Conceptual framework on barriers and facilitators to implementing perinatal mental health care and treatment for women: the MATRIx evidence synthesis

Article history

The research reported in this issue of the journal was funded by the HSDR programme or one of its preceding programmes as project number NIHR128068. The contractual start date was in September 2019. The final report began editorial review in December 2021 and was accepted for publication in October 2022. The authors have been wholly responsible for all data collection, analysis and interpretation, and for writing up their work. The HSDR editors and production house have tried to ensure the accuracy of the authors’ manuscript and would like to thank the reviewers for their constructive comments on the final manuscript document. However, they do not accept liability for damages or losses arising from material published in this manuscript.

Copyright statement

Copyright © 2024 Webb et al. This work was produced by Webb et al. under the terms of a commissioning contract issued by the Secretary of State for Health and Social Care. This is an Open Access publication distributed under the terms of the Creative Commons Attribution CC BY 4.0 licence, which permits unrestricted use, distribution, reproduction and adaptation in any medium and for any purpose provided that it is properly attributed. See: https://creativecommons.org/licenses/by/4.0/. For attribution the title, original author(s), the publication source – NIHR Journals Library, and the DOI of the publication must be cited.

2024 Webb et al.

Background research

Perinatal mental health (PMH) difficulties can occur during pregnancy or after birth. They affect up to one in five women and the cost to the UK is estimated to be £8.1 billion for every annual cohort of women, with 72% of this cost attributable to the long-term impact on the child. 1 PMH difficulties commonly consist of anxiety disorders, depression, post-traumatic stress disorder (PTSD) and stress-related conditions such as adjustment disorder. Many disorders are comorbid2,3 and severe postnatal mental illness is one of the leading causes of maternal death. 2

Perinatal mental health difficulties are particularly important because of the potential negative impact on women, their partners, children and families. This impact varies according to the type of mental illness, severity and timing (e.g. whether pre- or postnatal; acute or chronic; pre-existing or new onset) but, overall, the evidence shows a severe and enduring impact. For example, perinatal mental illness (PMI) is associated with a range of adverse outcomes for women, such as an increased risk of stillbirth4–6 as well as an increased risk of maternal suicide. 2,7–9 PMH difficulties are also associated with a decline in relationship satisfaction and an increased risk of relationship breakdown. 10–12 In terms of infant and childhood outcomes, PMH difficulties are associated with an increased risk of pre-term birth,13,14 and longer-term impacts on children’s cognitive15–18 and behavioural development,17,19,20 as well as being associated with an increased risk of children developing mental health difficulties themselves. 21–23

It is therefore important to identify and assess PMH difficulties quickly so that women who need treatment are able to access it. However, a survey of 200 women living in the UK found that 23% had not sought professional help for their symptoms. 24 It is also estimated that only 30–50% of women with PMH difficulties are identified and less than 10% are referred to specialist care. 25,26 This is likely due to a range of factors at individual, health professional (HP), interpersonal, organisational political and societal levels. For example, HPs not asking about mental health, lack of effective assessment, barriers to women seeking help or attending treatment, clinician barriers to diagnosis and treatment, lack of services to refer on to or limited understanding of effective treatments.

Recognising the barriers and facilitators to identification, assessment, referral and treatment for PMH difficulties is therefore important for health and social care services working with perinatal women. The need for this is evident in calls for research,27 UK strategy and policy,28,29 and clinical guidelines. 30,31 For example, in 2014, the National Health Service (NHS) set out plans for £365 million to be spent on PMH services from 2016–2128 as part of the Five Year Forward View. These services were to ensure 30,000 more women each year would access evidence-based specialist mental health care during the perinatal period. Similarly, the Scottish Mental Health Strategy aimed to improve the recognition and treatment of PMH difficulties. 31 The full implementation of these plans would mean women being asked about their mental health and well-being during antenatal booking visits, being screened for mental health difficulties, assessment within two weeks of referral and being provided with evidence-based psychological interventions within one month of initial assessment. 30–32 Furthermore, in 2019 NHS England set out a Long Term Plan for PMH, pledging an additional £2.3 billion a year and stating that by 2023/24, 66,000 women with moderate to severe mental health difficulties should have access to specialist care from pre-conception to 24 months postnatal. 33 Identifying barriers to women accessing treatment, as well as barriers to implementing PMH assessment and treatment in NHS services, is therefore important to inform these initiatives.

This evidence synthesis therefore aims to identify potential barriers and facilitators to identification, assessment, referral and treatment of PMH difficulties across the care pathway, both in terms of women accessing care or treatment, as well as in terms of NHS services implementing new assessment and treatment initiatives. This will be used to inform a conceptual framework of barriers and facilitators to PMH care that will inform health care services and practice, care pathways, and highlight where further research is needed.

Evidence explaining why this research is needed now

Perinatal mental health is a priority for UK strategy and policy,28,29 clinical guidelines,30,31 HP organisations26,34,35 and third-sector organisations. 36,37 While there have been large improvements in PMH service provision since the publication of the Five Year Forward View,28 in a progress review carried out in 2017 by the All-Party Parliamentary Group on Mental Health, The Royal College of Psychiatrists stated that teenage pregnancies, care beyond the baby’s first year and comorbid substance abuse remained areas lacking in focus and investment. The King’s Fund also reported that in some local areas, recommendations for expanding PMH services were being achieved by retraining existing staff without employing more, an approach that is not sustainable in the long term. 38 Furthermore, in 2020 the Maternal Mental Health Alliance identified that 20% of clinical commissioning groups (CCGs) in England still did not have specialist PMH services. These gaps in specialist PMH service provision are even higher in Wales (71.4%) and Scotland (85.7%). 39 These treatment gaps may mean women are not accessing the care that they need. 26 Given the recent provision of £2.3 billion a year to PMH services,33 this is a crucial time to understand what barriers exist and how they can be overcome.

Reasons why women are not identified and treated for PMH difficulties are complex and multifaceted and likely due to two broad factors: (1) lack of availability of services, which in the current context is possibly due to difficulties implementing services; and (2) difficulties accessing services from women’s perspectives. These difficulties are likely to occur at multiple levels (e.g. individual level factors, HP factors, organisational and wider political-social factors40) and across the care pathway. 41 The care pathway has been defined by Goldberg and Huxley (1992)41 who provided a framework for understanding how a person reaches mental health services. Their Pathways to Care Model shows how, as a person moves through the care pathway, certain factors act as filters that prevent people from accessing mental health care. The first filter is ‘illness behaviour’, where a person needs to pay attention to their symptoms and then make the decision to seek help. If this is not done, this is the first ‘filter’ out of the care pathway. The second filter is the HP’s ability to recognise mental illness, the third is referral on to mental health services and the last filter is admission to hospital beds.

Difficulties in implementing services and accessing services from women’s perspectives are also likely to be impacted by an environment where health care services are highly heterogenous, with variation both within and between services. In some cases, care pathways and treatments are based on organisational factors or assumptions that are not evidence-based. For example, prior to 2016 some CCGs had never commissioned a PMH service and women in these areas were referred to mainstream adult psychiatry services. 42

Guidelines for implementing PMH services have been developed by both NHS England in 201643 and the National Collaborating Centre for Mental Health in 2018. 44 These state the need for multi-agency working across all levels of care and services, expansion of workforce capacity, working with providers and those with a lived experience, and evidence-based service plans. Despite this guidance, large treatment gaps are still reported. 45 The lack of consistent implementation and the development of future implementation plans suggests it is both timely and important to understand what factors may affect implementation of PMH care and at present there are no reviews on barriers and facilitators to implementing PMH care in the NHS or other health care services.

Several systematic reviews of qualitative literature have identified potential barriers to women seeking help for PMH difficulties. Barriers include a reluctance to acknowledge symptoms and lack of support from partners and family members; a lack of ability for women to talk about their feelings openly due to perceived social pressures and stigma; fear of losing custody of their child; and a lack of knowledge of PMH difficulties among mothers. 46–48 Women also cite practical factors influencing their decision to seek help, such as the cost of treatment, finding childcare, lack of health insurance and transport issues. 49

Other factors that may influence whether women access care or treatment include HP factors. For example, in a survey of general practitioners (GPs) and midwives, just under a third reported having had no training in PMH. 34,50 This lack of training may be particularly critical given the importance of interpersonal factors in whether women seek help and access treatment. For example, qualitative studies of women’s experiences of PMH care suggest having a trusting relationship with a HP and being helped to discuss feelings in an unrushed, conversational manner are associated with increased acceptability and disclosure. 51–53

In terms of organisational factors, research suggests that lack of referral pathways, lack of specialised services and inadequate assessment influence whether women receive treatment or not. 46,49

Social and cultural factors are also important. Different cultural norms and health care systems will influence women and HPs understanding of PMH as well as the availability of assessment and treatment. Attitudes towards mental health, such as stigma, may affect women’s willingness to disclose their symptoms and seek help. 46–48

Over 20 published systematic reviews have explored women’s barriers and facilitators to accessing PMH care. However, each systematic review varies slightly in relation to its aim, inclusion criteria and analysis and no evidence synthesis has been carried out to combine the results into a single body of evidence. This would make it easier for health care providers and policy makers to access the information and use it to inform their decisions. 54,55

Conceptual framework

The above evidence highlights that many factors may influence whether or not women are identified, assessed, referred, and receive treatment for PMH difficulties. These are likely to operate at different levels, and across the care pathway. They may be due to barriers in implementing services, or barriers from women’s perspectives in accessing services.

