 317,684 women (note that women can have more than one scenario). The number of scenarios and number of women in each group are summarised in Table 7.

TABLE 7 - Defining the groups based on LARC use, events related to planning a pregnancy and contrary events, and conception

SD, standard deviation.

LARC use	Read code indicating an event	Pregnancy	Group	Number of scenarios	Number of women	Age (years), mean (SD)
1: LARC removal/inserted/in situ	A Read code to indicate the pregnancy was being planned (planned pregnancy code or trying/difficult to get pregnant) either between a LARC (removal/inserted/in situ) and pregnancy start, or between a pregnancy start and end	Yes	1: planning a pregnancy	1635	1616	29.0 (5.81)
2: LARC removal/inserted/in situ	No Read code to indicate that the pregnancy was planned (planned pregnancy code or trying or difficult to get pregnant) or not planning a pregnancy (alternative contraception and menopause). No Read code to indicate that this pregnancy was unplanned (unplanned pregnancy code)	Yes	2: possibly planning a pregnancy	10,902	10,387	28.5 (6.06)
3: LARC removal/inserted/in situ	A Read code to indicate that a pregnancy was not being planned (alternative contraception and menopause)	Yes	3: probably not planning a pregnancy	3851	3761	26.5 (5.98)
4: LARC removal/inserted/in situ	A Read code to indicate that a pregnancy was being planned (planned pregnancy code or trying/difficult to get pregnant)	No	1: planning a pregnancy	4871	4717	30.6 (7.26)
5: LARC removal	After a LARC removal code, no Read code or a Read code present to indicate that the woman was planning/not planning a pregnancy	No	Not enough information	73,290	69,455	32.4 (9.03)
6: LARC inserted/in situ	No code present to indicate that the woman had a LARC removed or was planning/not planning a pregnancy	No	3: probably not planning a pregnancy	379,495	277,144	33.0 (9.22)
7: LARC removal/inserted/in situ	Code to indicate a pregnancy was not being planned (e.g. alternative contraception), menopause indication (e.g. blood tests)	No	3: probably not planning a pregnancy
8		Yes	Unable to code	6	6
9		No	Unable to code	59	59

Table 8 shows, for each group, the number of LARC events, the proportion with a valid BMI measurement broken down by BMI categories: ≥ 25 kg/m² (overweight and obese), ≥ 30 kg/m² (obese) and ≥ 40 kg/m² (morbidly obese). Of all 474,044 LARC events with a valid BMI measurement, a small proportion (1.4%, n = 6506) were removed for planning a pregnancy, 10,902 (2.3%) were possibly planning a pregnancy and 383,346 (80.9%) were probably not having a pregnancy; 73,290 (15.5%) could not be grouped because of insufficient information. A total of 62% of women had a BMI recorded within 3 years of their LARC event, and this was similar across all groups (range 60–67%), with a 5% difference between women not planning a pregnancy and those planning a pregnancy (62% vs. 67%). Around 50% of women had a BMI of ≥ 25 kg/m², around 28% had a BMI of ≥ 30 kg/m² and < 5% had a BMI of ≥ 40 kg/m². Among those planning a pregnancy, 54% had a BMI of ≥ 25 kg/m², 28% had a BMI of ≥ 30 kg/m² and < 5% had a BMI of ≥ 40 kg/m².

TABLE 8 - Completeness of BMI and BMI categories

Reproduced with permission from Channon et al. 91 This is an Open Access article distributed in accordance with the terms of the Creative Commons Attribution (CC BY 4.0) license, which permits others to distribute, remix, adapt and build upon this work, for commercial use, provided the original work is properly cited. See: https://creativecommons.org/licenses/by/4.0/. The table includes minor additions and formatting changes to the original table.

Group	Events, n (%)	With BMI			BMI ≥ 25 kg/m2			BMI ≥ 30 kg/m2			BMI ≥ 40 kg/m2
		Category	n	%	Category	n	%	Category	n	%	Category	n	%
1: planning a pregnancy	6506 (1.4)	Yes	4328	66.5	< 25	2014	46.5	< 30	3130	72.3	< 40	4124	95.3
		No	2178	33.5	≥ 25	2314	53.5	≥ 30	1198	27.7	≥ 40	204	4.7
2: possibly planning a pregnancy	10,902 (2.3)	Yes	7034	64.5	< 25	3480	49.5	< 30	5368	76.3	< 40	6844	97.3
		No	3868	35.5	≥ 25	3554	50.5	≥ 30	1666	23.7	≥ 40	190	2.7
3: probably not planning a pregnancy	383,346 (80.9)	Yes	237,155	61.9	< 25	112,472	47.4	< 30	176,619	74.5	< 40	227,618	96.0
		No	146,191	38.1	≥ 25	124,683	52.6	≥ 30	60,536	25.5	≥ 40	9537	4.0
Not enough information	73,290 (15.5)	Yes	44,286	60.4	< 25	20,629	46.6	< 30	32,535	73.5	< 40	423,89	95.7
		No	29,004	39.6	≥ 25	23,657	53.4	≥ 30	11,751	26.5	≥ 40	1897	4.3
Total	474,044	292,803 (62%)

Objective 3: report the annual number of women in the UK requesting removal of LARC without replacing it with an alternative prescribed contraception

For this objective, we are interested in events of LARC removal without replacement with an alternative contraception (see Appendix 2, Table 23, for alternative contraception code). Of the 123,204 LARC removal events in the study population, 24,777 were followed by an alternative contraception code (20.1%). The time between LARC removal and alternative contraception ranged from 0 to 35.84 months, with a mean of 4.38 months (standard deviation 4.76 months) and a median of 2.73 months (interquartile range 0–7.82 months) (Figure 7).

FIGURE 7.

Time between LARC removal and alternative contraception (months) (n = 24,777).

Based on this information, and without a clear cut-off point identified from clinical practice, we decided to use 4 months as a cut-off point and considered contraception events outside this window as LARC removal events without replacement with an alternative contraception. Overall, there were 14,368 (58.0%) LARC removal events without replacement with an alternative contraception (Table 9).

Objective 4: identify events in general practitioner and hospital records to explore time from LARC removal to conception or appointments relating to difficulties conceiving (if possible)

For these 24,073 pregnancy events, we aimed to calculate an accurate time from LARC removal to conception. We could not calculate an accurate conception time for 9602 (39.9%) pregnancy events [411 events had a LARC removal event between the estimated pregnancy start and end date; 9191 events had no LARC removal code (presumed to be because the LARC was removed elsewhere) before a pregnancy start]. For the remaining 14,471 (60.1%) valid pregnancy events following a LARC removal, the median time to conception following LARC removal was 160 days (25th to 75th centiles = 62 to 398 days) (see Figure 11). In total, 1861 (12.9%) were within the first month, 3562 (24.6%) were within the first 2 months and 5021 (34.7%) were within the first 3 months following LARC removal (Figure 8) (see Appendix 3 for the details of time to conception for women in different age and BMI categories, as well as additional analysis on further contraception patterns).

FIGURE 8.

Time from LARC removal to pregnancy start (days) (n = 14,471).

Study objectives to be addressed

Study objectives 4 and 5 are addressed directly:

1. 4. to assess the willingness of health-care practitioners to raise weight loss in consultations and recruit eligible women to the intervention

2. 5. to assess LARC users’ views as to the acceptability and feasibility of the proposed intervention and of future research.

The findings from this chapter will also map onto study objectives 2 and 3:

1. 2. to identify opportunities to identify women with overweight/obesity who plan to have LARC removal for the purpose of planning a pregnancy and opportunities to intervene with a potential preconception WLI

2. 3. to identify suitable and acceptable interventions to be incorporated into a preconception WLI.

