Pelvic floor muscle training for women with pelvic organ prolapse: the PROPEL realist evaluation

Article history

The research reported in this issue of the journal was funded by the HS&DR programme or one of its preceding programmes as project number 14/04/02. The contractual start date was in April 2016. The final report began editorial review in March 2019 and was accepted for publication in August 2019. The authors have been wholly responsible for all data collection, analysis and interpretation, and for writing up their work. The HS&DR editors and production house have tried to ensure the accuracy of the authors’ report and would like to thank the reviewers for their constructive comments on the final report document. However, they do not accept liability for damages or losses arising from material published in this report.

Copyright statement

© Queen’s Printer and Controller of HMSO 2020. This work was produced by Maxwell et al. under the terms of a commissioning contract issued by the Secretary of State for Health and Social Care. This issue may be freely reproduced for the purposes of private research and study and extracts (or indeed, the full report) may be included in professional journals provided that suitable acknowledgement is made and the reproduction is not associated with any form of advertising. Applications for commercial reproduction should be addressed to: NIHR Journals Library, National Institute for Health Research, Evaluation, Trials and Studies Coordinating Centre, Alpha House, University of Southampton Science Park, Southampton SO16 7NS, UK.

2020 Queen’s Printer and Controller of HMSO

Introduction

Pelvic organ prolapse (POP) affects women of reproductive age and beyond; it is a condition seen in up to 50% of parous women and up to 75% of women attending outpatient gynaecology clinics. 1,2 Treatment options for prolapse include surgery and conservative management [with pessaries or pelvic floor muscle training (PFMT)]. A high proportion of women go on to have surgical intervention, which is often prone to failure, with the same prolapse recurring or another prolapse occurring in another location, which can lead to repeated operations. 3 Mesh-related complications are frequently reported, and have a removal rate of up to 35%. 4 In Scotland, in 2014, the use of synthetic mesh implants in the treatment of POP was suspended; this suspension was subsequently introduced in NHS England in 2018.

These highly publicised suspensions of synthetic mesh implants make the need for non-surgical options to treat this condition even more pressing. 5 Clear evidence of the clinical effectiveness and potential cost-effectiveness of PFMT in the management of prolapse is now available. This evidence concludes that PFMT should be recommended as a first-line treatment for POP. 6 The Pelvic Organ Prolapse PhysiotherapY (POPPY) trial was a multicentre randomised controlled trial of the clinical effectiveness and cost-effectiveness of individualised PFMT compared with a lifestyle-advice leaflet in women with newly diagnosed symptomatic stage I, II or III prolapse. 6 The POPPY trial constitutes the largest, most rigorous, pragmatic trial of PFMT for prolapse and, as such, provides the necessary evidence to inform future practice. Individualised PFMT provided by specialist women’s health physiotherapists (WHPs) was found to be effective in reducing women’s symptoms of prolapse and in improving prolapse-related quality of life. It also showed potential to be a cost-effective treatment. However, knowledge of efficacy and effectiveness is not enough to ensure implementation.

Background

In the UK, there is currently limited availability, and variation in the availability, of specialist WHPs to deliver PFMT to the large numbers of women who may benefit from it. 7,8 In the UK, there are approximately 800 specialist physiotherapists working in women’s health, as registered with the Pelvic, Obstetric and Gynaecological Physiotherapy (POGP) group. The number of women aged > 40 years in the UK (based on the 2011 census9) is approximately 15.9 million; taking a symptomatic estimate of 10% into account means that there are approximately 2600 symptomatic women for each specialist physiotherapist in the UK. It is unlikely that PFMT will be available to meet the demand unless it can be delivered in other formats, for example by other types of health-care professionals (HCPs).

Implementation science is an emerging field involving complex and multilevel processes. 10 It aims to advance knowledge of implementation by providing generalisable knowledge that will be useful for other settings and contexts. It can help to identify barriers to implementation, but should also extend this to how and why implementation processes are effective. 11 To do this, we need to study implementation strategies and the contexts and processes in which implementation strategies are delivered. Such research is a necessary step in the Medical Research Council’s evaluation of complex interventions framework. 12

Research to improve the implementation of evidence-based PFMT was required. Delivery methods that can enhance service capacity and increase the availability and choice for women are required, but these need to be tested to ensure that the outcomes achieved under trial conditions are maintained. We needed to know whether or not the NHS could deliver PFMT using different staff skill mixes and/or different numbers of sessions and still maintain the benefits observed under trial conditions. We also needed to know how PFMT is implemented in everyday practice and understand the barriers to and facilitators of successful uptake and delivery. It was anticipated that this knowledge would enhance the likelihood of PFMT being rolled out more widely if service models could be successfully tailored to suit different local circumstances and resources, thereby increasing the availability of such services for the many women who would benefit from this treatment.

In addition, trial follow-up rarely extends to more than 1 or 2 years post trial. An observed reduction in ‘further treatment’ following PFMT was initially established in the POPPY trial. A record linkage-based study of longer-term follow-up of the original POPPY trial participants would show whether surgery is prevented or delayed by the use of PFMT. These data would help inform NHS managers as to what long-term benefits they might expect if they implemented PFMT.

Aims

Methodology

Design

This study included the following components.

Realist evaluation

A realist evaluation was carried out that used case studies of implementation of PFMT delivery in three varying NHS settings (see Chapters 4 and 5). The realist evaluation allowed for substantial local stakeholder engagement and for local sites to make decisions on how to deliver PFMT [e.g. using different skill mixes such as specialist physiotherapists, women’s health nurses and junior (band 5) physiotherapists, as well as different numbers of sessions] (Figure 1). The realist evaluation would elicit local folk theories around how implementation was supposed to work [context–mechanism–outcome (CMO) configurations], track how implementation was working (including fidelity to the PFMT protocol) and lead to an understanding of what influenced outcomes.

FIGURE 1.

The PROPEL intervention flow chart. EQ-5D-5L, EuroQol-5 Dimensions, five-level version; ICIQ-UI SF, International Consultation on Incontinence Questionnaire – Urinary Incontinence Short Form; POP-Q, Pelvic Organ Prolapse Quantification; POP-SS, Pelvic Organ Prolapse Symptom Score; PROM, patient-reported outcome measure; PROPEL, PROlapse and Pelvic floor muscle training: implementing Evidence Locally. Reproduced from Maxwell et al. 15 This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated. This figure includes minor additions and formatting changes to the original figure.

Reproduced from Maxwell et al. 15 This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated. This figure includes minor additions and formatting changes to the original figure.

Intervention description

The PROlapse and Pelvic floor muscle training: implementing Evidence Locally (PROPEL) intervention that was being evaluated in this study is fully described using the Template for Intervention Description and Replication (TIDieR) checklist (see Appendix 1). 16 Participants attended an initial assessment and treatment visit with a PROPEL intervention clinician to determine their personalised PFMT protocol, and the patient was provided with any additional lifestyle information deemed necessary to assist with their treatment. Further appointments were scheduled at the patient’s and clinician’s discretion; the sessions progressed using the PFMT protocol until the clinician or patient decided that no further appointments were required, at which point the patient was discharged from the service. Alternatively, if it became apparent that the patient’s condition was more complex than initially thought, the clinician had the choice to refer the patient to specialist physiotherapy. During each appointment, clinicians completed a clinician assessment form to record all information.

Training to deliver pelvic floor muscle training

Staff who were identified in round 2 of the realist evaluation as potential staff to deliver the PROPEL intervention were approached by either specialist physiotherapists or a consultant within their trust. The staff who were to deliver the PROPEL intervention attended a 1-day training session held within their site. The training session was developed specifically for the PROPEL intervention and in conjunction with the POGP. It was delivered by two POGP-registered specialist physiotherapist trainers to a maximum of five new staff per site. In addition, a member of the research team was present at all training sessions to ensure that all questions regarding study specifics could be answered. Training manuals were produced and provided to the participants. Further details on the training are provided in Appendix 2.

Patient and public involvement

The Bladder and Bowel Foundation (BBF) (Kettering, UK) was an initial partner in this research and a co-applicant that contributed to the study design and provided expertise concerning the involvement in this study of women with prolapse. The BBF specifically provided input to the discussions on the delivery options women may receive, on the issues of recruitment of women and on the feasibility of patient data collection processes. They also provided members’ views of their experiences of health care and how they would value opportunities for PFMT, and the likelihood of its acceptability to women with POP.