Critically, there are no reviews on barriers and facilitators to implementing PMH care in NHS or other health care services. In addition, no evidence synthesis has been carried out to combine the evidence of multiple reviews about barriers from women’s perspectives in accessing services into a single body of evidence. Lastly, no reviews have combined information from both implementation literature and barriers and facilitators to accessing PMH care from women’s perspectives. Synthesising this literature coherently would make it easier for health care providers and policy makers to access the information and use it to inform their decisions about PMH services and care. One way to do this is through the use of a conceptual framework. A conceptual framework can be defined as a ‘network, or a plane, of interlinked concepts that together provide a comprehensive understanding of a phenomenon or a phenomena’. 56 The development of a conceptual framework can highlight areas for improvement and provide an empirical basis for recommendations for future practice and research.

Definitions and scope

The literature on PMH, identification, assessment and treatment is complex so it is important to define the key terms and scope of this synthesis.

Perinatal mental health difficulties include common affective disorders experienced during pregnancy or the first year after birth, such as depression and anxiety (e.g. generalised anxiety disorder, phobias, panic, obsessive compulsive disorder), stress-related disorders (e.g. acute stress disorder, PTSD), adjustment disorders, and other psychiatric disorders (e.g. psychosis, personality disorders). Symptoms can be mild, moderate or severe. All PMH difficulties were included. We excluded substance misuse disorders because they raise unique challenges in terms of assessment and treatment that may not be generalisable to other disorders.

We define PMH care as identification, assessment, referral and treatment for PMH difficulties.

How assessment of PMH is conceptualised is important. In particular, the distinction between assessment and case identification is important because they have different implications in terms of barriers and facilitators to accessing treatment. PMH assessment refers to identifying women who may be at risk for PMH difficulties, or who have PMH difficulties. Case identification uses psychiatric definitions of disorders, such as the American Psychiatric Association’s Diagnostic and Statistical Manual57 to identify women who fulfil diagnostic criteria for a disorder. Women who meet criteria (i.e. cases) are more likely to reach the threshold for onward referral and treatment. In this project we include both assessment and case identification to examine their implications in terms of barriers and facilitators to accessing care or treatment.

Perinatal mental health treatment refers to any treatment or prevention strategy to reduce PMH symptoms. Services offering treatment vary widely. In primary care or maternity care GPs, midwives and health visitors may offer supportive care for women with mild or moderate PMH difficulties. Evidence suggests primary care is the main provider, with 90% of common mental disorders being managed in primary care. 41 Specialist mental health services vary. For example, the NHS England specialist PMH services target the top 5% of women, that is, those with the most severe difficulties. These are likely to be women with severe postpartum depression, psychosis or other complex disorders, many of whom may require inpatient psychiatric treatment. In contrast, NHS Talking Therapies (formerly known as Improving Access to Psychological Therapies, IAPT) is a community-based outpatient service that predominantly treats moderate affective disorders such as anxiety, depression, PTSD and adjustment disorders. Increasing numbers of NHS Talking Therapies services now have a PMH specialist. In addition, there are increasing community services provided by third-sector organisations, such as peer support services for women with moderate PMH difficulties. These different settings (primary care, maternity care, psychiatric and specialist PMH services) will have different barriers in terms of implementing services and women accessing treatment. In this project we aimed to consider different health and social care settings throughout.

Some services provide interventions focused on outcomes associated with poor PMH. These include parent-infant, couple or family interventions. In this project we focus on interventions for maternal mental health and exclude literature that focuses on mother-infant, couple or family interventions because these are aimed at associated outcomes rather than PMH per se and raise different issues in terms of accessing services.

The scope of this project is also on PMH of women and those who identify as women, rather than partners or those who identify as men. Partner’s and men’s PMH is important but is an area that requires research and evidence synthesis in its own right. Current research into partner’s and men’s PMH is sparse compared to research on women, and there are likely to be differences in barriers and facilitators to PMH care for men compared to women.

Primary research objective

Our primary research objective was to develop a conceptual framework of barriers and facilitators to PMH care (defined as identification, assessment, care and treatment) to inform PMH services, and highlight where further research is needed.

Secondary objectives

Our secondary research objectives were to:

1. determine the barriers and facilitators to implementing PMH assessment, care and treatment in health and social care services;

2. identify differences in barriers and facilitators across different health and social care settings;

3. evaluate the quality of this evidence;

4. extract recommendations for implementation, practice and research based on the barriers and facilitators identified;

5. determine the barriers and facilitators to women accessing PMH care or treatment;

6. evaluate the quality of these reviews;

7. map the geographical distribution of the evidence to establish generalisability and gaps in the evidence;

8. map the findings on to a conceptual framework;

9. conduct a consultation of the conceptual framework and recommendations with a panel of expert stakeholders (e.g. women, GPs, commissioners, third-sector organisations, etc.);

10. make recommendations for practice and future research for PMH assessment, care and treatment.

Research overview

This research used a focused systematic review [Review 1 (R1)], a meta-review of reviews [Review 2 (R2)], conceptual framework and expert stakeholder panel to answer the research questions. It was carried out in three phases: Phase 1 was a focused systematic review of research into implementing PMH care into health and social care services (R1). Phase 2 was a meta-review of reviews into the range of barriers and facilitators to women accessing PMH care (R2). Phase 3 mapped the findings from phases 1 and 2 on to a conceptual framework and refined it through consultations with three expert panels of stakeholders (see Figure 1).

FIGURE 1.

MATRIx study overview.

Patient public involvement

This project was developed with patient public involvement (PPI) representatives from the National Childbirth Trust (NCT) in England (Jennifer Holly and Sarah McMullen) and the Maternal Mental Health Change Agents (MMHCA), a group of women with lived experience of PMH difficulties in Scotland (collaborators). These organisations ensured that we had PPI input from perinatal women generally as well as those affected by PMH problems. Representatives from the NCT and MMHCA co-ordinated PPI input throughout the research and were involved in the dissemination of the project.

Protocol and registration

Both reviews were registered on PROSPERO: R1 PROSPERO (CRD42019142854); R2 PROSPERO (CRD42020193107).

Ethical review

Ethical permission is not required for systematic reviews of available literature.

Search strategy

Literature searches and study selection are reported according to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines. 58

Search terms were identified through hand searches of PMH literature, scoping searches and in consultation with a specialised librarian at the University of Oxford Bodleian Health Care Libraries, Nia Roberts.

To identify papers for R1 we used a mixture of the PICO (Population, Intervention, Comparison, Outcome)59 and SPIDER (Sample, Phenomenon of Interest, Design, Evaluation, Research type)60 research question format to identify four clusters of search terms (see Table 1) relating to:

• Population: NHS and other health or social care services for women in the perinatal period treating women with mental health difficulties.

• Phenomenon of interest: implementing assessment, supportive care or treatment interventions, programmes or protocols for PMH into health or social care services.

• Outcome: qualitative implementation outcomes (e.g. acceptability, feasibility).

• Evaluation: barriers/facilitators.

Pre-planned searches were carried out by a specialist librarian, Nia Roberts, on MEDLINE (1946–present), EMBASE (1974–present), PsychInfo (1806–present) and CINAHL (1982–present). Full search syntax and databases searched can be found in the supporting information of the published review61 and Supplementary material S1.

To identify papers for R2 we used the SPIDER research question60 to identify four clusters of search terms relating to:

• Sample: women in the perinatal period (conception to one year postpartum).

• Phenomenon of interest: assessment, care or treatment for PMH.

• Design: systematic review papers.

• Evaluation: women’s barriers and facilitators.

Pre-planned searches were carried out by a specialist librarian, Nia Roberts, on MEDLINE (1946–present); EMBASE (1974–present); PsychInfo (1806–present); CINAHL (1982–present), Scopus; and Cochrane Database of Systematic Reviews (Issue 8 of 12, August 2021). Full search syntax and databases searched can be found in Supplementary material S2.

Search process

Medical subject heading (MeSH) terms and search terms described above were used to query the databases for literature published from inception to 11 December 2019 for R1 and to 4 August 2021 for R2. Forward and backward searches of included studies were carried out by Rebecca Webb and were completed by 31 March 2020 for R1 and 8 September 2021 for R2.

Eligibility criteria

For R1, the following parameters were used for inclusion in the review:

• Population: NHS and other international health or social care services for women in the perinatal period.

• Intervention: implementing assessment, care, referral pathways or treatment interventions, programmes or protocols for PMH into health or social care services.

• Outcome: implementation outcomes (i.e. barriers, facilitators).

Studies were included if they made statements about factors that either facilitated or impeded implementation of PMH assessment, care, referral or treatment. These statements could be from qualitative interviews with HPs or women; or from studies describing the implementation of PMH care.

For R2, studies with the following characteristics were eligible for inclusion in the review:

• Population: women in the perinatal period (conception to 12 months after birth) experiencing mental health difficulties, who may or may not have decided to seek help, accessed help or engaged in PMH care (defined as assessment, referrals and/or treatment/intervention programmes) from health or social care services.

• Outcome: barriers and facilitators (defined as any individual, HP, interpersonal, organisational, political or societal factors that women believed impeded (barriers) or aided (facilitators) them to seeking, accessing or engaging in help for PMH difficulties).

• Design: systematic reviews that used a PRISMA search strategy.

Studies were included if they made descriptive statements about barriers and facilitators to women deciding to seek help, accessing help and engaging in PMH care. These descriptions had to be drawn from perinatal women’s experiences. Only systematic reviews were included. Reviews that did not use a clearly reported PRISMA58 search strategy were excluded. Reviews were also excluded if they were not conducted on the target population (e.g. focused on men/partners, HPs); focused on substance misuse (which has unique challenges in terms of assessment and treatment); did not focus on the mental health of perinatal women; did not examine any barriers/facilitators regarding seeking help, accessing help and engaging in PMH care; and were non-English publications.

Study selection

For both reviews, search results were initially imported into Endnote and duplicates were removed by a specialist librarian, Nia Roberts.