Aims of the chapter

To answer the study objectives, the aims were to:

1. develop an understanding of how health-care services are expected to approach discussions of weight loss with women with overweight/obesity

2. develop an understanding of how LARC treatment pathways currently operate through an analysis of policy documents

3. assess current practice of how health-care practitioners approach discussions of weight loss with women with overweight or obesity and the inter-relationship between discussions about overweight/obesity and family planning through qualitative surveys with LARC users and health-care practitioners

4. assess the barriers to and facilitators of, including practical and ethical considerations, introducing weight and the option of delaying LARC removal during health-care consultations from the perspective of LARC users and health-care practitioners

5. identify feasible opportunities to intervene with a potential preconception WLI

6. assess the desired characteristics of a potential preconception WLI from the perspective of LARC users and health-care practitioners.

Analysis of policy documents

A review of policies, best practice guidelines, other clinical/advisory documents and grey literature pertaining to the use (and particularly the removal) of LARCs in the UK was conducted to meet aims A and B. Relevant data were identified via searches of websites and/or online document repositories belonging to four key organisations: Faculty of Sexual and Reproductive Healthcare of the Royal College of Obstetricians & Gynaecologists (FSRH), RCOG, NICE and British Pregnancy Advisory Service (BPAS) (the search strategy is detailed in Appendix 6). The initial results were presented to the SSC to ensure that all relevant policies had been included. A full-text analysis of included documents was conducted: key statements and/or excerpts relating to aims A and B were extracted.

Qualitative online surveys

Qualitative online surveys with LARC users and service providers (health-care practitioners and weight loss practitioners) were used to meet aims C, D, E and F.

Analysis of policy documents

A total of 15 documents were included in the review: 13 were identified via the search strategy detailed in Appendix 6 and two were identified subsequently by study team members. The document types identified included FSRH clinical guidance, FSRH Clinical Effectiveness Unit-produced statements, NICE public health guidelines and clinical guidance, RCOG Green-top Guidelines, and BPAS and RCOG press releases (see Appendix 7).

Engagement with LARC users and service providers

Aim D: assess the barriers and facilitators, including practical and ethical considerations, to introducing weight and the option of delaying LARC removal from the perspective of LARC users and health-care practitioners

The surveys for both groups of stakeholders included specific questions about the introduction of weight as a topic during a LARC removal and the acceptability of asking/being asked to delay the LARC removal in order to take part in a WLI before pregnancy.

Many LARC users (46.8%) described feeling uncomfortable at the prospect of discussing weight at LARC removal appointments; 38.4% said that they would be comfortable and 14.8% said that they would be neither comfortable nor uncomfortable with the conversation. However, the majority of health-care practitioners said that they would feel comfortable introducing the subject of weight with patients attending for LARC removal (65%), whereas 12% reported that they would feel uncomfortable and 23% would feel neither comfortable nor uncomfortable.

In answer to the specific question about whether it would be acceptable to ask LARC users with overweight or obesity to delay LARC removal to attend a WLI prior to trying to conceive (Figure 9), 39.9% of LARC users stated that they felt it would be acceptable, compared with 29.6% of LARC users who felt that it was unacceptable and 30.5% of LARC users who were unsure. The majority of health-care practitioners (63.6%) reported that they felt that women would be willing to postpone LARC removal to take part in a preconception WLI, whereas only 17.2% of health-care practitioners felt that women would be unwilling to postpone LARC removal and 19.2% were unsure.

FIGURE 9.

Stakeholder responses to the acceptability of delaying LARC removal to lose weight.

The themes that emerged from responses to these questions are described below in the broad categories of barriers to and facilitators of introducing preconception weight and delaying LARC removal at LARC removal appointments. Many of these themes repeated or resonated with those that emerged from the LARC user and health-care practitioner responses to the questions surrounding the discussion of weight in health consultations generally.

Barriers to the introduction of preconception weight and LARC removal delay at LARC removal appointments

Barriers are detailed according to the following themes: (1) the sensitivity of weight as a topic area, (2) LARC user engagement and past experiences, (3) knowledge and beliefs of the LARC user, (4) the context of the consultation and service setting, (5) health-care practitioners’ skills, (6) the information/interventions provided and (7) ethical implications of requesting a delay in LARC removal.

Sensitivity of the topic

Both stakeholder groups considered weight a sensitive topic and felt that conversations regarding weight at LARC removal appointments would be difficult. LARC users described that they would feel ashamed, attacked and judged for having a raised BMI if weight was discussed at LARC removal appointments; health-care practitioners were aware of the potential to upset patients and were concerned about possible repercussions:

I gained a good amount of weight on the implant, and I don’t like talking about it. I feel judged.

LU234

Following a recent complaint when I discussed weight with a patient at a contraceptive review, it has made me wary about initiating discussions.

HCP64

However, some in both groups were ambivalent, acknowledging both the discomfort with the conversation and the belief that the conversation was needed. Some LARC users appreciated the importance of the conversation and health-care practitioners’ good intentions; similarly, health-care practitioners described the necessity of discussing weight as a health risk with patients despite the potential outcomes:

Because weight is a very personal thing and it sometimes feels like an attack even though with a health professional I know it is in my best interest.

LU44

I am uncomfortable about it because it upsets women but think I it is my medical duty to mention it so try and do it as tactfully as possible.

HCP87

LARC user engagement and previous experiences

LARC users reported being reluctant to engage in discussions about weight loss during LARC removal appointments. Barriers included being content with their weight but also the complexity of weight loss that requires more support than a simple mention during a health-care appointment, which is ineffectual:

If there was a point. What are your aims to ask about weight? Weight isn’t as cut and dry as smoking or drinking alcohol. It’s not like you can ask ‘are you currently thinking about quitting food?’

LU5

Past experiences of weight discussions in health-care consultations could be a significant barrier to introducing the topic of weight during a LARC removal appointment. LARC users reported frustration that weight was so frequently a focus in consultations, often, in their view, inappropriately. As a result of these types of experiences, LARC users said that they avoided health care if possible, and health-care practitioners acknowledged that patients are unwilling to engage because of their previous experiences. Furthermore, LARC users reported occasions when they had attempted to seek help to lose weight but had been disappointed by the lack of support from the health-care practitioner and the quality of the care/intervention:

Almost every medical professional I have ever seen has suggested symptoms of everything from a chest infection to a sprained wrist would be improved if I lost weight.

LU4

Almost everything you go to see a health-care professional about has the opportunity for them to bring up the subject of weight. I’m rarely ill, but even if I am, I put off going to see anyone for as long as humanly possible.

LU8

I mention it to patients, feel that they are not motivated to change and get ‘fed up’ with health-care professionals telling them they are overweight when they already know they are overweight.

HCP27

Mental health issues, both related and unrelated to weight, may be a barrier to introducing the topic of weight at LARC removal appointments. Anxiety, experiences of bullying and low self-esteem were all described as barriers to engaging in conversations with health-care practitioners or WLIs. LARC users reported the potential negative impact that discussing weight could have on patients who have a history of disordered eating:

Before giving advice or recommendations on weight loss methods get to know the patient and their history better so it feels more like you’re being helped rather than being told off.

LU215

I have struggled with eating disorders in the past, and I know that any attempt at losing weight quickly turns to me starving myself for quick weight loss.

LU235

Knowledge and beliefs of the LARC user

LARC users felt that weight did not map on to health as closely as was suggested by health-care practitioners and mistrusted the emphasis on BMI. Some LARC users questioned the assertions related to risk, deeming the interpretation of statistical risk exaggerated, and they wanted clearer information/statistics to be able to make an informed choice:

I know all about the risks but lots are correlation rather than causation. I have not had any complications in my pregnancy despite a high BMI . . . Don’t worry about the confounding socioeconomic factors behind those statistics I am fat so I am doomed.