However, the intention that the BBF would continue to be represented throughout the conduct of this study as a full partner and member of the project management team was not realised, because the BBF ceased to exist just at the point of the PROPEL intervention commencing. We then worked to identify another organisation that could step in to fulfil the role of the BBF. On the advice of our Study Steering Committee (SSC) we approached PromoCon (Worsley, UK), which later became Bladder and Bowel UK, as an organisation that represented people with bowel and bladder problems, and PromoCon agreed to become a part of the PROPEL intervention team. However, the process of understanding the organisational commitment and who would be able to take on the lead role within the PROPEL intervention from within PromoCon was confounded by the organisation’s own impending changes and its move to forming Bladder and Bowel UK. Therefore, it took some time within the PROPEL intervention to secure the support of another patient and public involvement (PPI) organisation; this had an impact on our ability to recruit individual women with experience of POP to sit on our project management group (PMG) and our SSC.

With the help of Bladder and Bowel UK, we finally recruited two PPI representatives: one became part of the PMG and the other joined the SSC. We continued to try and recruit further PPI representatives well into the 2 years after the study initially began, but without any further success. However, the two women recruited to the PMG and the SSC remained with the study to the end and contributed greatly to our meetings and how we communicated with both women and HCPs.

Although our PPI representatives were not actively involved in data collection or analysis tasks, they did provide input to project management, commenting on project documentation and reports to the funder, but, specifically, they were extremely valuable in discussing local site problems. There were difficulties in getting sites up and running and there were issues concerning staff attrition rates (mainly as a result of illness or retirement), which all affected the ability of sites to recruit sufficient numbers of women. Our PPI representatives were sympathetic to these issues, but also offered insight and sometimes solutions, such as contacting other local women’s groups to raise awareness of the study.

Our PPI representatives had always been intended to support dissemination of these findings to lay audiences and were included as key participants in our proposed dissemination and implementation workshops, which were intended, for example, for NHS managers, service leads, urogynaecologists, and physiotherapists with a remit for POP. One of our PPI partners attended two workshop events (London and Glasgow) and was a powerful voice not only in the telling of her own experiences, but also in encouraging managers and HCPs to take the PROPEL intervention findings on board and act on them. The feedback from the dissemination events overwhelmingly rated the contribution of the PPI representative as ‘excellent’.

We will continue to work with our PPI representatives in producing further outputs for lay audiences and will also disseminate these via Bladder and Bowel UK.

The experiences of our PPI representatives of working with the PROPEL intervention are described as follows:

I welcomed the opportunity, my first, to be involved as PPI representative in this project.

In my view anything that reduces the number of women having to undergo surgery can only be good for patients. Having increased numbers of skilled staff and reducing the costs involved in surgery would also be of benefit for NHS trusts.

I did find it difficult to offer any specific input to the project, particularly latterly when I was less able to attend meetings. I found the experience interesting and would certainly participate as PPI in the future.

PPI representative 1 (PMG)

I feel very privileged to have been a PPI on this study as the subject matter is very relevant to my patient experience within the NHS. I work within the NHS (in an unrelated field) so have (a little) understanding of the difficulties faced on a daily basis with staffing and funding issues. But more importantly I have years of patient experience and know how difficult I personally found it to access women’s health physiotherapy in my area. Hence, I appreciated being involved in a study that could not only improve patient outcome, but look at the implementation aspect of providing an NHS service.

Having never been involved in a research project before, I found the team supportive, friendly and above all willing to listen. It is sometimes difficult partake in discussions when you are not experienced or qualified in the field, but I felt that the patient voice was heard. I feel my experience with administrating an online support group for women with pelvic pain and prolapse helped with my ability to voice the patient point of view.

My highlight was speaking about my patient experience at the dissemination meetings. It was well received and I felt it emphasised why the research was undertaken and what a difference it could make to patient outcomes. It was fantastic to hear of the success of the project and that physiotherapy works in terms of patient improvement and cost savings. I look forward to being involved in the next study. Thanks to the PROPEL team for all their hard work.

PPI representative 2 (SSC)

Introduction

The PROPEL intervention initially aimed to include three diverse sites across the UK in which to develop new models of PFMT service delivery. These three sites, A, B and C, had been identified during the funding application process and were keen to be involved in the study from this stage. It became apparent during the recruitment of women to the PROMs study and from the delays in sites to implement the new models that we would find it difficult to reach our recruitment target through these three original sites. Through one of our co-applicants who had previously been on the executive committee of the POGP, we sent out an invitation to around 20 women’s health services throughout Scotland and England. We had a number of positive responses, which resulted in the recruitment of two further ‘light-touch’ sites to the study. By this stage in the study, round 1 of data collection for the realist evaluation had been completed in sites A, B and C and the decision was made, with the agreement of the PMG, not to include these two new sites in the full realist evaluation, but to gain some reflection on their experiences of setting up and delivering PFMT in their regions. This decision was made to avoid further delays in recruiting women to the PROMs study and to maximise follow-up of women, while at the same time adding to our knowledge of ‘what works’ for implementation.

Overview of sites

Womens’ input into service planning for the PROPEL intervention

The women’s focus group data that were collected in round 1 of data collection for the realist evaluation were summarised and fed back at a SPM at each of the sites. The focus group study was designed so that each site had data on the issues that were important to women currently using these services; this included topics around awareness of women in the existing service, referral pathways and treatment received. These data were collected and fed back to sites so that the views of women could be considered in the process of designing a new model of delivering PFMT to women in the PROPEL intervention.

Overall, 26 clinicians were initially trained to deliver the PROPEL intervention (Table 1). In total, six of these withdrew from the study prior to treating any women using the PROPEL intervention protocol. They cited a number of reasons for this, including injury, lack of confidence, lack of capacity, organisational issues and moving away from their current post.

Training

Realist evaluation approach

Realist evaluation belongs to the family of theory-based evaluation approaches. Theory-based evaluations aim to clarify the ‘intervention theory’, that is clarify how intervention activities are understood to cause or contribute to outcomes and impacts. Realist evaluation was used in this study to understand how the intervention was implemented in different study sites, what contextual factors influenced its implementation and what ‘mechanisms of action’ led to successful (or unsuccessful) delivery and outcomes.

Realist evaluation has emerged in response to the need for knowledge that extends beyond that obtained by the traditional outcome-focused evaluation approaches that ask the question – is the intervention effective? Realist evaluation is founded on the premise that interventions are complex and are introduced into social systems that are also complex. It maintains that no interventions are universally effective, but that some things work for some people in some contexts. Realist evaluation therefore asks what is it about the intervention that works, for whom, in what contexts and why. 20

Realist evaluation contends that it is not interventions that work; rather, it is the people involved in interventions who make them work. Interventions introduce opportunities, resources or ideas for change, but whether or not these actually lead to the intended outcomes depends on how people react to, interpret and act on these resources. It is people’s reasoning and capacity in response to the intervention elements that represent the real ‘mechanisms of action’ in any intervention. These mechanisms of action are, however, contingent on the social context in which people work. Certain contexts enable people to act, whereas others place limits on people’s behaviour. 21 Realist evaluation thus seeks to explain the complex relationship between the mechanisms activated by the intervention, the context that influences their workings and the intended and unintended outcomes they produce. The explanatory proposition of realist evaluation is that interventions work (have successful outcomes – O) only in so far as they introduce appropriate ideas and opportunities (mechanisms – M) to groups in the appropriate social and cultural conditions (contexts – C). The task of an evaluation is to identify the linked patterns of contexts, mechanisms and outcomes (CMO configurations) to explain how particular outcomes were brought about by certain mechanisms being triggered in certain contexts. 20

This study aimed to implement a complex intervention, the delivery of PFMT using different staff skill mixes, in complex NHS systems consisting of a number of actors, varying resources, diverse geographical locations and service configurations. The implementation also involved actions and decisions from people in multiple roles at different levels, for example from service managers and finance directors at organisational level to front-line staff delivering and supporting the PFMT intervention, as well as women receiving PFMT. Given the interplay of multiple factors operating in different personal or organisational contexts with different priorities and goals, realist evaluation provided an appropriate framework and methodology to explore and explain the implementation of PFMT.

Realist evaluation typically involves three broad phases. Phase 1 seeks to identify the ‘folk’ theories about how and why the intervention will bring about change. This involves eliciting ideas about how the intervention is expected to be implemented and work, what intended and unintended outcomes are likely, what may be their mechanisms of action and what contextual factors may enable or constrain these mechanisms. Data are gathered from those involved in the implementation of the intervention and its key stakeholders. These data are used to build hypotheses about the causal relationships between specific contexts, mechanisms and outcomes; these are known as the CMO configurations. Phase 2 involves testing these theories by gathering data on the actual implementation process; this unfolds the mechanisms and outcomes and impacting contexts. In the third and final phase, the intervention theories are refined through iterative data analyses and interpretation to provide middle-range theory statements about why and how the intervention worked, for whom and in what contexts.