For R1 animal studies, case reports and book reviews were also removed. Remaining studies were imported into Eppi-Reviewer 4, where results were screened by title and abstract by Nazihah Uddin, the research assistant on the project. A proportion (10%) of the results were double screened by Rebecca Webb, the research fellow on the project. Decisions to include or exclude were concordant between reviewers in 88.11% of cases. Following abstract and title screening, full text screening was carried out by Rebecca Webb. A proportion of full texts (10%) were double screened by Nazihah Uddin and decisions to include or exclude were concordant between reviewers in 90.90% of cases. Disagreements for both title and abstract and full text screening were discussed and resolved by both researchers.

For R2, once duplicates were removed, the specialist librarian also removed papers relating to fetal distress, oxidative distress and those not published in English. Remaining studies were imported into Eppi-Reviewer 4, where results were double screened by title and abstract by Rebecca Webb and Georgina Constantinou, a research assistant in maternal and child health research. An additional proportion (n = 166, approximately 7%) of titles and abstracts were triple screened by Nazihah Uddin. Decisions to include or exclude were concordant between Rebecca Webb and Georgina Constantinou in 94.2% of cases and between Rebecca Webb and Nazihah Uddin in 99.4% of cases. Disagreements were discussed and resolved by all reviewers by applying the relevant inclusion criteria. Once title and abstract screening was complete, full text screening was carried out by Rebecca Webb and Georgina Constantinou. An additional proportion (n = 9, approximately 10%) were triple screened by Nazihah Uddin. Decisions to include or exclude were concordant between Rebecca Webb and Georgina Constantinou in 91.4% of cases and between Rebecca Webb and Nazihah Uddin in 100% of cases.

Data extraction

For R1, data extraction was carried out by Rebecca Webb using Eppi-Reviewer 4 which allows for line-by-line coding. A new ‘codeset’ labelled ‘data extraction’ was created and contained every item to be extracted from the data (e.g. year of publication, country of study). Each paper was read in full, and relevant parts of the text highlighted (e.g. the country of the study) and applied to the relevant code.

For R2, data extraction was carried out using Microsoft Excel (Microsoft Corporation, Redmond, WA, USA) by Rebecca Webb. Each paper was read in full, and relevant parts of the text inputted into the relevant part of the spreadsheet. Methodology of included reviews was copied on to one sheet, and results on to another to aid analysis. Double coding of extracted data was carried out for a proportion of included reviews (n = 3, 10%) by Georgina Constantinou. Data extraction matched in 85% of cases.

The data that were extracted was guided by the Cochrane Systematic Review for Intervention Data Collection form62 for both reviews, and the AMSTAR 263 critical appraisal tool for R2 (see Table 2 for extracted data).

Assessing the robustness of results

For R1, the methodology sections of included studies were assessed for quality with the Joanna Briggs Critical Appraisal Tools for qualitative research,64 cross-sectional studies65 and text and opinion. 66 Each point on the checklists can be coded as either yes, no, unclear or not applicable. Each tool was separated into domains that reflected the question of interest (see Box 1). Where most questions within a domain were answered with yes, this domain was rated as having high quality; where the majority were answered with no, this domain was rated as having low quality. Medium quality was when there was a mixture of yes and no answers.

Rebecca Webb completed the methodological quality assessments for the included papers, and dual critical appraisal of 16 (35%) papers was done by Nazihah Uddin. Nazihah Uddin initially screened nine papers, which were discussed, and conflicts were resolved. Following this screening, the final seven papers were screened by Nazihah Uddin. Coders assigned the same score to 13 (81%) of the 16 papers. All disagreements were discussed and resolved by both researchers, and the final appraisal for these 16 papers is based on agreed answers.

For R2, methodology sections of included reviews were appraised using the AMSTAR 263 criteria (see Box 2). Critical domains in the appraisal of systematic reviews according to AMSTAR 2 include protocol registration, adequacy of literature search, justification of study exclusion, risk of bias, appropriateness of meta-analytic methods, consideration of risk of bias when interpreting results and assessment of publication bias. If more than one critical domain is not met (critical flaw), a systematic review should be evaluated as having critically low confidence in the results of the review. One critical flaw means reviews should be evaluated as low confidence. More than one non-critical flaw means reviews should be evaluated as moderate confidence and no or one non-critical flaw means reviews should be evaluated as high confidence. 63

Given that all studies in this review were qualitative, the AMSTAR 263 items related to meta-analysis were not relevant and were removed. Further, given the debate in the literature regarding the appropriateness of conducting risk of bias assessments on qualitative research, we downgraded the items relating to risk of bias from being a critical flaw, to just a flaw. Quality appraisal of all studies was carried out by Nazihah Uddin and Rebecca Webb. Ratings were concordant in 90% of cases.

A large proportion of the reviews were rated as having low and critically low confidence in the evidence (see Chapter 4). A decision was made to include reviews where confidence in results was evaluated as low and critically low because some of these reviews focused more on marginalised women, such as refugees, migrant women, women with a low income and women living in lower middle income countries (LMIC). Including these reviews ensured that the experiences of these seldom heard women were captured. To improve validity of results, a qualitative sensitivity analysis was carried out to assess whether themes remained consistent across all reviews, regardless of their quality rating. The methods proposed by Harden (2007)67 and Carroll et al. (2012)68 were followed, so sensitivity analysis was carried out in two ways: (1) synthesis contribution; and (2) evidence of adequate description of themes.

To examine whether higher quality reviews contributed more to the themes, a measure of ‘synthesis contribution’ was calculated for each review (as outlined by Harden, 200767). This involved dividing the number of barriers and facilitators identified by the specific review in question (see ‘Number of themes’ column in Table 3) by the total number of barriers and facilitators identified in R2 (n = 62 themes). For example, the findings from Bina (2020)69 contributed to 31 out of 62 themes, giving this review a synthesis contribution score of 50% (see Table 3).

Each study’s synthesis contribution scores were plotted against the number of quality criteria the study met (see Figure 4). Statistical analysis (Pearson’s correlation) was used to help interpret the plots. To examine whether removing lower quality reviews influenced the number of themes, themes that were only supported by lower quality reviews were identified.

To examine whether removing lower quality reviews influenced the description of themes, data were assessed for ‘thickness’ or ‘thinness’ (as done by Carroll et al., 201268). A ‘thin’ description refers to a set of statements, such as this quote about HPs dismissing women’s symptoms:

[The study authors] found that women also felt that providers were downplaying the symptoms they were experiencing. 70 (p12)

A ‘thick’ description provides the context of experience and circumstances71 such as this description of HPs minimising symptoms:

Having symptoms dismissed or attributed to factors other than [postpartum depression] PPD by HPs led to women ‘remaining silent’. Some women perceived that their difficulties would only be taken seriously when there were concerns about risk of harm to themselves or the infant. One woman said, ‘I kept going to this doctor and he used to give me a pep talk and send me home …’. 72(p738)

It is argued that the extent to which a text provides a thick description shows evidence of the authenticity of the results. 73

Data analysis

Review 1 results were analysed by Rebecca Webb using thematic synthesis;98 line-by-line data extraction of statements referring to facilitators or barriers to implementing PMH assessment, care and treatment was carried out in Eppi-Reviewer. Next, codes were re-read and assigned a descriptive theme based on their meaning and content. Themes were developed and revised as each study was re-read. Once all codes had been assigned into themes, these themes were mapped on to a systems level model adapted from Ferlie and Shortell’s (2001) Levels of Change framework40 (e.g. individual level factors, HP factors, organisational factors and larger system factors) and then grouped to reflect different stages of the care pathway adapted from Goldberg and Huxley’s (1992) Pathways to Care model41 (e.g. deciding to disclose, assessment of PMH and access to care and treatment). Mapping of descriptive themes was developed deductively from the initial theoretical framework and then inductively revised as new themes emerged. The mapping of descriptive themes aided the development of the analytical themes. Here, inferred barriers and facilitators were generated. Following this, recommendations for implementing PMH care were drawn from a dictionary of implementation strategy terms and definitions. 99,100

Review 2 results were also analysed by Rebecca Webb using a thematic synthesis98 in NVivo (QSR International, Warrington, UK) and Microsoft Excel. First, line-by-line data coding of statements referring to facilitators or barriers to accessing PMH care from the results section of each paper was carried out. Next, codes were revisited and assigned a descriptive theme based on their meaning and content. Themes were developed and revised as each review was re-read. Once all codes had been assigned into themes, these themes were mapped on to a multi-level framework adapted from Ferlie and Shortell’s (2001) Levels of Change framework40 and the findings from R1. 61 Mapping of descriptive themes was developed deductively from the initial theoretical framework and then inductively revised as new themes emerged. The mapping of descriptive themes was discussed by the project research team before being finalised.

Study selection

Both reviews were reported in accordance with PRISMA guidelines. 58 Enhancing Transparency in Reporting the Synthesis of Qualitative (ENTREQ) research guidelines101 were also followed.

Results of searches are shown in Figures 2 and 3. For R1, database searching identified a total of 21,535 citations. After screening by title and abstract, 10,130 records were excluded, leaving 931 papers to be screened by full text. Screening of full texts left 43 studies to be included in the review. Forward and backward searches identified a further three papers. Therefore, 46 qualitative studies were included in the qualitative synthesis (see Figure 2). Excluded texts are given in Supplementary material S3.

FIGURE 2.

PRISMA flow diagram for R1.

FIGURE 3.

PRISMA flow diagram for R2.

For R2, database searching identified a total of 4086 citations. After duplicates, and studies not meeting inclusion criteria were removed, 2028 articles were left to be screened. Screening by title and abstract led to 1962 records being excluded, leaving 66 papers to be screened by full text. Screening of full texts identified 32 reviews, all of which were qualitative, to be included in the meta-review (see Figure 3). Excluded texts are given in Supplementary material S4.