LU6

LARC users believed that the LARC was a causative factor in their weight gain and wanted this to be acknowledged by health-care practitioners. One health-care practitioner reported using patients’ perception of LARC and weight gain as a way to introduce weight loss:

Sometimes a route in is in their perception of the LARC and their weight, i.e. ‘I am taking it out because I’ve gained weight’.

HCP63

Service setting/context

LARC users, and some health-care practitioners, suggested that weight is not relevant to contraception, and that discussion of weight or health matters at a contraception appointment is not appropriate unless raised by the patient:

Client is in LARC clinic not in a weight watchers’ class or slimming clinic!

HCP43

If it was just asked off the cuff with getting contraception removed I’d be a bit taken aback. I’m not there presenting with a health complaint, so why discuss weight?

LU5

Several barriers were suggested by health-care practitioners regarding the logistics of asking women to delay LARC removal when presenting at contraception appointments. Health-care practitioners and LARC users reported long waiting times for appointments and time limits placed on appointments:

I had to wait months for an appointment to remove my implant in the first place which was difficult as I was keen to try and get pregnant – I would have been upset if I’d had to consider delaying removal further.

LU18

It would be really hard to have this kind of conversation in a very short appointment in a successful and productive way.

LU44

In my surgery, patients book in without prior counselling for a 10 min (coil) or 20min (implant) removal so time is limited. We would have to change our system.

HCP93

The reason for a patient requesting LARC removal may determine whether or not weight should be introduced during the LARC removal appointment. Participants suggested that if patients requested LARC removal for reasons other than wanting to conceive, such as side effects including heavy periods, weight gain and changes in mood, then it may not be appropriate to raise the topic of weight:

I had a coil removed for mood swings, I couldn’t have spent any longer on it. God help if a stranger had told me to leave it in and lose weight!

LU75

The majority of women for whom I remove LARC state they want removal for side effects (perceived or actual) and it is not often that I see someone that wants removal purely for pregnancy.

HCP42

Finally, LARC users reported feeling vulnerable during intimate LARC removal appointments and felt that it would be inappropriate for practitioners to instigate discussions regarding weight:

For some females it can be incredibly nerve wrecking going to have a discussion with a total stranger about birth control. Especially if it’s the coil where they are at the most intimate part of a female’s body then they start talking about the person’s weight!

LU75

Health-care practitioners

Many LARC users thought that health-care practitioners lacked the required skills to deal with the sensitive topic of weight management. Similarly, health-care practitioners acknowledged that they could feel uncomfortable discussing weight with patients, particularly when providing preconception advice. Health-care practitioners wanted to be provided with guidance or accurate information regarding the risks of raised BMI in pregnancy or effective weight loss advice to enable them to have successful discussions with patients:

Making us feel crap about our choices and our bodies, is not a way to motivate anyone. Look up motivational counselling techniques. Look up life coaching techniques. How the healthcare service treats overweight and obese people is literally the opposite, with the effect that we feel crap and we get fatter.

LU8

If this HCP was someone I was only seeing as a one off. There was no follow up with the same person or I’d never met the person. No rapport or trust. I would be fairly insulted if they offered unsolicited advice.

LU5

Staff will feel uncomfortable unless they have some guidance on how and when to phrase things.

HCP87

I’d be willing to try it – but would like to have some training about how to raise weight loss anyway . . . would there be any supporting patient resources – then the patient might feel the suggestion is not coming personally from me.

HCP67

The weight status of the health-care practitioner also appeared to be a barrier to successful discussions regarding weight at LARC removal appointments. Health-care practitioners with overweight may appear hypocritical, whereas health-care practitioners with a healthy weight may appear unsympathetic and patronising:

It would be nice if the healthcare professional followed their own advice. Being out of puff and smelling of bacon is an example that rather frustrates the patient.

LU161

Some clinicians may not be keen particularly if obese themselves – practice what you preach!

HCP64

Health-care practitioners spoke about the requirement for referral pathways to effective interventions, and both stakeholder groups reported their frustration at the lack of efficacious interventions or at unsuitable interventions:

. . . unhelpful advice relating to solutions that are financially inaccessible and not accessible to single parents with limited support.

LU137

Ethical implications of requesting LARC removal delay

In many of the responses to the idea of delaying LARC removal to lose weight, respondents reported what can broadly be described as ethical barriers relating to personal choice, conception decision-making, impact on the care pathway and exclusion of non-LARC users.

Individual freedom to choose

LARC users and health-care practitioners highlighted the ethical implications of what could be interpreted as an imposition of a contraceptive decision on patients, suggesting that this should, ultimately, be the choice of the individual. Some LARC users referred to discrimination against people with overweight/obesity and said that the inference would be that ‘fat people shouldn’t breed’ (LU4). LARC users felt that state involvement in personal reproductive decisions was unacceptable. They were also concerned that people may feel pressured into agreeing to delay LARC, acknowledging that health-care practitioners, despite acting with best intentions, may be unaware of this consequence. However, health-care practitioners also acknowledged the potential to pressure patients in a ‘doctor-led approach’:

I understand that you are concerned about fat people having babies and the implications for the healthcare system but if a woman wants her coil taken out at any time for any reason that is her choice.

LU117

I would find it ethically complex to suggest to women that their freedom of choice to have their contraception removed may be compromised by their weight. I would however feel happy to discuss the options with them and share evidence about the pregnancy-outcomes in obese women.

HCP57

Because you’re denying someone the right to get pregnant due to their weight. That is a breach of their human rights.

LU180

People who are overweight are already treated like second class citizens by some professionals, this would feel like a form of eugenics unless treated very carefully.

LU23

. . . this is already happening in places, with women are reporting having been ‘refused’ to have their implant/coil removed, conditional on losing weight. That may not be what the healthcare professional thought they were saying, but that is the message women are hearing.

LU8

Complexity of timing in conception decisions

The complexity of decision-making surrounding trying to conceive may be a barrier to asking patients to delay LARC removal to lose weight; LARC users said that the decision to have a baby is a complex one, which may be informed by their relationship, their studies, their career choices, sibling age gaps, financial implications, etc.:

People do not lightly decide the time is right to try for a baby!?!? They will have thought long and hard about it, and undoubtedly already have concerns about weight.

LU8

Both LARC users and health-care practitioners commented on the risk of delay in older women, saying that a long delay may have implications for their chances of conceiving or lead to complications in pregnancy due to age. Additionally, women who have fertility issues may not want to delay LARC removal:

. . . may not achieve anything other than delaying pregnancy and LARC removal – so you end up with older overweight people getting pregnant.

HCP42

Impact on care pathway

It was suggested that, if patients were aware that they would be asked to delay LARC removal due to their weight, this may discourage patients from having a LARC fitted in the first place, or they may go elsewhere for removal. LARC users also questioned whether a process would be in place if patients declined to attend a WLI or if they were unsuccessful at losing weight and whether this would have implications on the care they received or if they were ‘allowed’ to have a baby. Furthermore, there was an expectation that, if the risks are greater in women with overweight/obesity, then supplementary screening/care should be offered to pregnant women:

It also carries the connotation that if they do undergo a weight loss programme but don’t succeed – what happens then? . . . I feel this kind of protocol carries a lot of weight and as such needs to be undertaken very carefully and sensitively.

LU63

Being asked if I would like to discuss potential risks/additional screening in the event of falling pregnant.

LU2

Exclusive intervention

Finally, it was highlighted that people may feel excluded from an intervention targeting women trying to conceive, suggesting that an effective WLI should be offered at a population level.

Facilitators of the introduction of preconception weight and LARC removal delay at LARC removal appointments

The facilitators are a mirror image of the barriers across the following themes: LARC user engagement and previous experiences, knowledge and beliefs of the LARC user, service setting/context, health-care professional and information provided/available interventions.