Phases and methods of realist evaluation

The realist evaluation was conducted in three broad phases, using a longitudinal, multiple case study design. The five study sites described in Chapter 3 were considered as ‘cases’, although the full realist evaluation was conducted in three sites only. Cases were defined at the level of the NHS trust in England and NHS Health Board in Scotland, as these represent the units through which health services are organised, governed and delivered in local areas. Defining the ‘cases’ at the level of these broad units helped to ensure that the influence of contextual conditions at various levels (i.e. from financial, organisational and managerial level to clinician, practice and patient level) was encompassed in the evaluation.

Theory refinement

Phase 1: folk theories of change – how was the intervention expected to work?

In rounds 1 and 2, participants in each site described the wider context in which their service operated in, the current configuration of the service for prolapse care, the perceived gaps in service and the key drivers for change. This description was followed by an account of proposed service models and implementation plans, anticipated and unanticipated outcomes, explanations of how the intervention was expected to work and the existing and potential barriers and facilitators likely to affect implementation.

The PROPEL intervention introduced an opportunity to deliver PFMT using different staff skill mixes to a wider population of women with prolapse than that currently reached by specialist physiotherapy services. The PROPEL intervention provided an opportunity for sites to reconfigure local service and referral pathways, and supported provision of training in PFMT delivery to new staff skill mixes. The PROPEL intervention also provided some limited resources to support the new models of service delivery, such as specialist physiotherapist time to support those newly trained while they engaged in recruiting and delivering PFMT to the study population. Across the three sites, the intervention was expected to affect three key sets of outcomes: (1) have an impact on public health by way of widening the reach and accessibility of PFMT to the target group in local areas, (2) have an impact on women’s health by way of improvements in prolapse symptoms and quality of life and/or reduction in surgeries and (3) have an impact on services by way of shortened waiting lists for PFMT and a reduction in specialist workload, so that their resources can be focused on more complex cases.

The context and organisation of prolapse care varied significantly across the three sites. Several contextual factors were identified in round 1 and 2 data that seemed likely to influence the implementation of the intervention, which, in turn, would affect the achievement of anticipated outcomes. In the following section, we describe the context in each site before articulating the initial CMOs that emerged from data explaining how the intervention was likely to work and what factors were expected to facilitate or impede the implementation.

Theories about how the intervention would work, for whom and in what contexts

Four sets of CMO configurations were identified from the data, which contained folk theories around how each of the intended outcomes would be brought about and what may facilitate or impede these processes (Figure 2). The data also revealed an unintended outcome that was expected to affect implementation. Each CMO is described briefly and illustrated via a diagram.

FIGURE 2.

The CMO configurations 1–4. C, context; M, mechanism; O, outcome. Reproduced with permission from Abhyankar et al. 17 This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data. The figure includes minor additions and formatting changes to the original figure.

Reproduced with permission from Abhyankar et al. 17 This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data. The figure includes minor additions and formatting changes to the original figure.

Context–mechanism–outcome configuration 1: widening the reach of pelvic floor muscle training through increased local provision of care

It was anticipated that using different staff skill mixes to deliver PFMT would widen the reach and accessibility of PFMT in local areas because it would increase the number of referrals to PFMT and increase the provision of PFMT in the community, closer to women’s homes, through adoption of a community model of service delivery. This mechanism was dependent on whether or not adequate facilities (e.g. private rooms in clinics) were available to carry out internal assessments and deliver PFMT, whether or not there was strong leadership in services to support the work and whether or not GPs and other potential referrers were aware of PFMT and new referral options. In sites B and C, there were concerns that a lack of appropriate facilities may be a challenge to successful implementation.

Context–mechanism–outcome configuration 2: improving women’s health outcomes through holistic and proactive care

The intervention was also expected to improve women’s symptoms, improve quality of life and reduce the need for surgeries. This would be achieved by enabling the newly trained staff to perform more holistic assessments of urinary incontinence and pelvic health issues, and provide proactive treatment in the form of PFMT by enabling them to tailor the treatment and advice based on accurate internal assessments and by informing women of the correct PFMT technique. Whether or not these mechanisms would materialise depended on whether or not there was adequate support from the service leads and/or management in terms of ensuring that there was dedicated time for staff to deliver PFMT; whether or not there was a sufficient workforce to deliver existing and new services, and whether or not they had manageable workloads; whether or not ongoing support by specialist physiotherapy staff was available to newly trained staff; and whether or not the differential training needs of different skill mixes (e.g. nurses vs. physiotherapists) were acknowledged and addressed. In sites B and C, the staff shortages and existing staff roles and workload were expected to act as barriers to successful implementation. Worries were also expressed about the impact on the workload of the newly trained staff if the demand for the number of referrals increased as a result of greater awareness of PFMT among GPs and women.

Context–mechanism–outcome configuration 3: improving service organisation through joined-up working

The intervention was expected to improve service delivery by reducing specialist workload and shortening waiting lists. This was expected to result from more joined-up working between physiotherapy teams and nursing teams in the community and acute settings. This was possible only if there were an adequate number of staff trained to deliver PFMT and if the referral pathways were straightforward. There were some concerns that the pressure to reduce the waiting lists may actually lead to inappropriate referrals to newly trained staff, which, in turn, would increase rather than decrease the waiting times and the specialist workload.

Context–mechanism–outcome configuration 4: implementation difficulties due to role protection issues

In addition to the intended mechanisms and outcomes, the interviewees anticipated a fourth, rather unintended, mechanism that had already started to unfold and affect the implementation processes during the initial stages, leading to some undesirable outcomes. With the intervention bringing the prospect of training other skills mixes, particularly staff in the lower banding/grade, this triggered perceptions of threat to the role of practitioners specialising in pelvic health or PFMT and activated feelings of role protection among these staff. A reluctance was sensed among the specialist staff, particularly in sites A and C, to train nurses of a lower band to perform the higher-skilled tasks (e.g. deliver PFMT, which has been a specialist job), as it was felt to be causing a disservice to the specialist profession. This reluctance was also observed in the form of some ‘hostility’ at the SPMs. The feelings of role protection were said to be always present, but these were observed to resurface and intensify as a result of the PROPEL intervention. This mechanism of ‘role protection’ hindered and delayed the process of implementation in sites A and C, and prevented the adoption of a community model of PFMT delivery involving band 5 community nurses. Site A continued their service through hospital- and community-based specialist physiotherapists without making any changes to the service model, whereas site C had to abandon the plans for training nurses or physiotherapists in the community and adopted a hospital-based model involving hospital-based urogynaecology nurses.

Phase 2: testing the folk theories

The way in which the PROPEL intervention was implemented and worked in each site is presented in Tables 3–5 in the form of site-specific CMOs. These CMOs report the outcomes of the implementation that were observed in each site and attempt to explain how and why these outcomes were achieved, for whom and in what contextual conditions. The outcomes include those that were expected in the initial theories of change, as well as those unintended and unanticipated.

Phase 3: refined intervention theory – how did the PROPEL intervention work, for whom and in what contexts?

The site-specific CMOs were compared and contrasted in a cross-case analysis to develop refined theories about the intervention. At this stage, the outcomes were also mapped explicitly to the indicators of impact that were suggested by the RE-AIM framework. During this process, we classified the specific outcomes of the PROPEL intervention as being related to one of the RE-AIM framework indicator areas, and sought to explain the patterns of outcomes observed in the theory-testing phase by comparing and contrasting their linked patterns of mechanisms and of triggering contexts in different sites. We first looked for the common patterns of mechanism and outcome links and then identified the contexts in which they were contingent. In the following sections, we present the refined CMOs that represent a middle-range theory about the impact that the PROPEL intervention had, what mechanisms of action brought about that impact, for whom and in what contexts.

Reach: impact on access to pelvic floor muscle training

In the PROPEL intervention, ‘reach’ referred to the extent to which the PROPEL intervention, the delivery of PFMT by different HCPs, reached the target population that it was intended to reach. It was hypothesised that the intervention would widen the access to PFMT by increasing the awareness of and referrals to PFMT and increasing the availability of trained staff who were delivering care close to women’s homes in the community. The extent to which this was achieved following implementation of the intervention varied across the study sites, depending on the presence or absence of two key contextual factors: (1) the receptiveness of the clinical setting and (2) the level of awareness of PFMT among potential referrers. The CMOs were refined based on this knowledge and are described in the following paragraphs and illustrated in Figures 3 and 4.

FIGURE 3.