Characteristics of included studies

For R1, studies were mainly (n = 39) carried out in higher income countries (HICs)102 with well-established highly ranked health care systems. 103 The majority were carried out in the USA (n = 16). Implementation occurred in a wide range of settings including hospitals (n = 14); primary care (n = 12); community-based care, such as community clinics or home visiting (n = 12); online or remote (n = 3); maternity care (n = 3) and specialist PMH care (n = 2). No studies examined implementation in social care settings. Most of the studies (n = 22) looked at the implementation of care services (including screening, referral and treatment); 18 studies were about the implementation of interventions and 6 were about screening only. For the intervention studies most were implementing cognitive behavioural therapy (n = 7) or another type of talking therapy (n = 8). For the screening studies, most were implementing the Edinburgh Postnatal Depression Scale (EPDS)104 (n = 5).

Ten of the studies were descriptive papers, describing the implementation of PMH care. The remaining were qualitative papers, interviewing key stakeholders about their views and experiences on the implementation of the care. Twenty studies interviewed only HPs, nine interviewed only perinatal women and seven interviewed both. Sample sizes ranged from 6 to 809 with a mean of 46.81; median = 24; interquartile range (IQR) = 16.25–33.35 (see Appendix 1 for more details).

For R2, reviews were published between 2006 and 2021 (M = 2017, Mdn = 2018; IQR = 2016–2019). The number of studies included in each review varied from 4 to 40 (M = 16, Mdn = 13, IQR = 9–19) with a total of 344 papers included across all systematic reviews. The number of women included in each review varied from 95–85,190 (M = 5080; Mdn = 463; IQR = 226–1715). The reviews included studies carried out in 24 different countries, the majority of which were HICs, mostly the USA and UK. One review included studies that were carried out only in sub-Saharan Africa. 97 Most reviews (n = 23) focused on perinatal depression, followed by a mixture of perinatal mood disorders (e.g. depression, anxiety, distress; n = 5). Only one study focused on postnatal psychosis105 and one on birth trauma. 92 Included studies were carried out across the perinatal period. Twenty-four of the reviews included studies that reported recruiting ethnic minority women. Six reviews included studies on the experiences of migrant women and/or ethnic minority women (see Appendix 2 for more details).

Analysis of the robustness of the results (sensitivity analyses)

For R1, most studies (n = 44) had a quality rating above 70% suggesting that studies were well-conducted with a low risk of bias. Seven studies were assigned a 100% quality rating (see Appendix 3).

For R2, the majority of reviews were evaluated as having low (n = 14) or critically low (n = 5) confidence with their results. The remainder had moderate (n = 8) or high (n = 5) confidence (see Appendix 4).

As described above, a sensitivity analysis was carried out for R2. In terms of synthesis contribution, there was no correlation between synthesis contribution and the number of criteria each review met (r = 0.142, p = 0.437) (see Figure 4). Furthermore, only four themes (cultural/spiritual causes of mental illness, age, previous diagnoses and appropriateness of care) were only identified by lower quality studies showing that the majority of themes (58 out of 62; 93.55%) were supported by both higher quality and lower quality papers.

FIGURE 4.

Synthesis contribution vs. quality appraisal criteria met for papers in R2.

In terms of richness of data, removing lower quality papers meant that the identified theme Language barriers lost some of its richness. For example, it led to the removal of quotes expressing frustration from women whose first language was not English:

… you don’t know where to go, what to do, who to trust, especially when you are coming by yourself … you believe that you speak English, but when you get here you realize that you don’t … 90(p18)

Sometimes when you have a baby, a woman comes from the hospital. Bengali girls don’t come with the midwife, we don’t understand what they say, we just sit there staring at their faces. 46(p695)

The removal of lower quality papers from the sub-theme Fear of being seen as a bad mum also led to the loss of richness of data, such as the removal of quotes from women who had migrated from their country of birth:

Back home, if someone has this problem, everyone gossips, you get this feeling that people are not dealing with you normally or as if you are abnormal almost …90(p12)

Lastly, the removal of lower quality studies meant important information was removed from the Characteristics of service sub-theme, such as women feeling services prioritised physical needs (n = 2), lack information about screening guidelines (n = 2) and the logistics of care (e.g. location, time of appointments) (n = 3).

Overall, the qualitative sensitivity analysis found that the majority of themes were supported by both the higher quality and lower quality reviews. Including all reviews meant there was more richness in the data and greater inclusion of marginalised women, such as refugees, migrants and women living in sub-Saharan Africa. This sensitivity analysis suggests that the results from R2 can be interpreted with reasonable confidence.

Frameworks for presenting the results

Care pathway

We used a care pathway based on Goldberg and Huxley’s (1992)41 Pathways to Care model to understand how a woman may access mental health services. In Goldberg and Huxley’s model, as a person moves through the care pathway, there are certain factors that act as filters, which prevent people from accessing mental health care. The first filter is illness behaviour, where a person needs to pay attention to their symptoms and then make the decision to seek help. If this is not done, this is the first filter out of the care pathway. The second is the HP’s ability to recognise mental illness; the third is referral on to mental health services and the last filter is admission to hospital beds.

Our care pathway is more detailed and includes the following eight stages: (1) deciding to consult for PMH difficulties; (2) first contact with HPs; (3) assessment/screening for PMH; (4) deciding to disclose PMH difficulties; (5) referral on to appropriate services; (6) access to treatment; (7) provision of optimal care and treatment; and (8) a woman’s experience of treatment (see Figure 5). It is possible that at each stage in the care pathway, a woman may fall through the gaps and ultimately not receive the care that she needs. The decision to disclose has been included after assessment because women have a choice about whether they complete an assessment honestly and thus disclose their symptoms. However, women may also decide whether to disclose their symptoms on first contact with a HP, especially if they are seeking help from their GP. Therefore, it is important to note that a woman may enter the care pathway either stages 1 or 3, and that some parts of the pathway are redundant in health care systems where the woman can contact mental health services directly (e.g. via NHS Talking Therapies services in the UK). Further, the process is not always linear, and some women might jump over certain stages or repeat certain stages.

FIGURE 5.

The MATRIx care pathway.

Determining the barriers and facilitators to perinatal mental health care

System level barriers and facilitators will be described at each step of the care pathway and include results from both reviews (R1 and R2). Please note that in the process of merging results from both reviews a new multi-level factor was added (commissioner level) therefore the levels and themes described may differ slightly to the published papers.

The most commonly cited barriers and facilitators will now be described at each stage of the care pathway, and for each level of the multi-level framework. The stage of the care pathway will be blue, bold and italicised; the level of the multi-level framework being described will be in italic font. For information on all of the multi-level factors at each stage of the care pathway (see Appendix 5).

Barriers and facilitators to perinatal mental health care in different health and social care settings

The most commonly cited barriers and facilitators in different health and social care settings will now be described.

Geographical distribution of evidence and mental health condition examined

Studies in R1 were mainly carried out in Western HICs. Geographical distribution of evidence is shown in Figure 6 and studies listed in Appendix 7. The majority of studies were carried out in the USA (n = 18), followed by the UK (n = 7); Australia (n = 6); Canada (n = 3), Norway (n = 2); and Ireland (n = 2). Remaining studies were carried out in LMICs, including Pakistan (n = 2); Israel (n = 1); Peru (n = 1); Ghana (n = 1); South Africa (n = 1); Uganda (n = 1) and Singapore (n = 1).

FIGURE 6.

Geographical distribution of studies from R1.

In R2 the country that individual studies in each review were carried out in was noted. Where over 50% of the studies in a review were from the same country or area, reviews were assigned the status of mainly: carried out in the UK (n = 9); Western HICs (n = 19); or LMICs (n = 1). Four of the reviews did not provide enough details of individual studies to be able to determine their geographical distribution.

Evidence in both reviews predominantly focused on perinatal depression. In R1 most of the studies (n = 32) focused on perinatal depression. The rest focused on unspecified PMH difficulties (n = 7), followed by anxiety (n = 1) and birth trauma (n = 1). Three studies did not specify which mental health difficulty they were targeting (n = 3). In R2 most reviews (n = 23) focused on perinatal depression, followed by a mixture of perinatal mood disorders (e.g. depression, anxiety, distress; n = 5). Only one review focused on postnatal psychosis105 and one on birth trauma. 92

What is a conceptual framework?

A concept relates to knowledge or an idea about a certain topic. 154 We will use the themes identified in the systematic reviews described above to refer to our concepts.

A conceptual framework can be defined as a ‘network, or a plane, of interlinked concepts that together provide a comprehensive understanding of a phenomenon or phenomena’. A conceptual framework can highlight areas for improvement and provide an empirical basis for recommendations for future practice and research. 56

Why is a conceptual framework needed?

Our primary research objective was to develop a conceptual framework of barriers and facilitators to PMH care to inform PMH services and practice, care pathways, and highlight where further research is needed.

The results from both reviews presented above provide an understanding of barriers and facilitators to implementing PMH care and to women deciding to seek help, accessing help and engaging in PMH care. However, in order to provide evidence-based recommendations for policy and practice related to PMH service provision, it is important that the results of both reviews are synthesised. A conceptual framework is one way of doing this.

Development of the conceptual framework

The method described by Jabareen (2009)56 involves eight stages. We describe how we conducted these stages below.

Integrating concepts

The aim of stage 5 is to integrate and group together concepts that have similarities to one new concept. 56 Concepts were grouped by multi-level factors of barriers and facilitators (see Figures 8 and 9) and the care pathway (see Figures 10 and 11).

FIGURE 8.

Multi-level model of barriers and facilitators to implementing perinatal mental health care (R1).

FIGURE 9.

Multi-level model of barriers and facilitators to women accessing perinatal mental health care (R2).

FIGURE 10.

Multi-level barriers and facilitators to implementation mapped on to the care pathway (R1).

FIGURE 11.