LARC user engagement and previous experiences

LARC users stated that they are accustomed to discussing weight at health-care appointments and appreciated the importance of the conversation and, therefore, despite the innate sensitivity of the topic, they would be co-operative and unlikely to be offended. Some LARC users described the prospect of delaying LARC removal as a reasonable proposal, and potentially would welcome the discussion as a motivating factor and any offers of support. Health-care practitioners also referred to patients’ attitude, wishes and motivation as facilitators of successful discussions regarding weight; if the patient has a positive attitude to weight loss, initiates the conversation or has a desire to improve their health, it allows for beneficial discussions. LARC users and health-care practitioners suggested that whether or not women find the conversation acceptable would depend on their circumstances, age, history and weight status; participants suggested that asking people to delay removal would be acceptable only if the individual had obesity and was not just ‘slightly overweight’:

I would be happy being asked to take part in a weight loss programme before trying to become pregnant. I would see it as a way to become more healthy before conceiving, which could also increase the chance of becoming pregnant if I lost weight.

LU9

Knowledge and beliefs of the LARC user

Nearly two-thirds of LARC users (64.2%) said that they were aware of risks of overweight/obesity in pregnancy, and each described between 1 and 10 risks, including conception risks (difficultly conceiving), pregnancy risks (miscarriage, pre-eclampsia, failed intubations, pelvic symphysis disorder, severe pregnancy symptoms, etc.), delivery risks (shoulder dystocia, stillbirth, cesaerean section, post-partum haemorrhage, instrumental delivery) and risks to the baby (cerebral palsy, spina bifida, large or small birthweight, premature birth). A total of 15.2% of LARC users stated that they did not know any risks, 2.9% were unsure of the risks and 17.7% did not respond. LARC users thought that knowledge about or being informed of the health benefits of being a healthy weight could be a motivator to engage in a WLI. They stated that communication of the risks of obesity in pregnancy was essential to allow patients to make an informed choice and maintain autonomy. However, as discussed previously, some LARC users were sceptical about the risks of overweight/obesity in pregnancy, saying that these are exaggerated by the health-care profession:

Personally, with the current weight I am, I would be less keen to become pregnant now. I would prefer to lose some weight and get fitter – but that’s because I know how pregnancy has an impact on my body and I would want to be as prepared as possible before hand to help support my back and hips (previous pgp).

LU5

It’s in their interests, they are free to decline . . . they may be unaware of the risks. Few service user participants these days mind not smoking and not drinking alcohol in pregnancy, weight loss may simply not be on their radar.

HCP20

Service setting/context

LARC users suggested that preconception and pregnancy may provide a window of opportunity to introduce weight loss in order to maximise fertility and improve health in pregnancy. In addition, the potential benefits to the unborn child may be a facilitator of healthy behaviours. Others described this opportunity in pragmatic terms, that is a time of increased contact with health-care practitioners and a potential willingness to discuss weight when it is considered as part of a broader preconception/healthy pregnancy package of health recommendations:

Because the healthier you are the better chance you have of getting pregnant and having a problem free pregnancy. Wouldn’t any mother want that?

LU150

Health-care professional

LARC users believed that it was the medical professional’s role to discuss health issues such as weight with patients and expressed confidence that health-care practitioners prioritise patients’ health. They valued an ongoing relationship with health-care practitioners and continuity of care, including the practitioner’s knowledge of the patient’s existing medical conditions, as vital for successful discussions.

LARC users highlighted the importance of good communication skills; they indicated that the health-care practitioner should be sensitive, compassionate, non-judgemental and not patronising and be aware of their tone and body language to ensure that the patient is at ease during the consultation. Having adequate time was seen as a facilitator of ensuring effective communication, as was not treating the conversation like a ‘tick-box’ exercise. An open, honest discussion was important to LARC users, as was practitioners having an awareness of terminology likely to cause offence, such as ‘obese’ or ‘fat’, and an understanding of the difficulties of losing weight. LARC users suggested ways of broaching the topic, rather than giving abrupt unsolicited advice. Health-care practitioners also considered themselves potential facilitators if they approached the discussion sensitively and had a good rapport with the patient, but they also requested additional training, guidance and potentially a script to ensure that they took the right approach:

If this HCP was someone I was only seeing as a one off. There was no follow up with the same person or I’d never met the person. No rapport or trust. I would be fairly insulted if they offered unsolicited advice.

LU5

Information provided/available interventions

LARC users said that, instead of simply instructing the patient to lose weight, health-care practitioners should provide support for and/or practical advice about losing weight or signpost to a choice of options for effective support. Practical solutions were recommended, such as diet tips, meal planning, free gym membership, WLIs and support networks. Similarly, health-care practitioners said that they could justify initiating such conversations if they were able to offer an effective intervention. Support that had a positive focus was important to LARC users, with the discussion being in the context of supporting health goals and highlighting the benefits of weight loss, rather than just focusing on the risks. Both LARC users and health-care practitioners suggested that, to increase acceptability and reduce stigma or risking patients feeling judged or singled out, weight and healthy lifestyle should be a standard part of preconception and contraceptive discussions:

If we could provide support/refer service user participants for intervention then I think service user participants would see that and find it acceptable/understand why they have been asked.

HCP44

It will be more acceptable if offered to everyone in this scenario.

HCP8

Both LARC users and health-care practitioners thought that referring to evidence-based guidance was a facilitator of acceptable discussions and enrolment in a potential intervention:

. . . information-giving discussion including introducing some of the evidence behind the issues to help it feel objective.

LU33

Timing in the context of the decision process

The decision to have a baby is often complex, thought through (also with partner), and timing is part of an emotional, social, cultural and financial web of circumstances. Although many have LARC removal for reasons other than wanting to become pregnant straight away, for example because of side effects, they would still not want to delay removal, and for those where it is a deliberate step to pregnancy, the removal appointment is considered too late.

Woman’s right to choose

The proposed intervention interweaves two complex sociopolitical issues in weight and conception, which can be captured by the concept of a ‘woman’s right to choose’, a phrase that has become synonymous with women’s reproductive rights. While there was never any intention that the intervention would remove a woman’s choices, the underlying message that some stakeholders identified was discriminatory, suggesting that women with a raised BMI should not become pregnant, and there was perceived to be a risk that women would feel pressured to participate, however unintentionally, by the health-care practitioners’ description of the option to delay the LARC removal.

Is weight part of health/family planning?

Throughout the themes is a thread of whether weight is regarded as a health/family planning issue.

In summary, several factors were important to consider when designing a potential preconception WLI and were to be explored further through the SAGs and phase 2:

• In the discussion of the intervention, it should be clear that the woman retains the right to choose to delay or not delay the removal of the LARC. This could be achieved by timing the discussion better so that it takes place well in advance of a LARC removal appointment, or ensuring that the delay of the removal is not one of the eligibility criteria for taking part in the intervention.

• The intervention could be offered to the general population, so that the LARC removal appointment is simply an opportunity to engage.

• Information that is clear for all about risk should be included in referral and intervention information.

• Health-care practitioners should be provided with training in communication skills and information regarding risks.

• Women’s experiences with weight loss should be taken into account, as these encounters set the stage for their likelihood of engaging with any future intervention.

• The referral and intervention should have a positive approach, focusing on the benefits of weight loss, rather than the risk of current weight.

• The intervention should be free and available to all.

• The intervention should be part of a healthy pregnancy approach.

Overall research questions for the review

• Preconception WLIs: what works, for whom and in what contexts?

• What are the barriers to and facilitators of engagement in preconception health behaviour change interventions?

• Is anything additional to be learned from WLIs during pregnancy and postnatally that could contribute to the development of a preconception WLI?

Aims of the realist review

To answer the research questions and consider any potential intervention designs, the aims were to develop an understanding of and potential theories about:

1. intervention characteristics and approaches to recruitment

2. contextual influences on the intervention

3. potential CMO configurations.