Refined CMOs explaining the impact of the PROPEL intervention on access to PFMT: role of receptiveness of service. C, context; M, mechanism; O, outcome. Reproduced with permission from Abhyankar et al. 17 This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data. The figure includes minor additions and formatting changes to the original figure.

Reproduced with permission from Abhyankar et al. 17 This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data. The figure includes minor additions and formatting changes to the original figure.

FIGURE 4.

Refined CMOs explaining the impact of the PROPEL intervention on access to PFMT: role of team’s awareness of PFMT services. C, context; M, mechanism; O, outcome. Reproduced with permission from Abhyankar et al. 17 This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data. The figure includes minor additions and formatting changes to the original figure.

Reproduced with permission from Abhyankar et al. 17 This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data. The figure includes minor additions and formatting changes to the original figure.

The access to PFMT widened in areas where the service was receptive to the PROPEL intervention (Figure 3). This included the staff and management having a keen interest in pelvic health issues and women’s health, the service already having started providing additional training to other staff to increase their capacity and having good peer support networks in place. This was mainly observed in the case of musculoskeletal and general physiotherapists, and some nurses. In these contexts, the PROPEL intervention training enhanced staff’s knowledge of prolapse and their skills and confidence in its assessment and treatment. The transition to doing internal assessments and PFMT was easier and faster for these skill mixes as a result of their basic physiotherapy training or continued professional development training in pelvic health assessment and management. Nursing staff who dealt with urinary incontinence extended their knowledge and skills beyond prolapse and applied this to women with incontinence and other pelvic health issues. Their enhanced knowledge led them to become more holistic in the assessment of pelvic issues and proactive in providing treatment.

However, different sets of mechanisms and outcomes were observed in the case of nurses working in the community who had no prior experience or training in pelvic health. A key underlying mechanism was low levels of adoption by community nursing staff, which was reflected in their withdrawal from the study. For the district nurse, the training was felt to be inadequate to address their knowledge and skill gap, as they did not have the level of understanding of physiology that physiotherapy training provided. This lowered their confidence and led to withdrawal from the study. For the advanced nurse practitioner for incontinence, their existing workload and a mismatch of expectations from their role led to withdrawal. The community setting was considered inappropriate for prolapse assessment and PFMT delivery; internal assessments were not feasible in people’s homes or care homes and PFMT was deemed unsuitable for older people and those with dementia, both of whom comprised the main aspect of the community nursing role. As a result, contrary to original intentions, PFMT failed to be extended to community settings.

Another reason for the difference in the reach of PFMT was the level of awareness of PFMT services among professionals who could refer women to PFMT (Figure 4). In contexts where the teams made an effort to raise other HCPs’ awareness of PFMT, there was good communication and co-ordination among referrers and there were favourable attitudes to PFMT, access to care was widened. This was because there were a variety of direct referral routes available: GPs and consultants were referring patients to these services, as well as women who were more aware of PFMT spreading the word in community. In contrast, the referrals were restricted in areas where not much effort had been made to raise awareness among potential referrers, which meant that their referral patterns remained unchanged. Referrals were also restricted because the consultants remained distant and disengaged in the PROPEL intervention, and other possible referral routes were not exploited (e.g. through urogynaecology nurses). Referral involved study recruitment, which was difficult to do in busy, short-staffed clinics. The process of referral to the PROPEL intervention was seen as being more complex than routine referrals hence it was often forgotten.

Effectiveness: impact on symptoms, quality of life and care experience

Effectiveness in the context of the PROPEL intervention was concerned with the impact of PFMT delivery by different staff skill mixes on women’s symptoms, quality of life and experience of care. It was anticipated that the PROPEL intervention would improve women’s symptoms and quality of life, and reduce the need for surgery by improving their PFMT technique, providing better tailoring of treatment and improving holistic assessment of urinary incontinence and related issues. The outcome study revealed that there was a significant improvement in outcomes after the intervention compared with before, but there were no significant differences in outcomes across the five sites and across different delivering staff and across different service models. This suggested that the outcomes were comparable regardless of which study site PFMT was delivered in, by whom and through which model. However, the staff’s and women’s perceptions of their outcomes were found to differ from the actual clinical evidence of improvement and across different sites, owing to differences in implementation processes. The refined CMOs below explain these differences in perceived outcomes (Figure 5).

FIGURE 5.

Refined CMOs explaining the impact of the PROPEL intervention on women’s health. C, context; M, mechanism; O, outcome; UI, urinary incontinence. Reproduced with permission from Abhyankar et al. 17 This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data. The figure includes minor additions and formatting changes to the original figure.

Reproduced with permission from Abhyankar et al. 17 This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data. The figure includes minor additions and formatting changes to the original figure.

Many women and staff reported improvements in their prolapse symptoms (e.g. muscle control and strength) and incontinence problems. In some cases, the experience of improvement was greater than the clinical evidence. When improvements in symptoms were reported, the underlying mechanisms appeared to include women receiving better explanations of prolapse and the role of PFMT from the delivering staff, receiving feedback on their performance from internal assessments, provision of a tailored and structured PFMT routine, receiving advice on the correct PFMT technique and receiving techniques for exercising regularly, all of which improved women’s adherence to exercises. The reported improvements in quality of life resulted from staff enabling women’s self-management of their symptoms by offering tips and advice on lifestyle, which helped to improve women’s bladder control and increase their confidence to resume previous activities. There was an overall person-centred approach, as staff were seen as approachable and motivational, to dedicate enough time, to reduce women’s embarrassment and to make women comfortable during their visits. This resulted in a positive and satisfactory experience of care. All of the above mechanisms were triggered in contexts where the staff were adequately trained and confident, where the staff had expert support available throughout and when the referrals were appropriate for the level of care they provided.

In contrast, certain contextual factors failed to trigger the mechanisms that led to perceptions of improvement in symptoms, quality of life and care. This was particularly the case when the staff delivering PFMT were located in an acute ward setting. Both the delivering staff and the women reported less improvement in symptoms, despite the clinical improvement that was observed in the PROMs study indicating otherwise. A number of factors present in these contexts seemed to trigger certain mechanisms that led to a lesser experience of improvement. First, there was a time gap between training and PFMT delivery that led to an initial loss of confidence among delivering staff. Second, the specialist physiotherapy team was located separately and there was a lack of clarity and communication on both sides about the delivering staff’s support needs. This triggered another mechanism, whereby the staff resorted to supporting each other by doing joint clinics, which helped with care and their confidence.

Adoption: impact on service delivery

Adoption refers to the willingness of institutions and staff to implement an intervention and to support the intervention’s adoption into routine practice. In this study, one of the indicators of adoption, the uptake and continued participation (or drop out) by staff, was already reflected in the mechanism, leading to wider (or restricted) reach of PFMT to target populations. Another indicator of the extent to which the intervention was adopted into practice is the effect of the PROPEL intervention on the services in which PFMT was delivered by other skill mixes. Favourable impact on service outcomes would indicate higher levels of adoption, whereas unfavourable impact would indicate lower levels of adoption. It was hypothesised that the intervention would reduce waiting times and specialist workloads by creating more joined-up working between acute services and community services. Although there was little evidence for this mechanism in the theory-testing phase, a number of other outcomes and their underlying mechanisms ensued, for instance, to impact workloads and pressures on delivering staff, as well as their professional colleagues, the organisation and functioning of the services and the perceived support from wider team (Figure 6).

FIGURE 6.

Refined CMOs explaining the impact of the PROPEL intervention on service delivery. C, context; M, mechanism; O, outcome. Reproduced with permission from Abhyankar et al. 17 This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data. The figure includes minor additions and formatting changes to the original figure.

Reproduced with permission from Abhyankar et al. 17 This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data. The figure includes minor additions and formatting changes to the original figure.

The intervention was better integrated and adopted into the routine service in some areas than in others, and had little disruptive impact on workload, service organisation or waiting times in such areas. This was observed particularly in contexts where practitioners, mainly physiotherapists and, in some settings, nurses, had some degree of autonomy over managing their bookings and caseloads and the wider team supported their involvement in PFMT delivery. Their service management also provided support by allowing dedicated time for PFMT delivery, making rooms available for assessment and treatment, and providing adequate staffing resources. This triggered the mechanism wherein PFMT was seen as part of their core role, rather than as an add-on. It also enabled the staff to juggle their own caseload to allow for the prolapse patients.