Multi-level barriers and facilitators to women accessing perinatal mental health care mapped on to the care pathway (R2).

Description of the conceptual frameworks

The MATRIx conceptual frameworks aim to understand key barriers and facilitators to PMH identification, assessment, care and treatment in order to improve PMH services. Syntheses of the reviews identified 78 key factors that can impact on PMH care. These are summarised in two conceptual frameworks which provide pictorial representations of 66 barriers (see Figure 12) and 39 facilitators (see Figure 13) across the care pathway and at multiple levels (note: there is overlap with 27 of the barriers and facilitators; see Appendix 9).

FIGURE 12.

The MATRIx barriers to perinatal mental health care conceptual framework.

FIGURE 13.

The MATRIx facilitators to perinatal mental health care conceptual framework.

These conceptual frameworks were used to inform the development of evidence-based recommendations which aim to address these barriers and ensure that all women are able to access the support they need. Recommendations were made for policy, practice and researchers.

Confidence in results

The confidence in the results will now be discussed, based on the results from the CERQual ratings (see Appendix 6). It should be noted that all papers were included in the CERQual ratings and were not split by stage of the care pathway.

Of the 78 concepts identified, 14 were assigned a rating of high confidence with the evidence. These included women’s fear of social services involvement or their child/children being removed from their care; HPs knowledge about PMH services and referral pathways; HPs being dismissive or normalising women’s symptoms; HPs valued characteristics (e.g. warm, genuinely interested, kind, empathetic); trusting relationship between women and HPs; language barriers; adequate workforce provision; high-quality training for all HPs; continuity of carer; culturally sensitive care; lack of appropriate or timely services to refer women on to; and stigma, culture and maternal norms.

Just under half of concepts (n = 33) received a rating of moderate confidence. These concepts included being judged as a bad mum, support (or lack of support) from family and friends; HPs confidence in addressing PMH; HPs not recognising help-seeking or PMI; HPs carrying out assessment in a person-centred way with open and honest communication between women and HPs; collaboration within and between services; clear assessment and referral procedures; services that offer logistical support for women; individualised and appropriate care provision; a lack of clear referral pathways; and the economic status of women.

Slightly fewer (n = 25) concepts received a rating of low confidence, suggesting more research is needed. These concepts included women’s knowledge and understanding of the causes of mental illness, and where to go to seek help; demographic factors such as the woman’s ethnicity or current symptoms/diagnoses; HPs focusing too much on the infant; shared decision-making between women and HPs; co-location of buildings; care with a dedicated mental health champion; and care that offers an opportunity to talk.

Only four concepts received a very low confidence rating, suggesting more research is needed into women’s age or previous diagnoses/symptoms impacting help-seeking and access; the provision of supervision within organisations; and organisational guidelines.

MATRIx framework of barriers to perinatal mental health care

The MATRIx conceptual framework of barriers to PMH care is made up of factors that prevent the implementation of, or women accessing, PMH care. The MATRIx conceptual framework of barriers is shown in Figure 12.

Individual level barriers with moderate and high confidence in the evidence included (in order of evidence confidence): being scared of social services involvement or being judged to be a ‘bad’ mum; having a lack of support from family and friends or them having negative perceptions about PMI; being socially isolated; not understanding HPs’ roles in relation to PMH; not understanding what PMI is, or believing PMI symptoms are due to physical causes, or are a normal part of motherhood; believing the best way to cope with symptoms is to ignore them, or minimise them; and lastly, previous negative experiences of mental health care.

HP level barriers with moderate and high confidence included: HPs being dismissive or normalising women’s symptoms or not recognising help-seeking or symptoms; appearing too busy; having poor knowledge about services, referral pathways and PMH in general; HPs having low confidence about addressing PMH; and lastly HPs carrying out assessment or screening in a tick box or impersonal way.

Interpersonal level barriers with moderate and high confidence were: no trusting relationship between HPs and women; language barriers; and a lack of open and honest communication.

At the organisational/service manager level, barriers with moderate and high confidence in the evidence were: inadequate workforce therefore HP’s workload is too heavy; inadequate provision of PMH training for HPs; lack of continuity of carer; lack of culturally sensitive care; difficulties with technology related to care; lack of collaboration within and between services; lack of logistical support offered by a service; insufficient information provided about the care; inflexible care; care that is not appropriate to women’s needs; confusing wording of assessment tools; assessment of screening viewed as unacceptable; and lastly, unclear or confusing assessment and referral processes within an organisation.

At the commissioner level, all three barriers had high or moderate confidence with evidence, and these were: lack of appropriate or timely services; complexities of funding, resources and financial reimbursement; and lastly, confusing referral pathways.

Political level barriers rated as having moderate confidence with the evidence were: women being a refugee or immigrant, and a woman’s economic status tied in with the cost of health care.

At the societal level, stigma, culture and maternal norms were all rated as having high confidence with the evidence.

MATRIx framework of facilitators to perinatal mental health care

The MATRIx conceptual framework of facilitators to PMH care is made up of factors that aid the implementation of, or women accessing, PMH care. The conceptual framework for facilitators to PMH care is shown in Figure 13. Fewer facilitators to PMH care were identified which suggests more research is needed.

Individual level facilitators with high or moderate confidence in the evidence were women recognising that something is wrong, having supportive family and friends and a strong support network. Previous positive experiences of mental health services were also a facilitator.

At the HP level, facilitators with high confidence ratings were: HPs possessing valued characteristics, such as being trustworthy, empathetic, kind, caring with a genuine interest; going above and beyond to meet women’s needs; and having knowledge of other services and referral pathways. Facilitators with moderate confidence were HPs having similar demographics to women; having good knowledge and understanding of PMH; feeling confident in addressing PMH; making time to address PMH; and carrying out assessment in an individualised way.

Interpersonal level facilitators were the direct opposite to the barriers. Development of a trusting relationship and rapport between HPs and women; and women being able to communicate open and honestly with HPs.

At the organisational/service manager level, facilitators with a high level of confidence were the provision of continuity of carer and culturally sensitive care for women; adequate workforce provision; and provision of optimal training in PMH. Furthermore, technology that worked well and was fit for purpose was a facilitator to PMH care. Facilitators with moderate confidence were individualised, person-centred, flexible care that is appropriate to women’s needs and delivered face to face; the provision of logistical support for women; or care that is delivered at home; group or peer support; sufficient information about available services; collaboration within and between services; and clear organisational assessment and referral processes.

At the commissioner level, one facilitator with moderate confidence was clear referral pathways.

Outputs from the MATRIx conceptual frameworks

The MATRIx conceptual frameworks were used as the basis to develop multiple outputs. The aim of these outputs was ensuring recommendations are disseminated and implemented. These outputs include:

• an animation aimed at HPs and service managers to show what ‘best practice’ services look like;

• an animation aimed at policy makers to show how PMH services should be designed;

• infographics for HPs, service managers, commissioners and policy makers with recommendations on the best ways to design and deliver PMH care;

• an infographic for women and families about ways to navigate an imperfect system;

• a summary report with more detailed information about the MATRIx study, the conceptual framework and recommendations;

• a website with clickable links, quotes and information about each part of the conceptual frameworks;

• dissemination events during Maternal Mental Health Awareness Week (May 2022) where the conceptual frameworks and recommendations were presented to large audiences, and a Question and Answer session was held.

International recommendations for implementing perinatal mental health care

The barriers to implementation identified in R1 were formulated into recommendations. Where consistent barriers were identified (e.g. lack of training), a recommendation to overcome this barrier was made (e.g. provide health care professionals with training). Where consistent facilitators were identified, a recommendation to utilise this facilitator was made. Finally, implementation strategies that matched the recommendations were drawn from a dictionary of implementation strategy terms and definitions. 99,100 These recommendations were made from an international perspective and are summarised in Table 7 and are discussed briefly below.

R1 found that for successful implementation and delivery to occur, PMH care had to meet women’s needs. Therefore, the evidence suggests that service design may benefit from co-production. One potential way this could be done is illustrated by a UK based co-production service (Croydon Service User Network) where both the design and delivery of care is carried out by professionals and service users. This is a network where members participate in the running of the service, feedback their views and work alongside staff to help run group care programmes. 165 In addition, services could utilise toolkits such as The Co-Production Star which ‘enables organisations … to map how much co-production is already taking place, improve existing co-production approaches, identify the potential for new approaches and scale out co-production across services and communities’. 166

R1 found that choice of care and personalised care was key to implementation and delivery. Therefore, the evidence suggests that in services where this is not already implemented, multidisciplinary teams should be created to facilitate choice and personalised care and ensure an adequate workforce to meet women’s needs. The building of a coalition of health visitors, midwives, GPs, therapists, psychologists and psychiatrists is needed to encourage referral and reduce the risk of women falling out of the care pathway. Collaboration between services is also needed with a focus on the identification and building of working relationships and networks with other services (e.g. Citizens Advice).

Many individual level barriers to implementation and delivery of PMH care identified in R1 related to logistical issues. Women could not always attend appointments because of the inflexibility of services. The evidence suggests that increasing the flexibility and accessibility of services could be done through offering home visits. Where this is not possible, providing out-of-hours appointments in a location with good transport links and an accessible building to allow for pushchairs or provision of virtual consultations using web-based platforms such as NearMe (approved for use by the Scottish Government)167 or Livi (an accredited NHS supplier)168 is recommended.