Search strategies methods

In realist methodology, concept mining describes the search process across a wide range of published evidence to explore potentially relevant mechanisms and how those might be affected by context in order to help build theories. This could include reviews, intervention studies, protocols and qualitative research.

Several systematic reviews and meta-analyses related to obesity and the perinatal period have been published in recent years (see Appendix 8). The conclusions drawn in those reviews and the studies they cite were examined to identify (1) possible intervention designs and (2) relevant information on contexts, mechanisms and outcomes. Supplementary searches, based on the review research questions, were performed to ensure that more current publications were included, and to bring together additional relevant evidence to support the theory development (see Appendices 4–6).

For each search, a search strategy was developed and run, and relevant papers were identified through purposive and snowball sampling (see Appendix 6 for details of the search strategies and the papers identified).

The use of middle-range theories

As part of the development of programme theories to guide the intervention development, existing middle-range theories that could relate to preconception weight loss and identified barriers and facilitators were explored, as were theories that may be considered central to weight loss in pregnancy and may generalise to the preconception period. Middle-range theories are those that, on the continuum of theory development, would fall between working hypotheses (as can be generated within individual projects, often described as programme theories) and the grand theories that aim to present a unified theory of the social world. 106 They operate at a more abstract level than the programme theories, specify mechanisms by which change may occur and, therefore, enable the development of more generalisable hypotheses.

Theories were identified in three ways. First, a list of relevant theories familiar to the research team from their work in studies relating to weight loss, GWG and behaviour change was generated. Second, theories that were stated as underpinning the interventions were identified through the searches. Third, a Google Scholar (Google, Mountain View, CA, USA) search specifically focused on theory107 was conducted to identify any texts that had not been identified in relation to preconception interventions (see Appendix 6 for the search strategy).

The consolidated explanatory accounts that were developed from the research literature were then considered alongside the formal theories identified through the searches. The next step was to integrate all of the data sources to produce a theory to take forward to the next stage of development. This process is outlined in the sections below.

Identified middle-range theories

The Google Scholar search identified two papers that described theoretical models relevant to preconception and obesity. The B’More Fit programme108 described the work of a coalition tackling obesity in a preconception population and used the social-ecological model, emphasising the multilayered nature of influence and the relationship between behaviours and the social environment. The second source included from this search was Liang et al. ,109 a review of theoretical components of guidelines for physicians. Although not specifically related to obesity, this source was thought to be relevant given the importance of obesity guidelines. In this review the theory of planned behaviour (TPB) and theoretical domains framework were most frequently used (in 16/42 and 10/42), with NPT also used (2/42).

The theories identified as of potential relevance and the source(s) used to identify these are listed in Table 17.

The lack of theory underpinning preconception interventions was identified in a review of preconception health behaviour research. 38 Many of the papers in this review either refer to interventions in broad terms, such as modifying behaviour, lifestyle modification and increasing motivation, but without references or pinpointing specific underpinning theory, or refer to constructs such as knowledge and attitudes, but often lack a definition of the construct. As a consequence, to include them would require interpretation and assumptions about potential theoretical mechanisms of action [e.g. impact of knowledge could be classified as Capability, Opportunity and Motivation to Behaviour (COM-B) self-efficacy], introducing bias and inaccuracies at this stage.

Each of the theories identified could potentially have a bearing on the review and were considered by two qualitative researchers. The four theories selected as of particular relevance and also encompassing many elements of the remaining theories were the TPB, COM-B, NPT and the social ecological model. Although areas of overlap exist between these four theories, each provides a unique structure that enables clear consideration of three central aspects of PCC programmes, namely individual behaviour change, practitioner behaviour change/embedding changes in professional practice, and relationships between individual behaviours and the social/societal contexts in which they occur.

These theories are briefly described, and their role in relation to the hypothesised programme theories is then considered in more depth.

CMO configurations

The review process, taking information from the searches through into explanatory accounts and consolidated accounts, resulted in seven CMO configurations that together form the basis of a hypothesised programme theory by which a preconception WLI could work in the general population. Each configuration is described, citing the evidence from the literature that underpins it, followed by the fit with the middle-range theories and the stakeholder survey results. Configurations 1–6 integrate findings from these three types of sources; configuration 7 did not emerge from the literature but has been developed from the stakeholder feedback and is considered in the light of the theories.

CMO configurations:

1. Reaching out to people.

2. Recognising the diversity and wealth of the individual’s experience.

3. Build the health-care practitioners’ confidence and commitment to weight management in their practice of PCC.

4. This is something for me.

5. An intervention that is fit for purpose.

6. Building confidence and motivation.

7. Weight loss discussions should be founded on the principles of informed choice and a client-centred approach.

Aim of the chapter

Work package 2 identified potentially suitable preconception/pregnancy-related WLIs and the theories underpinning them using realist methods, in addition to assessing the feasibility and acceptability of a preconception WLI to stakeholders (LARC users and health-care practitioners). The aim of this chapter is to describe the work with the SAGs, refining the intervention components and associated theory, and to consider the key questions to ask participants in phase 2.

Methods

The findings from the phase 1 work were synthesised to describe the core and optional components of an intervention. These findings from phase 1 were presented at two SAG meetings: a LARC user SAG and a health-care practitioner SAG. Questions to be addressed at each SAG were agreed by the study team (see Report Supplementary Material 5 and 6). The health-care practitioner SAG was held at a British Association of Sexual Health and HIV (BASHH) audit event, and the LARC user SAG was held remotely via Zoom (due to COVID-19 restrictions).

Results of the Stakeholder Advisory Groups

Three LARC users attended the LARC user SAG meeting and 34 health-care practitioners attended the health-care practitioner SAG. The feedback from the two SAGs is summarised in four domains: (1) general response to the model of the intervention, (2) specific components of the programme, (3) barriers in practice and (4) questions to include in phase 2 interviews.

Aim of the chapter

The aim of the chapter was to further refine outputs from the phase 1 survey and the SAGs. The potential intervention design and associated CMO configurations informed by phase 1 findings were tested with targeted qualitative interviews with LARC users and health-care practitioners.

The aims of the qualitative interviews were to assess:

• LARC users’ views regarding the acceptability and feasibility of the potential WLI

• health-care practitioners’ views regarding the type of intervention and their willingness to recruit women to such an intervention.

The findings from the interviews were then presented to the two SAGs for discussion to refine a final version of a proposed intervention.

Participant identification/selection

Participants in all qualitative interviews and SAGs were identified from the pool of participants who had previously consented to be contacted in the WP2 phase 1 online surveys.

Potential participants in the LARC user and health-care practitioner qualitative interviews and the LARC user SAG were contacted by a member of the study team using the person’s preferred method (e-mail/telephone) with information about the phase 2 interviews/SAG. Participants were purposively sampled to ensure a range of BMI and reproductive histories/plans (LARC users) and professional backgrounds and genders (health-care practitioner interviews). All potential participants in the health-care practitioner SAG were invited by e-mail to take part in the SAG.

Participant informed consent

Participants in the qualitative interviews and the LARC user SAG were e-mailed the PIS (see Report Supplementary Material 10–12) and consent statements (see Report Supplementary Material 13 and 14) and were given sufficient time to consider the information prior to the interview/SAG. At each interview/SAG, participants were given the opportunity to ask any questions, and consent statements were read aloud by a member of the research team. All participants were asked to consent verbally, and this was recorded.

All potential participants in the health-care practitioner SAG were e-mailed an information sheet (see Report Supplementary Material 15) and link to the Padlet discussion forum. Participants were informed that participation in the discussion was considered consent to participate.

Withdrawal

Participants were able to withdraw at any time by leaving the interview/SAG, but their contributions up to that point were unable to be withdrawn.