Contrary to this, service disruption was reported in some areas in the form of disturbed ward organisation, delays in treating routine patients, and increased pressure on delivering staff, particularly in the cases of nurses in the acute ward setting and the community setting. This was because PFMT delivery in these settings was seen as an add-on and not given a priority or due recognition. Staff had to juggle two roles and required additional administrative support. Team colleagues were perceived to be discontent with staff’s involvement in PFMT, as they often had to provide cover for the PROPEL intervention staff in their day jobs. This resulted in doubts about the extent to which such an intervention may be supported by the wider team. Several unfavourable factors in the context played a role in triggering these mechanisms. First, delays in implementation and inflexibility of the patient booking systems meant that slots for the PROPEL intervention patients remained unfilled during early phases and led to delays in the treatment of routine patients, as slots could not be allocated to them. Second, existing workload pressures in the service areas, as well as a lack of dedicated time for PFMT, a lack of appropriate facilities and inadequate staffing resources, put additional pressure on delivering staff.

Implementation: impact on implementation plans

Implementation refers to the extent to which the PROPEL intervention was delivered/implemented as intended by the services. During the implementation phase, the services in sites A, B and C considered and debated various models of service delivery and potential staff skill mixes, with the overall intention of widening access to PFMT in local communities. Whether or not this intention was accomplished was determined by the strength of the tendency of staff groups to protect professional roles and boundaries (which could prevent change), which was triggered in certain contextual conditions (Figure 7).

FIGURE 7.

Refined CMOs explaining the impact of the PROPEL intervention on implementation plans. C, context; M, mechanism; O, outcome. Reproduced with permission from Abhyankar et al. 17 This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data. The figure includes minor additions and formatting changes to the original figure.

Reproduced with permission from Abhyankar et al. 17 This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data. The figure includes minor additions and formatting changes to the original figure.

In remote and rural contexts, where there were significant staff shortages and service was inaccessible in many areas, there was heightened awareness of the service needs among the implementers, at least at the implementation stage, which led to an increased readiness to adopt a community model and train nursing staff in the community. However, in other contexts with adequate specialist capacity where there was less of a need for such an intervention, the implementers were less convinced of the requirement for the PROPEL intervention. This triggered feelings of threat to their role and protective feelings about their role among the specialist or higher-grade staff. These staff were also concerned that training lower-grade staff may not address women’s complex care needs adequately and would have a negative effect on standard of care. These feelings resulted in preventing the adoption of a community model, resistance to train lower-grade nurses and significant delays in implementation.

Summary of findings from the realist evaluation

In summary, an outcome evaluation of the PROPEL intervention, showed that women’s outcomes improved regardless of the service model and skill mix through which PFMT was delivered. The realist evaluation provided more insights into the extent to which, how and why the intervention worked (or not) in diverse contexts that differed in their geographical locations, service organisation, type of service model used, skill mix trained and availability of support structures and resources. Findings from the realist evaluation suggest that implementation of PFMT that is delivered by different clinicians in the context of the UK NHS would probably be successful if:

• the training of PFMT delivery is adequately tailored to the needs of the different skill mix

• there is awareness about prolapse and PFMT among women, GPs and other HCPs

• the referral systems are open, flexible and well co-ordinated

• there is sufficient buy-in at organisational levels and strong managerial support in terms of resources, training, time, autonomy and staffing

• the wider team supports the delivery of PFMT through different staff skill mixes

• the tendencies to protect professional roles and boundaries (and thereby prevent change) are acknowledged, addressed and managed effectively and balanced with the population needs.

Primary research questions

• Does the intervention, as implemented, improve POP symptoms in women at the 6-month follow-up? (research question 2).

• Does the effect of the intervention on POP symptoms at the 6-month follow-up differ when the intervention is delivered by specialist physiotherapists compared with when it is delivered by other HCPs? (research question 3).

Secondary research aims

• Applying the primary research questions to POP symptoms at the 12-month follow-up (research question 2).

• Applying the primary research questions to secondary outcomes, including other pelvic floor-related outcomes (severity of prolapse and incontinence) and service-related outcomes (number of health service contacts) (research question 2).

Study design

The study was an observational prospective cohort study that compared PROMs pre and post intervention. The PROMs study was concerned with answering the specific aims to explore the differences in clinical outcomes with the different skill mixes and to understand the variations of PFMT protocol. Furthermore, data from the PROMs study were used to establish the costs and benefits of each model.

Participant flow

A total of 102 women were recruited to the study. Baseline questionnaires were available from 91 women, and 71 women completed the 6-month follow-up questionnaire, for whom 68 questionnaires were matched to baseline questionnaires. Seventy-two participants with baseline or 6-month follow-up questionnaires also had a clinical assessment form. Eight participants had a clinical assessment form only, and 59 participants had baseline and 6-month follow-up questionnaires plus a clinical assessment form. Participant flow is summarised in Figure 8.

FIGURE 8.

The PROMs study participant flow.

Results

Primary outcome results (Pelvic Organ Prolapse Symptom Score)

The distribution of scores on the primary outcome (POP-SS) was explored visually in a histogram (Figure 9). Before intervention, that is, at baseline, there were 88 POP-SSs available. The distribution of these scores is shown in Figure 9. Similarly, the POP-SSs available at the 6-month follow-up were examined to evaluate the distribution of the scores. At 6 months post intervention, 68 POP-SSs were available. The distribution of these scores is shown in Figure 10.

FIGURE 9.

Histogram of the primary outcome, POP-SS: all baseline questionnaires.

FIGURE 10.

Histogram of the primary outcome, POP-SS: full sample.

Results for women who had paired data available at baseline and 6-month follow-up

The primary comparison is between the scores at baseline and the scores at 6-month follow-up among the women for whom both sets of observations are available. There were 65 women with both baseline and 6-month POP-SSs. Initially, we examined the distribution of the scores at baseline and 6-month follow-up visually in frequency histograms (Figures 11 and 12, respectively).

FIGURE 11.

Histogram of the primary outcome, POP-SS paired data: baseline.

FIGURE 12.

Histogram of the primary outcome, POP-SS paired data: 6 months.

The primary quantitative research question regarding whether or not there was a significant change in the POP-SSs between baseline and 6-month follow-up (research question 2) was evaluated with a paired t-test (Table 8). A Wilcoxon matched-pairs signed-rank test was also significant (p < 0.001). Prolapse symptom severity was significantly lower at the 6-month follow-up than at baseline. The estimated difference in POP-SS between baseline and 6-month follow-up is 3.20 points [95% confidence interval (CI) 2.00 to 4.40 points].

TABLE 8 - Paired t-test of POP-SSs between baseline and 6 months

CI, confidence interval; df, degrees of freedom.

Variable	n	Mean	Standard error	SD	95% CI	Paired means t-test
POP-SS (points) at baseline	65	10.18	0.70	5.63	8.79 to 11.58	t = 5.3142, df = 64; p < 0.001
POP-SS (points) at 6 months	65	6.98	0.65	5.23	5.69 to 8.28
Difference (points)	65	3.20	0.60	4.85	2.00 to 4.40

Estimating the effects of the intervention from the mixed-effects model

The effect of the intervention was also evaluated with linear mixed models (Table 9). Linear mixed models allow computation of appropriate standard errors for correlated, repeated-measures data and estimate subject-specific conditional effects with adjustment for covariates. 27 The first model had POP-SS as a dependent variable and time as an independent variable (baseline vs. 6-month follow-up). That is, the coefficient for ‘time (baseline vs. 6 months)’ in the model estimates the conditional difference in means on the POP-SS between baseline and 6-month follow-up. A random effect of participant was included in the model to control for repeated measures. The second model also included parity and age, as assessed at baseline, and BMI, as measured at baseline and 6-month follow-up. The third model included ethnic group. A second random effect of site was also tested in the models; however, the effect was not significant and so was dropped and is not included in the results presented. We recognise that the degrees of freedom (df) in the described models are very low. The models are presented for illustration only, as they were part of the planned analysis. The estimated effect is very similar to the paired t-test, which estimated the treatment effect as a 3.20-point improvement in POP-SS, and here the fully adjusted model estimates the effect as a 3.48-point improvement.

TABLE 9 - Effect of intervention on the POP-SSs (points) as indicated by the effect of time (before vs. after) in mixed-effects models

Variable	Coefficient (95% CI)
	Model 1	Model 2	Model 3
Time (effect of the intervention)	–3.34 (–4.46 to –2.22)	–3.50 (–4.96 to –2.06)	–3.48 (–4.93 to –2.04)
Parity (baseline)		1.03 (–0.04 to 2.10)	0.88 (–0.23 to 2.00)
BMI (change from baseline)		0.11 (–0.15 to 0.38)	0.11 (–0.15 to –0.38)
Age (baseline)		–0.01 (–0.10 to 0.12)	0.01 (–0.10 to 0.13)
Ethnic group: white			0
Ethnic group: Asian			6.80 (–3.86 to 17.45)
Ethnic group: missing			1.21 (–2.47 to 4.88)
Random effect of participant estimate	16.34 (10.60 to 25.18)	14.42 (8.07 to 25.76)	13.79 (7.68 to 24.76)
Total (n)	91	56	56

Next, we began to explore the second research question by examining how the baseline POP-SSs varied by site (Table 10). The mean POP-SSs ranged from 9.57 points (SD 5.52 points) in site E to 15.71 points (SD 3.25 points) in site C.