R1 found that technology can be a facilitator to implementation, for example through reminders to assess women’s mental health,116 online referral systems127 and online interventions. 114,146 However, where technology was not fit for purpose, this was a barrier. This evidence suggests that technology should be co-produced with HPs and women, to ensure ease of usability and integration into the workflow. Evidence outside of the review suggests it is also important that in-person consultations are offered where possible because of the challenges of using virtual consultations in areas of high deprivation and with those for whom English is not their first language,169 as there is a risk of increasing health inequalities. 170,171

R1 found that a lack of training was a significant barrier to implementing PMH care. The evidence suggests that HPs should be provided with necessary training in PMH in order to provide a high-quality, evidence-based service. Ensuring HPs provide innovative care can be encouraged through creating accreditation or membership requirements and a learning collaborative. It is also important that HPs work in an organisation that supports their efforts to provide high-quality PMH care. Involving executive boards and ensuring HP implementation team meetings will encourage managerial understanding and should therefore promote effective implementation.

The evidence from R1 suggests that funding complexities and difficulties are a barrier to implementation. Funding is required to ensure high-quality care provision. This suggests that funding needs to be available, easily accessible and ring-fenced at a local level in order to prevent essential PMH funds being diverted to other local services. 172 Funding structures may need to be revised depending on the needs of the community in which the service is delivered (e.g. affordable health insurance where free health care is not available). 173,174

International recommendations for perinatal mental health care

The barriers identified from women’s perspectives in R2 about help-seeking and accessing PMH care were used to develop recommendations about the optimal characteristics of PMH care. These are made from an international perspective and are summarised in Table 8.

R2 highlighted a complex interplay of multi-level factors that influence women’s help-seeking and access to PMH care. Societal factors such as stigma, maternal norms and culture play a large role in women accessing care and the effects can be seen in all system levels. Research suggests that public mental health campaigns can increase knowledge about mental illness and improve attitudes about people with mental illness. 175–179 This evidence therefore suggests public health efforts need to be made to increase women’s, families’ and the public’s mental health literacy through public health campaigns, education within the community, such as antenatal education, and at health care appointments.

R2 found that maternal norms were associated with women believing that they needed to be strong and show they could cope. Maternal norms were a barrier to women accessing PMH care. Research suggests that there may be some potential to change societal beliefs around maternal norms through increasing societal expectations about fathers’ roles in the family through more equal parental leave and rights. For example, in countries where parental leave is more equal (e.g. Finland), the uptake of paid paternity leave is higher. 180 Changing society’s maternal norms could also be achieved through increasing women’s equality. For example, research suggests that stereotypes of what a mother or a woman should look like are beginning to change in countries where women have gained more participation in the labour force181–183 and have the right to access contraception and abortion. 184–186 However, research is needed to corroborate these findings.

At the political level, R2 identified that immigration and economic status and health care costs were barriers to women accessing PMH care. The results also show how race and gender interact to influence women’s experiences of the health care system (intersectionality). 187 This finding is supported by research in general health care that has found ethnic minority and migrant women are disproportionately affected by existing barriers to accessing health care. 188,189 R2 found that these barriers include language and communication barriers, stigma, the cost of health care190 and the inability to access culturally appropriate services. 191 The evidence shows the need for equal rights to health care, regardless of immigration or economic status. The evidence also suggests that changes at the legislative level are needed to protect those who have migrated to a different country from being penalised for accessing health care. 190,192

At the organisational level, R2 identified a range of factors that women viewed as ideal care. Women appreciated the opportunity to discuss screening results with HPs and for it not to be filled out as a tick box exercise. 89 In terms of treatment, women wanted the opportunity to talk to someone (a HP or a peer) about their difficulties. 72,75,81,84,94,193 They found peer support offered them a sense of validation which they appreciated. 79 Further, the evidence suggests that the length of treatment should be flexible and based on women’s needs. Women did not want a one-size-fits-all approach but wanted personalised care that was culturally appropriate. 46,48,49,72,75,84,90,91,95,96,193

At the interpersonal and HP level, the characteristics of the HPs were important, as was their communication with women. Women reported that some HPs normalised their symptoms or were dismissive of their attempts to seek help at first contact, or assessment. This could be a reflection of inadequate training. 194–197 Another key training need is cultural sensitivity and cross-cultural understanding of PMH. Some reviews in R2 identified that women were treated in a culturally insensitive way by HPs, and that ethnic minority women were less likely to be offered treatment or be asked about their mental health. 198–206 The evidence suggests that it is therefore crucial that communication, cultural sensitivity and cross-cultural mental health training are provided to HPs.

In terms of individual level factors, many of these barriers can be improved through the recommendations suggested above. For example, improvement of knowledge around mental health is likely to reduce a woman’s fear of judgment, self-stigma and increase her awareness of the symptoms she is experiencing which may encourage help-seeking. 207,208

UK recommendations for policy and practice

Recommendations for UK practice and policy were developed from recommendations provided by the stakeholder consultations and the conceptual framework. During the stakeholder consultations, attendees were asked ‘In your view, what are the top recommendations for clinical practice?’ Answers to this question can be found in Table 9. In terms of the conceptual frameworks, where the confidence with the evidence was low or moderately low, recommendations for future research were made (see Chapter 9). Where a concept had high or moderate confidence in the evidence, a recommendation to enact this concept in practice was made. This was firstly done by reframing the barriers into answers to the question ‘What would help to improve PMH identification, assessment and treatment?’, and by looking at the guidance provided by stakeholder groups in relation to recommendations. Examples of good practice were also taken from the stakeholder consultation events, and from the NHS Future Platform (see Table 10). Each recommendation has a number next to it which relates to the audience the recommendation is aimed at. Additional information related to the recommendations can be found in Box 3. Recommendations are for third-party organisations, HPs, service managers, government and commissioners and the four devolved NHS oversight organisations (e.g. NHS England).

It is important to note that these recommendations were developed during the COVID-19 pandemic, which had a large impact on the delivery of health services. In terms of PMH care, level of demand for PMH services increased, whereas staff capacity decreased due to illness, school closures and staff being redeployed to COVID-19 health care. There was a significant increase in digital service provision and a reduction in face-to-face services. These changes posed challenges for both services and women. The impact of these changes is still being felt now, and the Maternal Mental Health Alliance has called on the UK government to future-proof PMH services against future pandemics and public health crises. They recommend that this is done by the UK government guaranteeing a minimum high standard of mental health care for pregnant women and mothers of young infants. Thus, although the recommendations of the MATRIx conceptual frameworks may be difficult to achieve in the current climate, the recommendations can be used alongside the Maternal Mental Health Alliance’s recommendations to design and deliver best practice, and future-proof care. 229 While recovery from COVID-19 and capacity issues may still be being resolved, it is still important to look towards the future in the design and delivery of services.

Summary of results

The research detailed in this report reviewed the evidence on barriers and facilitators to implementing PMH care, and to women seeking help and accessing PMH care and treatment. Results from reviews were then synthesised and two conceptual frameworks of key barriers and facilitators to PMH identification, assessment, care and treatment were developed. The conceptual frameworks (see Figures 12 and 13) were developed using an eight-stage approach set out by Jabareen (2009)56 and are based on evidence from 46 primary research papers on implementing PMH care (R1), and 32 systematic reviews on women’s barriers and facilitators to accessing care (R2).

These frameworks were used to provide evidence-based recommendations for international and NHS policy, practice and future research. The MATRIx frameworks led to the identification of 22 evidence-based recommendations for practice and commissioning. Despite being aimed at different stakeholder groups, these recommendations are all highly intertwined and the uptake of one would be likely to have positive effects on others. For example, the continuation of prioritising funding for PMH services by the government231 and NHS England33 will impact on the amount commissioners can allocate to PMH services, thus impacting on the workforce, increasing opportunities for continuity of carer models, staff training and other resources such as translators and logistical support.

Relevance to the wider literature on perinatal mental health

At the individual level, barriers identified included no family support for mental health difficulties, lack of awareness or knowledge about PMH, beliefs about medication, reluctance or inability to attend mental health services, previous experiences of mental health services and additional personal difficulties. It is important to note that these individual level factors do not develop in isolation but often compound one another. For example, women’s beliefs about health services and HPs are likely to come from their previous experiences of health services. This is supported by research carried out with other populations such as young people,232 suicidal people207 and refugees233 that found previous experiences of health care influences help-seeking behaviour.

Societal factors, such as stigma, maternal norms and culture, are likely to play a role in women’s fear of judgment about acknowledging PMH difficulties. The systematic reviews showed a clear overlap between feeling judged as a bad mother, maternal norms to be strong and able to cope, and stigma. Further, some cultural understandings of PMI increase this fear and stigma. R2 found that women from black, Asian and Hispanic backgrounds living in Western countries were more likely to believe symptoms of mental illness were seen as a sign of weakness, or failure, and such symptoms were highly stigmatised in their culture. 46,47,89,90,95 This finding is supported by previous research that has found stigma is one of the leading barriers to help-seeking234–236 and that certain cultural beliefs may amplify the effects of stigma. 237–239

Women also faced logistical challenges such as lack of childcare and lack of transport facilities to access care. These factors were linked to political factors such as economic status. For example, where women had low or no income, other factors such as unstable housing took priority. 82 Women who had migrated into a country had additional barriers such as fear of deportation for accessing health care, or an inability to obtain healthcare insurance. 70

At the HP level, a facilitator to implementation was HPs having a positive perception of the care provided. For example, where HPs internalised the value and importance of assessment, they would be more likely to assess women. This is in line with several implementation theories, such as the internalisation aspect of the Normalization Process Theory,240,241 the Diffusion of Innovation Theory and the Technology Acceptance Model, all of which suggest that users’ perceptions of an innovation are important for their decision to use an innovation. 242–244

Other facilitators to implementation were HPs who were genuinely interested in women, took time to listen and were kind and caring. This genuine interest in women suggests that intrinsic motivation, which is where individuals perform a certain action or behaviour for personal satisfaction without any external reward (e.g. praise or money),245 may play an important role in the implementation of PMH care. Health care providers are increasingly utilising payment for performance models,246,247 such as the payment by results system used within the NHS to improve implementation. These models are based on performance improving with extrinsic motivation and, while there is some evidence that this method works,247 the results from this evidence synthesis did not reflect this. Furthermore, for HPs to act on intrinsic motivation to be kind and to care for people, they need to be working in a well-resourced setting and not experiencing burnout. Burnout is associated with compassion fatigue,248 which is a term used to describe HPs becoming disconnected from or desensitised to patients and patients’ families. 249 Both burnout and compassion fatigue are associated with negative outcomes for patients, such as HPs being less engaged with patients, as well as negative outcomes for HPs themselves. 250 Therefore, it is important service providers ensure that they have an adequate workforce and have support in place to ensure staff well-being.