Procedure

Topic guides for interviews were based on the synthesis of phase 1 results, described in Table 12 (see Chapter 6; and see Report Supplementary Material 16 and 17). For the LARC user SAG, the study team presented study progress, which was followed by a discussion based on the synthesis of phase 1 results (see Report Supplementary Material 18). For the health-care practitioner SAG, a study update was provided using a combination of written text, a PowerPoint summary and a video of the chief investigator presenting the main findings. Participants were asked to respond to three questions, each focusing on a service-related topic: (1) how to contact women who use their service who might be interested in a preconception WLI, (2) how to advertise it more broadly in the community and (3) what training practitioners would want in talking about weight.

Questions to be addressed in interviews and with SAGs were agreed by the study team and used as a foundation for discussions, but prompts were used based on participant responses.

Analysis

Both SAGs were recorded, anonymously transcribed and analysed thematically. The results were further grouped and are reported according to the CMO with which they most closely align. Any repetition of the findings in phase 1 (see Chapter 6) has not been reported in detail.

CMO 1: reaching out to people – maximising engagement

Qualitative feedback relating to CMO 1 included interviewees’ general impression of the proposed intervention and their views on the recruitment route and the approach to discussing risk.

CMO 2: recognising the diversity and wealth of individuals’ experience

All of the LARC users interviewed had experiences of weight loss attempts and had tried various approaches or programmes, which would inform their approach to any new weight loss programme. The aspects of previous weight loss plans that they found useful included (1) weekly weigh-ins (increased motivation due to accountability); (2) group support elements, including apps (although not suitable for all); (3) different apps to track progress; and (4) flexibility in order to fit in with people’s lives:

I just think physically going somewhere and knowing in a week, that actually you’re going down and step on those scales on Tuesday in front of Joe Bloggs keeps you motivated.

LUI01

So if, if you’re having a bad day and you, you know, you’d have a day when you’d eaten far more than you should have, they were very supportive and, you know, with all being there and, you know, just get back on it tomorrow.

LUI03

CMO 3: building health-care practitioners’ confidence and commitment to weight management in their practice of preconception care

Practitioners were asked about how they and their colleagues could be best supported to have discussions about preconception weight with patients in order to introduce the intervention. They felt that the communication skills required should already be integral to their role, but some suggested that training in motivational interviewing158 may be beneficial. Additional information via fact sheets or existing guidance documents and training would help overcome their lack of knowledge about some of the risks. Some thought that a script could potentially be useful for introducing the topic, and training videos and role playing using the script were suggested as ways to help practitioners become confident. Furthermore, a thorough understanding of the intervention itself would be essential for engaging patients:

I mean certainly as a GP my understanding of risk, I know it is desperately poor and my ability to articulate that and explain that in most circumstances could be better.

HCPI01

A lot of staff don’t feel confident about asking any number of things about alcohol, substance abuse . . . but what we found is that if they can be given some key strategies or phrases, they’re often happy to run with it.

HCPI10

Where it’s an awkward conversation [it] needs to be scripted. You need to have practised it.

HCPI03

I think they need to have a good idea of what the intervention was, in terms of how long it lasted, and the kind of steps that people would go through . . . when they could withdraw from it if they wanted to or not. If they became pregnant early on, what they would then do after that. And the safety of those things.

HCPI06

CMO 4: this is something for me

Three key topics in this CMO related to eligibility for and engagement with the programme where we needed feedback from stakeholders to progress the design: (1) the lower threshold of BMI for the programme, (2) the acceptability and feasibility of weighing women in clinic and (3) the preferred mode of delivery of the intervention.

CMO 5: an intervention that is fit for purpose

The questions for stakeholders that arose from this CMO concerned the acceptability of explicitly describing the 5–10% weight loss and the inclusion of meal replacements as an option. We also explored the practicalities of weight loss feedback, which resources interviewees had found useful in their weight loss experience and any obstacles they could see that had not been brought to the fore by other interview topics.

CMO 6: building confidence and motivation

Participants were asked about their experiences of successful consultations about weight to try to elicit the key ingredients for establishing strategies to build confidence and motivation.

Some LARC users said that they had never had a good conversation about weight with health-care practitioners. However, those who had done identified various aspects that had made a difference, including the attitude of the practitioner, for instance if they were non-judgemental and matter-of-fact. Some participants talked about having a good relationship with their health-care practitioner, and participants said that the experience is improved when health-care practitioners take the time to listen to the patient. One health-care practitioner said that this might be easier for nurses, who are allocated more time for consultations. Health-care practitioners felt that factual starts to the conversation might also help:

I appreciated what she said actually, she didn’t really pussy foot around . . . she didn’t shame me . . . she just said, ‘keep walking, what you’re doing, but just control maybe your portions and don’t eat as much’ [laughs]. And she was right, she was right [laughs], what can I say?

LUI05

Both groups of stakeholders felt that the conversation needed to be led by the patient, either that they introduced the topic or that it was given as an option and left to the patient to decide whether they wanted to explore it. The practitioners said that it was important to ensure that they took the position that the patient was an expert in their own condition, tailoring the discussion and acknowledging their knowledge:

At the point where you go in and say, ‘I would like some help with my weight’, that’s the point where it should be a positive experience. They should be like, ‘That’s really great you’ve come in. These are your options’.

LUI11

CMO 7: weight loss discussions should be founded on the principles of informed choice and a client-centred approach

The theme of ensuring that women were given information and able to make their own choices came through several of the areas of discussion in the interview. As with the survey responses, when discussing the design, LARC users wanted it to be really clear that delay was just one option, that it was not a prerequisite for attendance and that the idea of informed choice was front and centre of the discussion. This includes presenting the risks and advantages of different decisions so that the woman can fully understand the choices being presented and is fully aware of her options. Health-care practitioners linked the discussion of risk and options with their professional responsibility and the discussions they have in relation to other procedures. The aim is to make sure that the woman can make an informed choice, fully aware of the risks and benefits:

So yes, I think, but I do think that informed choice is important and you can’t make an informed decision if you haven’t been presented with all the facts around weight and pregnancy and if you haven’t been given an option to delay with support to lose weight.

LUI04

Phase 2 Stakeholder Advisory Groups

The LARC user SAG was attended by three members of the research team and five women, two with children and pregnant/planning to be pregnant, two planning to be pregnant in the next year and one planning to be pregnant within the next 5 years. The results are presented under the three main themes of the discussion: (1) engagement, (2) outcomes and (3) support and improving practitioner skills.

Three main points were made by the SAG with regard to maximising engagement with the programme. The advertisement of the programme would need to be wide reaching, involving physical spaces such as community pharmacies (not just online), and also be clearly identified as NHS-endorsed to be seen as trustworthy. The emerging evidence of the potential effectiveness of the intervention would need to be widely disseminated, and not just shared with clinicians, to enable women to make decisions about the suitability of the programme for them. The health coach was seen as key to retention as they would offer continuity, being the person with whom participants felt safe and connected.

In terms of engagement with the WLI, the stakeholders stressed that it would be important to consider what outcomes (not just weight loss) would be valued by service users, which could therefore be included as outcomes of any research but also reflected in the advertising. The SAG members identified the following potential benefits of importance: (1) increasing energy, (2) confidence, (3) feeling proud of yourself, (4) feeling healthier, (5) able to exercise, (6) posture and (7) less weight on joints. They also said that the critical shift, the thing that makes the difference between this and other times in a person’s life, is that preparing for family life and doing the best for your baby’s health were seen as powerful psychological motivators, far more so than thinking of one’s own health and well-being alone.