TABLE 10 - Baseline and 6-month follow-up mean POP-SSs (points): comparison by site (complete case)

Site	Total patients (n)	Baseline POP-SS, mean (SD)	6-month follow-up POP-SS, mean (SD)	Mean difference (95% CI)
A	22	10.50 (5.95)	6.55 (4.34)	–3.95 (–6.11 to –1.80)
B	19	9.16 (5.00)	6.53 (5.54)	–2.63 (–5.28 to 0.02)
C	5	16.00 (2.74)	12.2 (3.63)	–3.80 (–6.49 to –1.11)
D	10	9.90 (5.63)	6.40 (4.93)	–3.50 (–7.67 to 0.67)
E	9	8.67 (6.22)	6.78 (6.92)	–1.89 (–4.67 to 0.89)

Table 10 indicates that the effect of the intervention is significant when considering either site A or site C alone only. This analysis, however, does not allow us to compare effects across sites. The comparison by site was conducted in a general linear model. POP-SS at 6 months was the dependent variable. The effect was estimated by an interaction term between ‘site’ and ‘baseline POP-SS’. The interaction term and the main effect of ‘site’ were not statistically significant. That is, we cannot reject the null hypothesis of no difference in effect of the intervention between sites. This result was expected, given that the study was not powered to detect between-site differences.

Comparison of intervention delivery by specialist physiotherapists versus delivery by other health-care professionals

The second quantitative research question (research question 3) concerned whether or not the effect of the intervention varied according to whether it was delivered by specialist physiotherapists or other HCPs. The sample was divided into two groups: women who were seen by specialist physiotherapists and women seen by another HCP. The difference in the change in POP-SS between baseline and 6-month follow-up was compared between the two groups; the results are shown in Table 11.

TABLE 11 - Independent-samples t-test of change in POP-SSs between baseline and 6-month follow-up comparing those seen by specialist physiotherapists with those seen by other HCPs

Group	Total (n)	Mean change in POP-SS (points) from 0 to 6 months	Standard error	SD	95% CI
Seen by specialist physiotherapist	22	–3.95	1.03	4.86	–6.11 to –1.79
Seen by other HCP	43	–2.81	0.74	4.86	–4.31 to –1.32
Combined	65	–3.20	0.60	4.85	–4.40 to –2.00
Difference		1.14	1.27		–1.41 to 3.69

There is no evidence to reject the null hypothesis of there being no difference between the intervention delivered by specialist physiotherapists and the intervention delivered by other HCPs. In the mixed-effect model, the effect of ‘being seen by a specialist’ was not significant before and after adjustment of covariates and random effect of centre.

Stage of prolapse (Pelvic Organ Prolapse Quantification System) at 6-month follow-up

There were 34 POP-Q scores available at baseline and 27 POP-Q scores available at the 6-month follow-up. From these scores, most women were deemed to be at stage II of prolapse. At baseline, eight women (23.5%) were at stage I, 25 women (73.5%) were at stage II and one woman (2.9%) was at stage III of prolapse. At the 6-month follow-up, 12 women (44.4%) were at stage I, 14 women (51.9%) were at stage II and one woman (3.7%) was at stage III of prolapse.

There were 26 women with a POP-Q score at both baseline and 6-month follow-up. A Wilcoxon matched-pairs signed-rank test was used to test the equality of the distributions at baseline and at 6-month follow-up; there was no significant difference between the distributions (p = 0.0956).

Number of visits to a health-care professional

Figure 13 shows the distribution of health-care visits in the 6 months before baseline; there was a mean of 1.76 visits (SD 1.40 visits, range 0–10 visits, n = 85 participants).

FIGURE 13.

Graph of the distribution of health-care visits 6 months before baseline.

The mean number of health-care visits in the 6 months between baseline and 6-month follow-up was 0.92 visits (SD 1.33 visits, range 0–4 visits, n = 65 participants) (Figure 14). The distribution of the count of health-care visits is not normally distributed; therefore, a non-parametric matched-samples test was employed to test the difference in distributions between baseline and the 6-month follow-up.

FIGURE 14.

Number of GP and nurse appointments between baseline and the 6-month follow-up questionnaire.

Table 14 summarises the number of appointments for women with paired data before and after the intervention (n = 59). There was a significant reduction in the number of health-care visits after the start of the intervention (Wilcoxon signed-rank test z = 3.663; p < 0.001).

TABLE 14 - Number of health-care appointments before and after the intervention

IQR, interquartile range.

Appointments by period	Total (n)	Mean (SD)	Median (IQR)	Range
Total appointments in the 6 months prior to baseline	59	1.73 (1.51)	2 (1–2)	0–10
Total appointments between baseline and 6-month follow-up	59	0.96 (1.36)	0 (0–2)	0–4

Association between adherence and prolapse symptoms

Adherence was assessed as the self-reported frequency of performing pelvic floor exercises. At the 6-month follow-up, one woman (1.45%) reported that she did not do pelvic floor exercises, two women (2.90%) reported that they did them a few times per month, one woman (1.45%) did them a few times per week, six women (8.70%) did them once per week, 10 women (14.49%) did them once per day and 49 women (71.01%) reported that they did pelvic floor exercises a few times per day.

Most women reported performing pelvic floor exercises a few times per day. We compared the change in the POP-SS from baseline to 6-month follow-up between women who performed pelvic floor exercises a few times per day and women who performed pelvic floor exercises less frequently (i.e. once per day or less); we found no evidence of a difference between these groups. The mean change in the POP-SS for those performing exercises a few times per day (n = 49) was a reduction of 3.75 points, and the mean change in those performing pelvic floor exercise less frequently (n = 19) was a reduction of 2.05 points. This was not a statistically significant difference, as determined by an independent-samples t-test (t = 1.2642, df = 61; p = 0.2110). The non-parametric correlation between the pelvic floor exercise frequency and the change in POP-SS between baseline and 6-month follow-up is also not significant (Spearman’s rho = –0.135; p = 0.2912).

Methodology

Results

The baseline characteristics of participants included in the analysis are summarised in Tables 16 and 17. The distribution of characteristics is similar to that of the full POPPY trial sample.

During the follow-up period, 15 participants (5%) died (Table 18). The mortality rate was 5% in both groups.

Analysis of time to first treatment

We examined whether or not there was any difference between the intervention group and the control group in the time until treatment during follow-up. That is, we investigated whether or not the intervention was effective in delaying any further prolapse treatment. First, we examined whether or not the overall follow-up periods were similar between the intervention group and the control group. The time in follow-up (i.e. the time at risk and under observation) was found to be comparable between the groups. The mean time in follow-up in the intervention group was 9.36 years (SD 262.5 days) and the mean in the control group was 9.32 years (SD 267.9 days).

The survival function for the intervention group and the control group is shown in Figure 15. The survival function shows how the proportion of women in the two groups who had not received any further treatment declined over time.

FIGURE 15.

Kaplan–Meier survival estimates by intervention status.

A Cox regression was used to test whether or not women in the control group were likely to receive treatment sooner than women in the intervention group. The Cox regression model (Table 22), which was stratified by preference for surgery and included 264 subjects and 125 failures, estimated that there was a significant hazard ratio in favour of the intervention group of 0.65 (95% CI 0.46 to 0.94; p = 0.020). This indicates that the intervention did delay the need for further treatment and that, during follow-up, women in the treatment group had a 35% lower hazard for treatment.

TABLE 22 - Cox regression on time to any treatment

HR, hazard ratio.

Variable	HR (95% CI)	p-value
Intervention	0.65 (0.46 to 0.94)	0.020
Prolapse symptom severity score at baseline (missing imputed at median)	1.06 (1.02 to 1.09)	0.001
Stage of prolapse II (reference category stage I)	1.48 (0.77 to 2.85)	0.245
Stage of prolapse III (reference category stage I)	2.40 (1.13 to 5.13)	0.023
Age band 45–64 years (reference category < 45 years)	1.66 (0.88 to 3.10)	0.115
Age band 65–79 years (reference category < 45 years)	2.58 (1.29 to 5.17)	0.008
BMI	0.95 (0.91 to 0.99)	0.014

When the outcome was restricted to time to surgery (Table 23), the number of failures was 91 and the hazard ratio was 0.71 (95% CI 0.47 to 1.08; p = 0.112).