Another HP level barrier was normalising women’s symptoms or not recognising their attempts to seek help, particularly at first contact. This is probably partly due to a lack of time and training. For example, one of the most cited barriers identified across both reviews was HPs’ heavy workloads, meaning they do not have the time to address PMH. For example, research suggests that consultations where mental health problems are discussed take longer, and HPs feel there is not enough time to address concerns fully. 251,252 Research also suggests that inadequate training in mental health is associated with feelings of anxiety and fear around patients with mental illness, a desire to avoid them, and less effective treatment. 194 This may mean HPs close down conversations about PMI prematurely. Women’s decision to disclose is related to HPs getting it right the first time. Therefore, it is important there is adequate time and training for HPs to feel confident addressing PMH, so every woman gets the response she needs the first time, whomever she sees.

At the interpersonal level good communication, allowing for clear information provision and shared decision-making between women and HPs facilitated women continuing along the care pathway. The systematic reviews in R2 suggested that women were not always included in decision-making regarding medication48 or referral72 and this impacted on their experience of care. Poor communication was further exacerbated by language barriers and women experienced culturally insensitive care from HPs. 89,96 Previous research also supports these findings, showing that communication with HPs influences individuals’ experiences of care. 253,254 Good communication between women and HPs is more likely to develop where there is continuity of carer. A real-life example of this is the lack of relationship-based care carried out in general practice. In a survey of 43 GPs, less than a quarter had responsibility for ongoing contact with the same patients, meaning it impacted their ability to see women in the perinatal period on a regular basis. 34 As identified in this work, lack of contact on a regular basis limits opportunities to develop trusting relationships.

At the organisational level, lack of services, the characteristics of the services (i.e. prioritisation of physical health, no language support services, difficult to access location, the need to attend appointments without the baby) and poor collaboration across services were barriers. These barriers have also been identified in other systematic reviews of help-seeking and health care access. 232,255,256 Poor collaboration across and within services is often due to working in ‘silos’. Organisations are often made up of multiple teams, divisions or departments. These can act as physical silos which prevent certain groups of individuals from working with one another. However, silos are not always physically present within an organisation, they may also be based on employees’ beliefs257 (e.g. we are the midwives, they are the doctors). It is argued that silos provide a feeling of safety by keeping ‘those who are not like us’ out. 258 However, this can create an ‘us-and-them’ mentality which can fragment organisations. 259

Furthermore, working in silos promotes groups achieving their own goals, rather than everyone working together to meet an overall goal. 260 It is therefore not surprising that research suggests that silo working within the NHS can have a negative impact on care. For example, one study found that silo working led to increased length of hospital stay. 261 The NHS has published multiple documents on breaking down silo working, both within the NHS and in terms of links with outside organisations. For example, in the Five Year Forward View28 one of the aims was to ensure better integration between health and social care through multidisciplinary working and providing more holistic care. 262 Furthermore, the NHS Long Term Plan33 sees the future of the NHS as investing in and working with the local community to improve the health of those living within that community. 263 Therefore, it is important that these plans are implemented successfully to improve the care provided to these individuals.

Limitations and recommendations for future research

For both reviews, a decision was made to only synthesise literature carried out with perinatal women. This means that the results from this review may not be generalisable to fathers, partners and families. Fathers’ and partners’ PMH is important so this is an area that requires research and evidence synthesis in its own right. Given that the NHS Long Term Plan has set aside money for the delivery of mental health checks for fathers/co-parents of women receiving specialist community PMH care,33 it is important to conduct research to understand the barriers and facilitators of PMH care for fathers, birth partners and co-parents.

These reviews also excluded services for substance misuse because these disorders raise unique challenges in terms of assessment and treatment that may not be generalisable to other disorders. Future research is therefore needed to understand the needs of women who suffer from substance misuse disorders. Furthermore, for both reviews a limitation of the methodology is that only reviews published in academic journals and written in English language were included. Relevant reviews from health services, charities, third-sector organisations and other grey literature may have been missed.

For R1 the large number of citations meant a decision was made to double screen 10% of abstracts so some papers may potentially have been missed. However, the high concordance of the double screening conducted makes this unlikely. Similarly, only 10% of papers included in R1 had dual critical appraisal of methodological quality which may have influenced the results of this appraisal. However, no papers were rejected on the basis of quality making this less problematic. Similarly, in R2 only 10% of studies had duplicate data extraction. However, concordance was high, so it is unlikely that any key themes were missed.

In terms of the development of the conceptual frameworks, the use of CERQual to evaluate confidence in the findings is a strength, but ratings were done by one researcher (Rebecca Webb) which may mean they are slightly less valid. However, the CERQual approach is described thoroughly and specific rules for each of the assessments were discussed and agreed with the research team to ensure ratings were standardised (see Appendix 6).

There are also some limitations in terms of the evidence included in the reviews. For example, there was a lack of research carried out in specialist services or for women with severe PMI. None of the implementation papers in R1 examined these. Only two reviews in R2 included studies of women with postnatal psychosis or postnatal PTSD, and only three included studies directly related to specialist PMH services. There may be different barriers for other PMH difficulties difficulties therefore future research should focus on researching the barriers and facilitators to women with disorders other than depression (anxiety, PTSD, OCD) as well as more severe PMH difficulties, and issues for neurodivergent women (such as those with autism who are at high risk for anxiety and depression). 264,265 Linked to this, only three studies from one review were directly related to admission to hospital beds, which is the fourth filter of the Goldberg and Huxley (1992)41 model. Given the large gaps in inpatient PMH service provision across the UK and globally,266–268 future research is needed that focuses on the implementation of mother-baby psychiatric units, or international equivalents. Furthermore, more research is needed that focuses on the provision of PMH care for women in universal services. UK evidence suggests primary care is the main provider for PMH care, with 90% of common mental disorders being managed in primary care services. 41 Therefore, we need to understand what changes are needed to improve care provision in universal services.

Most research included in R1 and R2 only looked at barriers to PMH care. This can be seen from the conceptual frameworks, where far fewer concepts are included in the framework of facilitators (see Figure 13) compared to barriers (see Figure 12). Furthermore, most of the research identified was carried out in HICs, meaning we still are not able to have a full cross-cultural picture of barriers and facilitators to PMH care. There was some evidence that beliefs that mental health difficulties being caused by spiritual factors were a barrier to women accessing help, because it was believed the best person to help would be a spiritual leader. However, studies carried out with ethnic minority women and those living in non-western countries or cultures were sparse and we were unable to draw conclusions about this. More research is therefore needed to understand barriers relating to more diverse populations and to include women who migrate from LMICs to HICs. Furthermore, research carried out in a variety of different countries may further our understanding of different barriers and facilitators based on health care systems across the world (e.g. free vs. paid health care).

Lastly, no identified studies or reviews focused on the experiences of the lesbian, gay, bisexual, transgender and queer (LGBTQ) community. Carrying out research with the LGBTQ community is a research priority. 269 Furthermore, this project did not assess the needs of fathers/birth partners and therefore it is unclear if the frameworks are applicable for anyone other than women in the perinatal period.

Other individual level factors that were identified by the conceptual frameworks that need further research are women’s beliefs about health services, such as whether they are approaching the correct HP/service, whether the service will have the capacity to help and whether the service offers more than medication. Furthermore, there was some emerging evidence of other individual level barriers such as certain demographics, and women not feeling psychologically ready to receive treatment, or whose symptoms prevented them for engaging with treatment.

At the HP level, one factor that may warrant further research is whether the impact of good supervision is a facilitator to implementing and delivering high-quality PMH care. Furthermore, one barrier to treatment may be HPs not having a good understanding of medication use during the perinatal period, and this should be researched further. At the interpersonal level shared decision-making with women about their future care options may also act as a facilitator, but more research is needed.

The importance of incorporating outcome measures into the conceptual frameworks was based on feedback from members of the research programme management group. This was not identified from the literature and may reflect the nature of service commissioning in the NHS, where services need to show that they are working in order to be recommissioned. Thus, outcome measures to evaluate services need further attention. It is important that outcome measures used are adequately measuring what the services aim to treat (such as a reduction in symptoms and increased quality of life) but that they are also culturally appropriate and sensitive to women’s needs so as not to discourage women for accessing further care due to the completion of inappropriately chosen measures. Previous research has examined this but implementation of it has been poor and may need updating given the findings of this research programme (e.g. culturally appropriate measures). 270 Further research at the organisational level should focus on ideal timing and length of care and whether more open inclusion criteria and clear organisational structures are facilitators to care.

Recommendations for further work

It is important that results from this work are disseminated as widely as possible to ensure positive changes can be made to PMH health policy and practice. Ideas for dissemination were identified during the stakeholder group meetings and are summarised in Table 11.

Based on these stakeholder recommendations, results from both systematic reviews, the conceptual frameworks and evidence-based recommendations we propose a series of possible dissemination strategies shown in Table 12. A knowledge translation framework was applied to these dissemination strategies – the interaction-focused framework. 280 This framework highlights the need to identify the most appropriate mode of interaction and the level of detail that should be provided. Therefore, research was identified about where each stakeholder obtains their information from, and the type of information they find the most helpful.