The peer support element of the proposed intervention divided opinion. Some felt that it was really important to be ‘trying to achieve something together’, whereas others felt that this was something that needed to be approached with caution, not just because they found group work unhelpful, but also because non-experts may start offering advice regarding pregnancy. In terms of training practitioners, the group reflected that it is very difficult to train people in fundamental communication skills if they do not have them already; therefore, it is important to advertise the programme so that women know to ask for it, and also to design a script for health-care practitioners to ensure that, as a minimum, the opening of the conversation is done well.

The time frame of the programme was discussed, reflecting on the recent findings from the Prepare trial in the USA,53 and it was suggested that a preconception WLI may need to support women until the end of pregnancy to counteract a rebound of preconception weight loss with greater GWG.

In summary, the LARC user SAG identified important considerations in engagement and recruitment, the centrality of the continuity of care provided by the health coach and the value of ensuring that advertising was widely disseminated not just to let people know about the programme but also to provide knowledge so that service users would be empowered to initiate the conversation with their health-care provider and suggest attending the WLI. The individuality of women’s needs around weight management means that, if possible, tailoring elements of the programme, such as the nature of provision, intensity of support and goals, would be beneficial. The programme would need to offer ‘light-touch’ support after the initial preconception programme.

Nine responses were received across the three questions in the health-care practitioner SAG. The ideas suggested for contacting women in their service included wall posters, waiting room QR codes, a link on the pre-procedure questionnaire sent for all LARC attendances, taking permission for health promotion text messages at registration, a link on the clinic website on the page for pre-pregnancy planning and also on the sexual health website. The general advertising followed similar approaches, with QR codes in toilets, gyms, and so on, but it also involved a social media campaign, including developing community champions. The themes in the responses to the training question echoed those from the interviews, of really working on improving the motivational interviewing training for practitioners and also the central importance of having an effective intervention to refer to. The biggest barrier to opening up the conversation about weight is if nowhere is offering support.

Potential intervention

The outline of a potential WLI for women in the preconception phase and the principles underlying it based on evidence and stakeholder feedback are shown in Figure 11. The WLI is framed as a healthy pregnancy programme for women with BMI of ≥ 30 kg/m², advertised widely in the community for self-referral but also discussed by practitioners involved in family planning services when working with women of reproductive age who are considering conception in the next 12 months, including those presenting for LARC removal. For women still using contraception, the discussion with the practitioner will include the potential benefits of continuing with contraception for the next 4 months to maximise their chances of losing a clinically significant amount of weight prior to conception. Practitioners would receive clear evidence-based guidance on the benefits of reduced BMI that they could share with their patients, and would also receive training in how to introduce the topic using a loose script co-produced with service users and the research team.

FIGURE 11.

Outline of a potential preconception WLI.

The intervention itself is a signposting intervention, delivered by a health coach who has been trained in motivational interviewing and has an understanding of the evidence-based principles of behaviour change, incorporating planning and goal-setting. With knowledge of the resources available locally and online, the health coach will support the woman to identify her own goals and understand the key components of her experience of successful and unsuccessful weight loss attempts. The information provided will focus on preparation for a healthy pregnancy and will be provided initially via an online platform and then in discussion with the health coach. Signposting to a range of interventions inherently maximises a sense of autonomy, and this will be partnered with support from the health-coach, in person and via text, introducing accountability with weight monitoring. Additional support via a group peer-support network will be offered if requested by the service user. The exact nature of the full programme will be tailored to the individual but also to the resources available locally (e.g. NHS-provided support).

Objective 1: to identify the annual number of women of reproductive age in the UK who request LARC removal and subsequently have a pregnancy

The pathway from LARC removal to pregnancy is not easily captured. The stakeholder narratives and the CPRD routine data show that the reasons for removal are many and varied, of which planning a pregnancy is just one. The current structure of routine data collected through CPRD and SHCs does not enable the sequential relationship between removal of LARC and subsequent pregnancy to be reliably determined. The main barriers include the quality of the routine data, in particular the multiple Read codes relating to LARC use in CPRD and the very limited coding in the Sexual Health data; and the lack of connectivity between SHC and CPRD data, which, given women’s freedom to choose the service they wish to access to remove the LARC (which may well be different from the setting where it was inserted), means that it is not possible to generate a reliable timeline of LARC-related events. Having a ‘clear’ pathway in the data between removal and a pregnancy within 15 months (our assumed definition of ‘subsequent’) may well relate to a minority of the women for whom that is indeed their lived experience.

In conclusion, the relationship between LARC removal and subsequent pregnancy is not clear enough to be able to reliably answer this objective with a number. It would not be feasible, using the current routine data sets relating to LARC use and pregnancy, to reliably identify women who request a LARC removal with the intention of becoming pregnant.

Objective 2: means of identifying women at study sites who are overweight/obese and plan to have LARC removal for the purpose of planning a pregnancy and opportunities to intervene

The barriers identified in objective 1 would also hold true in relation to objective 2. In terms of BMI information, there was an overall average of 62% of women having a BMI recorded within 3 years of a LARC event. Of those, 51–54% had a BMI of ≥ 25 kg/m², which is in line with the National Maternity and Perinatal Audit 2019,1 in which 50.4% of women giving birth had a BMI in the overweight/obese category, suggesting that there is no particular skew towards higher rates of recording of higher BMI. This rate of recording could be considered adequate for identifying women at study sites with overweight/obesity. However, the routine data barriers would impede this route in the current circumstances.

The quality of the routine data, in particular the identification of the exact purpose of the consultation, would also preclude using them to identify opportunities to intervene and introduce the topic of weight. Even if the data were more reliable, it is questionable whether using LARC-related events would be acceptable as an opportunity to intervene. When the idea of opportunities to talk about weight was explored with stakeholders, opinions ranged from feeling that weight should only ever be raised by the woman to feeling that any consultation was potentially an opportunity. Insertion was the most commonly identified LARC-related opportunity in which it felt appropriate for weight to be discussed. However, many others felt that any LARC event was a situation in which women were physically and emotionally vulnerable, and, therefore, introducing the topic of weight was not appropriate.

To conclude, from the routine data it is not possible to reliably identify women with overweight/obesity and who plan to have LARC removed for the purpose of planning a pregnancy. The acceptability of the intervention to stakeholders (described under objectives 3–5 in the following sections) would also preclude this approach to identifying opportunities to intervene.

Objective 3: suitable and acceptable interventions that can be incorporated into a pre-pregnancy weight loss intervention

The published research on preconception weight loss is dominated by studies relating to IVF that focus most commonly on hypocaloric deficit with pregnancy as the primary outcome. IVF as the context of the studies puts the emphasis on the relationship between BMI and conception, in relation to either the BMI threshold for eligibility for IVF or exploring the relationship between BMI and fertility. There is rarely any detail of support on offer or theoretical underpinning of any intervention other than the biological mechanisms relating to calorie deficit or fertility. These studies provide useful information on the level of weight loss possible with this very motivated group. However, the issue of fertility is the dominant narrative and so the studies may not offer direct guidance on the suitability of particular interventions for a broader population.

The literature on interventions in managing GWG is significantly more developed in terms of details, models, theoretical underpinning and so on. The key crossover with the preconception context is in terms of motivation for change, with our stakeholders citing the health of the baby and their own health in caring for their child as prime motivators for behaviour change at this particular point in their lives. It is useful to incorporate the evidence of the effective mechanisms underpinning intervention design in pregnancy-focused studies into a preconception WLI, with planning and feedback/monitoring being key to success. 153 Some of the barriers to WLI in pregnancy, in particular the reduction in effectiveness of physical activity interventions,117,159 would need to be considered, that is to say whether or not a preconception WLI should follow protocols closer to those of general adult or pregnancy WLIs.