TABLE 23 - Cox regression on time to inpatient surgery in days, stratified by preference for surgery

HR, hazard ratio.

Variable	HR (95% CI)	p-value
Intervention	0.71 (0.47 to 1.08)	0.112
Prolapse symptom severity score at baseline (missing imputed at median)	1.08 (1.04 to 1.12)	< 0.001
Stage of prolapse II (reference category stage I)	1.47 (0.67 to 3.22)	0.335
Stage of prolapse III (reference category stage I)	2.65 (1.09 to 6.44)	0.031
Age band 45–64 years (reference category < 45 years)	1.56 (0.78 to 3.15)	0.211
Age band 65–79 years (reference category < 45 years)	2.22 (1.00 to 4.91)	0.049
BMI	0.95 (0.91 to 1.00)	0.036

Summary of findings

The data linkage study provides evidence that PFMT reduces the overall long-term risk of requiring hospital treatment for pelvic floor disorders, over a post-intervention period of > 10 years. There is also evidence that PFMT extends the time for which hospital treatment is not required.

Introduction

The economic evaluation for the PROPEL intervention was concerned with the associated costs and outcomes of different service delivery models for delivering PFMT. In addition, an economic assessment of the long-term costs that were associated with accessing further pelvic prolapse treatment over time was conducted for the original POPPY trial participants, who were resident in Scotland. The multicentre randomised POPPY trial had found that individualised PFMT was effective in reducing the symptoms of prolapse, provided improvements in quality of life and showed potential to be cost-effective in women. 6

Methods

Results

The POPPY trial longitudinal outcomes

Long-term costs that were associated with accessing further pelvic prolapse treatment were estimated using follow-up data for 293 women involved in the POPPY trial. NHS Scotland Scottish Morbidity Record (SMR) data for F2 Gynaecology and CB Urology health-care utilisation was accessed for the period 2007–18. For the SMR-00 data set, no F2 and CB episodes had a main condition recorded; therefore, it was not possible to filter in or out only POP-relevant episodes on that basis. The main operation codes were used and were predefined as P18, P22–P24, P26, Q07–Q08, Q54, M51–M58, A70 and H57. It should be noted that operation code is a non-mandatory code field for SMR-00. No relevant operation code for POP was recorded in CB episode data for either SMR-00 or SMR-01. No further analysis was carried out for CB specialty. Resource use costs for each episode’s main operation were valued in prices using the UK NHS Reference Costs 2017/18. 34

Table 29 presents the resource use costs for available data regarding F2 gynaecology outpatient and inpatient episodes of care. Owing to issues of confidentiality, SMR-01 data cannot be broken down further. Information is presented in aggregate to meet the data protection requirements of the NHS National Services Scotland’s Statistical Disclosure Control Protocol Version 3.0. 39 Resource use costs were estimated for 111 F2 outpatient episodes of care, using the main operation code of interest. Resource use costs were estimated for 109 F2 inpatient episodes, using the main operation code of interest recorded. Inpatient costs (2017–18 prices), excluding excess bed-days, were higher for the control group (£172,458.70; 53%) than for the intervention group (£154,544.02; 47%). Including excess bed-days, the difference in costs between the groups was £21,480.26, retaining a 53% and 47% split. Longitudinal follow-up data about health-care resource use for prolapse treatment by the POPPY trial participants indicated that the overall total costs for the control group were higher than for the intervention group (53% and 47%, respectively). The difference is potentially an artefact of the adjustment of 2017–18 prices, using the UK deflator series. 35

TABLE 29 - The POPPY trial participants’ health-care utilisation costs for procedures of interest (F2 gynaecology), from 2007 to 2018 (GBP 2017–18 prices)

16% of SMR-00 records had procedure codes recorded.

Excess bed-days calculated as the difference from the average number of bed-days of the POPPY trial participants with same main operation code.

Cost sources: NHS Reference Costs 2017/18;34 gross domestic product deflators. 35

SMR record type	Physiotherapist (treatment) costs (£)	Leaflet (control) costs (£)
SMR-00: outpatienta	9112.08	10,200.04
SMR-01: inpatient	154,544.02	172,458.70
SMR-01: inpatient – excess bed-daysb	8723.24	12,288.82
Total	172,379.36	194,947.56

Outpatient (SMR-00)

In total, 5004 episodes of outpatient care was recorded for all participants. A total of 1419 F2 episodes were recorded between 2007 and 2018. No main condition code was recorded. A total of 84% of records had no main operation or other operation code recorded. Of 226 F2 records with a main operation code, 129 F2 records had an operation code of interest recorded as a main or other operation. Excluding ‘did not attend’ episodes, resource use costs were estimated for 111 episodes of care that had a procedure of interest recorded as main operation.

Figure 16 compares all episodes of care between POPPY trial groups, and compares identifiable POP-relevant episodes of care between groups (2007 to 2010 are supressed because of low numbers). The numbers of episodes of health-care treatment shown in Figure 16 indicate a broadly similar pattern of resource use by both groups.

FIGURE 16.

Comparison of the health-care resource utilisation between the POPPY trial control and treatment group participants, gynaecology specialty: all episodes of care vs. POP-relevant coded episodes of care. Note that lines start at different points owing to small numbers within years.

New appointments accounted for 41% of the intervention group and 30% of the control group episodes of care. Table 30 summarises the POP-relevant outpatient resource use costs.

TABLE 30 - Costs for outpatient appointments (GBP 2017–18 prices)

Unit costs adjusted for year using gross domestic product deflators. 35 Costs source: NHS Reference Costs 2017/18. 34

POPPY trial group	SMR-00	Units (n)	Unit cost (£)	Total (£)a
Intervention	P26 procedure (MA23Z, outpatient)	52	183.52	9112.08
Control	P26 procedure (MA23Z, outpatient)	59	183.52	10,200.04

The total estimated outpatient resource use costs were lower for the intervention group than for the control group: £9112.08 and £10,200.04, respectively (2017–18 GBP prices). Given that it was not possible to use a condition code to filter records in or out, and that 84% of records did not have any procedure codes recorded, it was not possible to determine how many data were missing. It is not possible to conclude whether or not there is a difference in the absolute number or timing of use of outpatient health-care resources between groups since the POPPY trial.

Surgery (SMR-01)

In total, 1153 episodes of inpatient care had been recorded for all participants, including 166 F2 episodes. Of 166 episodes, 137 episodes recorded a main condition code and 126 a main operation code. A total of 109 F2 episodes recorded at least one procedure code of interest. Table 31 indicates that, overall, the control group received more procedures of interest than the intervention group.

TABLE 31 - Procedures of interest (F2 gynaecology), number per POPPY trial group

Number of procedures (tally count of all operation codes of interest recorded). Not equivalent to individuals. Some patients will have received more than one procedure of interest during a single health-care episode.

Procedures of interest	Number of proceduresa
	Intervention group	Control group
Colporrhaphy/repair (P22, P23, P24, Q54)	52	67
Hysterectomy (Q07, Q08)	5	5
Incontinence (M53)	< 5	8

Comparison between groups indicates that patterns of admission were broadly similar during the period 2007–18, with a total of 50 health-care episodes for the intervention group and 59 for the control group. The total number of excess bed-days (based on the average stay of the POPPY trial participants by main operation code) was 20 for the intervention group and 28 for the control group. The timing of more expensive procedures, cost-adjusted to 2017–18 prices using the UK deflator series,35 drives some of the difference in total cost. For the physiotherapy group, 35% of the main operation codes were for expensive operations (> £4000), most of which had occurred in earlier years and, therefore, were deflated for cost-adjustment. In contrast, 21% of the main operation codes for the control group were for expensive operations carried out in later years and, therefore, costs were subject to less adjustment.

Discussion

The economic analysis assessed the costs of each clinical site’s delivery model of PFMT. The total cost of the specialist physiotherapist delivery model was calculated at £434.13 per woman attending for PFMT during the PROPEL intervention (2017/18 GBP). This includes the cost of consultant referral, specialist physiotherapist triage, the initial assessment appointment and treatment sessions delivered by specialist physiotherapists (AfC band 7 or band 6), and proportional overheads. At the other PROPEL intervention sites, delivery varied in terms of referral pathway, the amount of clinical contact time to deliver PFMT that women could access and the mix of clinical staff delivering PFMT. Services that were delivered using higher-band physiotherapists only were more costly than services delivered by other staff mixes.