For example, results suggest that mothers’ beliefs about PMH (e.g. the causes and where to seek help) and fear of judgement and social services involvement may prevent women from seeking or accessing help. It is therefore recommended that infographics, reports and animations for women and their families, similar to those reported above, are developed, which aim to inform families about PMH and break down cross-cultural barriers. These should be disseminated through universal health services, such as GP surgeries and maternity services; via social media and online, for example, websites such as Netmums, Tommy’s and Mumsnet.

Results from the reviews also identified that HPs’ knowledge about PMH can either be a barrier or facilitator to women accessing care and continuing along the care pathway. We therefore recommend the development of good practice guide(s) which cover the different symptoms and appropriate responses, as well as further training opportunities. This could be disseminated via multiple organisations such as the clinical networks for PMH, local maternity system, PMH workstreams, service managers, webinars, social media, unions, professional journals (e.g. Practice Midwife, BMJ, British Journal of General Practice) and training courses (e.g. Readwell and Health Education England Training hubs).

There were many barriers to women accessing and receiving optimal care in the design and the delivery of services. It is therefore recommended that a document for service managers is designed which utilises the results from the reviews to design a good practice guide. This guide could include examples of good practice and ways to overcome barriers such as language barriers, workforce, training and technology issues. Dissemination should be via networks such as the Perinatal Quality Network; Maternity Networks for NHS improvement; NHS Talking Therapies Clinical Networks; PMH Clinical Networks; Academic Health Science Network; Health Education England Training Hubs and the Community Mental Health Transformation work.

The conceptual frameworks (see Figures 12 and 13) and reviews identified commissioning and political barriers that also prevent women from seeking help, accessing care and experiencing optimal care. It is therefore recommended that a policy and commissioning guide is developed which covers topics such as ‘what is PMH and why is it important?’, ‘what should PMH care look like?’ and ‘how can this be commissioned?’, as well as how to break down postcode lottery in service provision. This guide could be disseminated to local commissioning groups and disseminated via PMH clinical networks in England, Wales, Scotland and Northern Ireland.

Finally, results suggest that stigma is still entrenched within societies and is a barrier across the care pathway. It is therefore recommended that future work focuses on the development of reports, infographics and animations, such as those mentioned above, that aim to break down this stigma. This could be through psychoeducation about mental illness and/or public mental health campaigns. This could be developed in collaboration with charitable and third-sector organisations such as Mind, The Maternal Mental Health Alliance and the 1001 Critical Days All-Party Parliamentary Group.

Future work should be carried out to build on the dissemination strategy in Table 12. Research suggests that active, targeted and multifaceted dissemination strategies are more effective in encouraging behaviour change,281,282 therefore this should be taken into account. Given the range of findings identified by this research programme, to ensure effective and impactful dissemination, a priority setting strategy of what should be disseminated would be useful. Future work could use a similar method to Barber et al. 283 to identify important findings to be disseminated. Barber et al. 283 propose a three-stage method to identify key priorities: (1) stakeholder meetings to identify key priorities; (2) literature reviews to ensure priorities are in line with best practice and existing measures; (3) modified Delphi panel. Once dissemination priorities are identified, dissemination strategies can be put in place to ensure impactful dissemination. A dissemination planning tool, such as the one proposed by Carpenter et al.,284 can be used to help with this. This tool involves identifying exactly who you want to share your results with, working with dissemination partners, ensuring the message is communicated effectively and lastly, evaluating success.

Recommendations for policy

Many elements of the conceptual frameworks can be modified by policy makers and government activity (e.g. workforce provision, health care capacity, training, etc.). Therefore, we recommend that policy makers review the frameworks and take comprehensive, strategic and evidence-based efforts to ensure that there is an effective system of PMH care.

The evidence suggests that funding is required to ensure high-quality care provision. This is particularly important given the impact of the COVID-19 pandemic on PMH services. 229 Therefore, the provision of a comprehensively researched and adequate budget is needed so that all health care needs for that financial year can be met. Funding needs to be adequate for service needs and easily accessible. Funding structures may need to be revised depending on the needs of the community in which the service is delivered (e.g. affordable health insurance where free health care is not available).

The evidence suggests health inequalities are a barrier to PMH care. It is therefore advisable that policy is put in place: (1) to improve equality between the sexes/genders by ensuring equal rights for women and men; (2) in terms of ethnicity, for example, changes at the legislative level are needed to protect those who have migrated to a different country from being penalised for accessing health care; and (3) in terms of income, a fair and easily accessible welfare system is needed to prevent health inequalities based on deprivation.

Recommendations for practice

The evidence suggests that women want choice in the care they receive, and that care is appropriate to their needs. Therefore, it is recommended that care is co-produced with women and is personalised and culturally appropriate. Increasing the flexibility and accessibility of services should be done through offering home visits and, where this is not possible, providing out-of-hours appointments located in an area with good transport links and an accessible building to allow for pushchairs. In addition, service managers could consider the provision of virtual consultations using web-based platforms, but women should be given the choice about whether virtual consultations are the right choice for them.

Culturally sensitive care and increased accessibility of care is required for women who are unable to speak, or have difficulty speaking, the country’s official language. This can be done via pictorial aids, the purchase of products such as Language Line or through collaboration with translation agencies.

The evidence suggests that technology can be a facilitator to PMH services in terms of assessment, referral and intervention. However, where technology is not fit for purpose, this is a barrier. It is therefore recommended that technology systems should be co-produced with HPs and women to ensure ease of usability and integration into the workflow.

The evidence shows that a lack of collaborative working within and between services was a barrier to PMH care. It is therefore recommended that multidisciplinary teams should be created which facilitate choice and personalised care and ensure an adequate workforce to meet women’s needs. We need to break down silo working and encourage collaborative and joint working within and across services. Collaboration between services is needed with a focus on the identification and building of working relationships and networks with other services (e.g. Citizens Advice). Furthermore, the building of a coalition of health visitors, midwives, GPs NHS Talking Therapies practitioners, psychologists and psychiatrists is needed to encourage referral and reduce the risk of women falling out of the care pathway.

The evidence shows that HPs who lack knowledge and confidence related to PMH can be a barrier to care. Therefore, HPs should be provided with high-quality training that is delivered face to face and incorporates role-play simulators where appropriate. This should include training in cultural sensitivity and cross-cultural mental health needs. Training time for HPs should be built into workloads and time protected.

Recommendations for research and third-party organisations

Future research should focus on addressing the gaps identified by this project such as: father/birth partners, severe PMH disorders, specialist and primary care services, diverse samples, outcome measures and facilitators to implementation of and access to PMH care.

Further work and dissemination based on the results of the MATRIx project include the development of infographics, reports and animations for women and their families, the use of public health campaigns to break down stigma, and the development of guidelines for HPs, service managers and commissioners. Furthermore, future work should look at designing a strategic dissemination plan to ensure that the results from the MATRIx study are disseminated as impactfully as possible.

Contributions of authors

Rebecca Webb (https://orcid.org/0000-0002-8862-6491) conducted the research as research fellow and was supported and supervised by the core management team and PPI respresentatives. She wrote the report.

Elizabeth Ford (https://orcid.org/0000-0001-5613-8509) was part of the the MATRIx Programme Management Group who oversaw and developed the MATRIx research programme and part of the core management team who supported and supervised the research.

Judy Shakespeare (https://orcid.org/0000-0003-0770-8098) was part of the the MATRIx Programme Management Group who oversaw and developed the MATRIx research programme and part of the core management team who supported and supervised the research.

Abigail Easter (https://orcid.org/0000-0002-4462-6537) was part of the the MATRIx Programme Management Group who oversaw and developed the MATRIx research programme and part of the core management team who supported and supervised the research.

Fiona Alderdice (https://orcid.org/0000-0003-2404-5644) was part of the the MATRIx Programme Management Group who oversaw and developed the MATRIx research programme.

Jennifer Holly (https://orcid.org/0000-0003-2492-1552) supported and supervised the research as PPI representative.

Rose Coates (https://orcid.org/0000-0002-6944-6428) provided specialist expertise and input.

Sally Hogg (https://orcid.org/0000-0003-1796-0369) provided specialist expertise and input.

Helen Cheyne (https://orcid.org/0000-0001-5738-8390) was part of the the MATRIx Programme Management Group who oversaw and developed the MATRIx research programme.

Sarah McMullen (https://orcid.org/0000-0001-5325-0458) was part of the the MATRIx Programme Management Group who oversaw and developed the MATRIx research programme and supported and supervised the research as PPI representative.

Simon Gilbody (https://orcid.org/0000-0002-8236-6983) was part of the the MATRIx Programme Management Group who oversaw and developed the MATRIx research programme.

Debra Salmon (https://orcid.org/0000-0003-2562-2116) was part of the the MATRIx Programme Management Group who oversaw and developed the MATRIx research programme.

Susan Ayers (https://orcid.org/0000-0002-6153-2460) was part of the the MATRIx Programme Management Group who oversaw and developed the MATRIx research programme and part of the core management team who supported and supervised the research.

All authors were involved in the design of the research, and oversaw and provided input into the conduct of the research and development of the frameworks. All authors read and contributed to the report.

Ethics statement

Ethical permission is not required for systematic reviews of available literature.

Data-sharing statement

Access to data extraction forms, data analysis tables and NVivo analysis documents are available on request to the corresponding author.

Disclaimers

This manuscript presents independent research funded by the National Institute for Health and Care Research (NIHR). The views and opinions expressed by authors in this publication are those of the authors and do not necessarily reflect those of the NHS, the NIHR, the HSDR programme or the Department of Health and Social Care. If there are verbatim quotations included in this publication the views and opinions expressed by the interviewees are those of the interviewees and do not necessarily reflect those of the authors, those of the NHS, the NIHR, the HSDR programme or the Department of Health and Social Care.