The review of the literature and relevant middle-range theories using realist principles, combined with the stakeholder experiences, generated seven CMO configurations that together create a potential programme theory. A model of a signposting intervention is proposed, with focused preconception content delivered via online modules and individual sessions with a health coach. The use of a signposting model seeks to address the question of whether the intervention should be a specialist preconception WLI or essentially add bespoke content to a more generic adult WLI. The stakeholders were really supportive of the idea of not reinventing the weight loss programme wheel; there are many generic interventions available that women have used, are using or would be happy to explore. However, preconception is a special context, so the intervention requires tailoring to the preconception/pregnancy context with elements of individualised support via a health coach. This would include nutritional and physical activity information about preparation for pregnancy, motivation for change, understanding the goals in terms of clinical impact of the weight loss on perinatal health, time frames and so on. These elements came through several of the CMO configurations in the review, identifying the importance of sensitive messaging during the recruitment process so that women know that the intervention is for them (CMOs 1 and 4), that their experience (positive and negative) of weight management will be recognised (CMO 2) and that it is fit for purpose in a preconception context (CMO 5).

Objective 4: willingness of clinicians to raise weight loss in consultations with eligible women and recruit them to the intervention

The majority of clinicians reported that they were willing to raise the topic of weight in consultations, seeing discussing this as part of their role either in relation to contraceptive choices (in SHCs) or as part of general health (in primary care roles). However, practitioners reported many barriers to these discussions (in line with the published literature160,161), which ranged from the practical, in terms of time, to the sensitivity of the topic, their skills and also broader ethical concerns. These concerns included whether it was their role to discuss weight if this was not directly related to contraceptive options (e.g. if a woman’s BMI prohibited certain choices) and whether weight may be such a complex issue that a woman needed to raise the topic herself.

Over 60% of practitioners would have considered recruiting to a WLI that focused on delaying LARC removal. This would have been facilitated by information on the programme, skills training in the recruitment approach and evidence of potential effectiveness of the intervention. However, practitioners identified significant barriers that resonated with LARC users’ views, which included those to discussing weight in general (described under objective 3) but also the critical issue that the discussion was inappropriate during the LARC removal appointment. They had concerns about the impact on the woman, their professional relationship, the feasibility of recruiting in their service context with their skills/role and, generally, that weight discussions ought to be patient-initiated.

Objective 5: views of eligible women as to the acceptability and feasibility of the intervention

The views of the women who completed the survey, all of whom had experience of having a raised BMI and using LARC, were divided fairly evenly on the question of the acceptability of this WLI in its basic form, of delayed LARC removal. The argument could be made that, with nearly 40% of the respondents thinking that this would be an acceptable model of intervention, it might be worth pursuing as an option, as it may suit some people. The themes that emerged from the free-text comments, the phase 1 SAG and the interviews help to clarify the context of that ‘acceptability’. The core conditions for the intervention’s acceptability included the need for it to be approached sensitively, using a person-centred approach that respectfully acknowledged their history and difficulties with weight, their ambivalence and so on, and also in a way that puts them in control of the decision-making with a positive focus. Delaying LARC removal is not something that would be embraced enthusiastically; it is something that could potentially be tolerated with the importance of the end goal in mind. However, as well as the barriers relating to communication, there were also practical and ethical considerations, ranging from concerns about the practicalities of getting another appointment to a view that this was an unethical intervention that damaged the woman’s right to choose when she could conceive.

Across the two phases of the study, the stakeholder responses led to the conclusion that, in its basic form, an intervention comprising the delay of LARC removal in order to take part in a weight loss programme prior to conception would not be acceptable or feasible. However, including this as one option in a preconception health and weight loss programme that is designed with the key principle of informed choice at its heart would be acceptable and potentially feasible if it also followed the other aspects of the proposed programme theory. This is discussed in objective 6.

Objective 6: future potential intervention based on feasibility and acceptability to stakeholders

A potential preconception WLI has been proposed, designed as part of a healthy pregnancy programme based on feasibility and acceptability to stakeholders and informed by the hypothesised programme theory. It is based on a broad population-based recruitment approach, with individual meetings with a health coach, whose work would be founded on the principles of the COM-B model. 110 Their role would be supporting women to feel competent and confident in relation to their weight across the preconception period and pregnancy and signposting to existing programmes, supplemented by tailored online information. The proposed WLI recruitment incorporates the opportunity for discussion of options presented by LARC removal, but, in recognition of all of the ethical and pragmatic complexities of making that the sole focus, the idea of delaying removal is one potential choice and the eligibility criteria would be much wider. Short-term interventions that simply focus on a relatively brief preconception period have recently been shown53 to be unlikely to work if the aim is to impact on health and weight management throughout the pregnancy rather than, at best, in the first trimester. Therefore, although the focus of the intervention is on introducing change in a 12- to 16-week preconception period, a form of support needs to be incorporated to support women to consolidate the changes over a longer time frame.

Weight management and weight in relation to health are very complex contexts. Ranging from the highly personal to the wider population system, there are layers of meaning to weight that go well beyond the simple ‘calories in, calories out’ model that is often reflected in the interventions delivered. Similarly, LARC removal is not a straightforward procedural opportunity; it is an appointment that has been made in response to an intricate web of life choices and decisions, some of which may relate to pregnancy. Individuals’ beliefs about the health impact of weight and how it fits with other health issues is critical to their engagement with any weight intervention, and this is true for both the service user and the practitioner. One of the messages that came through from the stakeholder work was the current conceptual separation of weight from other components of a healthy pregnancy (e.g. smoking, folic acid). Similarly, there was representation in both groups of stakeholders of the view that contraception is not part of wider health and that somehow it is a separate issue. To improve preconception health, programmes need to reflect this complexity and, as with the potential programme theory described here, draw on middle-range theories that incorporate individual- and systems-level change. As identified by Bull and Willumsen162 in their review of evidence to prevent childhood obesity via the continuum of preconception, pregnancy and postnatal interventions, no single intervention will suffice; action is needed at multiple levels and a life course approach needs to be taken. Preconception programmes must incorporate both individual and societal components, underpinned by socioecological as well as individual behaviour change theories, if we are not simply going to continue making minor changes and exacerbate the social inequities that already exist in nutritional and behavioural aspects of preconception health. 27,29

The ethical considerations in delaying LARC removal as part of the original proposed intervention interlaced two complex sociopolitical issues in weight and conception, which can be captured by the concept of a woman’s right to choose. Although there was never any intention that the intervention would remove women’s choices, the underlying message that some stakeholders identified was discriminatory, suggesting that women with a raised BMI should not become pregnant, and there was perceived to be a risk that women would feel pressured to participate, however unintentionally, by the practitioner’s description of the option to delay LARC removal.

In many ways, preconception does not need a dedicated WLI. Most, if not all, general population WLIs would be suitable or could be adapted. The key differences include the unique primary motivation to succeed; the intervention can also be wrapped in preconception health provision that offers advice on a wider range of important health issues. The stakeholders seemed to feel that this would be more acceptable. However, the reason for this must be considered, along with whether it is a positive step to, in some ways, disguise the focus on weight to make it acceptable. One of the barriers to any new programme, as identified by our stakeholders, is difficulty engaging people, as they need to know that their effort will be rewarded. Being clear about the evidence (e.g. relating to health gains from a 5–10% weight loss) and also about what support will be provided would be positive steps to improving engagement.

Both groups of stakeholders recognised the need for greater awareness of the importance of weight and preconception health at a population level. It may be that investment in a public health campaign (e.g. as suggested by Barker et al. 29 and Stephenson et al. 27) has more of a lasting impact, changing the attitudes and awareness of future generations, as well as those of people who are currently of reproductive age. As identified by Hemsing et al. 37 in a review of preconception interventions, taking a more population-based approach, using gender transformative principles, would also benefit everyone (given the general health benefits of the suggested behaviour changes) and reduce the stigma experienced by women who feel solely responsible for the health of their baby.