The headline costs indicated that, generally, services that were delivered using higher-band physiotherapists only were more costly. In common with other clinical studies, the introduction of the PROPEL intervention was accompanied in some clinical sites by an increase in staff for the duration of the project. Changes in resources may provide the opportunity to reduce waiting lists, as was the case for site D during the PROPEL intervention, or to enable access to treatment for more patients. Future research into sustainable changes to resources is advised.

Changes in utility, measured using the EQ-5D-5L (secondary outcome), indicated that the PROPEL intervention participants had a high self-reported level of HRQoL, with 31% of participants reporting maximum utility at 6 months (37% at 12 months). This was also reflected in the participants’ responses to the EQ-VAS. Utility improved by 0.019, on average, over 6 months: 0.035 for specialist physiotherapists and 0.012 for other clinicians (with no significant difference between these groups). The primary outcome indicated that statistically significant differences in POP-SSs were found in sites A and C, comparing baseline with 6 months. Comparison between sites was not possible. The economic outcomes indicated that no statistically significant difference was found between baseline and 6 months within any site. There was a greater magnitude of improvement in EQ-5D-5L scores for women who received PFMT treatment from specialist physiotherapists than for women who received the treatment from other HCPs, but this was not a statistically significant difference.

During the POPPY trial, the mean cost of PFMT was £306.86 (95% CI £250.74 to £368.29) for one woman taking part (2009 prices) [mean cost £349.45 (95% CI £285.54 to £419.41) for 2018 prices]. 6 Longer-term benefits were not considered as part of the study. Longitudinal follow-up data of the POPPY trial participants did not indicate a difference in the use of health-care resources between the groups. Interpretation of outpatient health-care resource use between the groups is limited by the absence of coding reference information. On the basis of the available follow-up data obtained from the administrative data sets, there is no observed difference in resource use between the groups at 11 years post PFMT.

Introduction and methods

The PROPEL intervention aimed to deliver an evidence base for alternative models of care to increase the capacity within services for delivering the evidence-based PFMT intervention for POP (as supported by the POPPY trial). To support further dissemination of this evidence (from the POPPY trial and the PROPEL intervention), we included dissemination activities that would also extend opportunities for further data collection on potential barriers and facilitators beyond the PROPEL intervention sites to a wide range of interested stakeholders across the UK.

Two ‘dissemination for implementation’ workshops were held: one in London and one in Glasgow, with 72 participants and 48 participants, respectively. The participants constituted specialist physiotherapists, physiotherapists, consultant urogynaecologists, women’s health fitness instructors and nurses. The majority of participants rated the workshop as meeting their objectives extremely well/very well, with a small minority (n = 6 across two events) rating the workshop as average in meeting their objectives. Most of the participants had attended the workshop to ‘identify how services may be changed to reach more women’ and to ‘find out about the results of PROPEL’. Following the workshop, the majority thought that their service could initiate some change and could identify actions to take forward.

Workshops consisted of sharing the preliminary findings of the PROPEL intervention but also included an ‘action-planning’ component for service managers and those responsible for the delivery of non-surgical interventions for prolapse, which allowed them to start planning a local service redesign that could deliver PFMT to larger numbers of women. These events also acted as ‘data gathering’ opportunities: gathering further data on the potential opposition and barriers to implementation, and how services can make sense of the study findings and apply them to their own services. We encouraged all participants to leave with an action plan.

The dissemination and implementation workshops were delivered according to a theoretically driven structure that detailed problems, solutions (mechanisms for action) and actions. We have used this theory-driven workshop structure previously.

Individual and group perceived barriers to and facilitators of delivering PFMT and delivery using non-specialist roles were mainly recorded using sticky notes posted on group boards, alongside group feedback of key points at several intervals during the day-long workshops. The sticky notes were gathered and collated at the event by the research team and were collectively reported back to the event audience to reflect the broad consensus on the barriers to and facilitators of delivering PFMT. These sticky notes were retained and their key themes were confirmed in a secondary coding exercise by ED, KB and MM, with specific attention given to understanding the levels of opposition and action, the discussion of barriers and whether or not the study findings influenced the development of solutions to these barriers, what other study findings are likely to influence adoption and implementation (e.g. financial data) and which models of delivery seem more favourable for wider adoption and implementation.

It is not surprising that most of the barriers, often representing the absence of something, were also noted as facilitators when the same thing was present or available. The same themes were evident across both workshops; therefore, the data from both events were combined. The main themes are presented and their content is briefly described in the following sections.

Workshop findings

Conclusion

These workshops have added to the knowledge that was generated in the PROPEL intervention, confirming and identifying additional barriers to and facilitators of introducing new models of delivery for PFMT for increasing the capacity and availability of PFMT for women with prolapse. The following activities could follow on from this study to begin to address some of the concerns and raise the potential for change on a wider scale across the NHS:

• Develop awareness-raising training for GPs and information for urogynaecologists about the benefits of PFMT as a first-line treatment.

• Develop new guidelines for practice nurses, midwives and continence services to capitalise on existing opportunities for early identification and intervention, for example by advising women about PFMT during routine smear tests and perinatal appointments.

• Develop public health campaigns for women of all ages (including teenagers), including, for example, a digital versatile disc (DVD) or animation to encourage understanding of pelvic health and exercises.

• Engage with gynaecological service leads to reconfigure care pathways that include PFMT as a first-line treatment option.

• Undertake work to improve care pathways for women with prolapse, including self-referral pathways for PFMT.

Implementation and outcome studies

The models of delivery included a specialist WHP-delivered service, which was used as a gold-standard comparison for other models. It was also seen as important to understand the delivery of such specialist-delivered services outside the trial conditions, in the real world of the NHS. All other services across four other sites implemented various models, with musculoskeletal physiotherapists, other physiotherapists and nurses delivering PFMT. All services except one operated a model whereby women were triaged by a specialist or consultant and referred, as appropriate, for PFMT delivery by the newly trained staff groups.

A realist evaluation was conducted alongside a robust PROMs study to understand ‘what works’, taking into account the impact that the local context has in influencing what works. The realist evaluation was based on multiple case studies of the implementation and outcomes of PFMT delivery in three contrasting NHS settings, sites A, B and C (see Chapter 4) (with the PROMs study involving five sites: sites A–E). The realist evaluation methods allowed for engagement of local stakeholders from the outset, which permitted a detailed exploration of how local sites made decisions on how to deliver PFMT (e.g. using different skills mixes) and how these were implemented. This enabled an understanding of what helped or hindered service delivery, of how the services integrated or affected other services or HCPs and of any potential influences on outcomes for women. The focus of ‘what works and for whom’ included the views of service managers, consultants/referral agents and those delivering PFMT, as well as the views of the women in receipt of PFMT services. However, given the overall positive outcomes for women (all models were effective for women), the presentation of findings focused more on the different models and HCPs delivering PFMT and what worked for ‘implementation’.

When there is a perceived need for services (as a result of a lack of an existing service), this is a driver for change. The service with the most established model of specialist physiotherapist-delivered PFMT was the service that was the most resistant to trying alternative models, regardless of need (e.g. staff absences and increasing waiting times for women were reported during this study). The lack of available services in rural areas was a driver for training many different staff types and initiating ‘workarounds’ to enable triage and referrals to the new services.

Staff who were newly trained within the PROPEL intervention and who had some level of prior knowledge of women’s health issues and of the physiology of the pelvic region were more comfortable with this new role and were more likely to feel confident following the 1-day training.

The following are required to successfully implement PFMT for prolapse:

• The provision of adequate training that is tailored to needs of different staff skill mixes.

• An increased awareness of PFMT among GPs and other HCPs.

• Well-co-ordinated and flexible referral systems.

• Wider (multidisciplinary) team support and/or buy-in.

• Organisational and managerial support with effective leadership.

• A receptive setting (some community nursing and acute wards may not be appropriate).

For the PROMs study, five NHS sites were recruited to deliver PFMT using a range of delivery models. Site A constituted an exclusively WHP-specialist service model that then acted as a comparator for models involving delivery by ‘other’ staff groups (i.e. nurses, physiotherapists), as developed by sites B–E.

A total of 102 women were recruited to the PROMs study from across the five sites and a total of 65 women provided both baseline and 6-month POP-SSs (the primary outcome measure). The mean POP-SS at 6-month follow-up was significantly lower than that at baseline, indicating that prolapse symptom severity had reduced. The mean difference between baseline and 6 months represented a clinically meaningful difference. A significant improvement was still observed when comparing baseline with 12-month POP-SSs (although numbers were low).

There was no statistically significant difference in the improvement of symptoms obtained from PFMT delivery by specialist physiotherapists (mean change in POP-SS = –3.95) compared with PFMT delivery by other HCPs (mean change in POP-SS –2.81 points) (mean difference 1.14 points, t = 0.89, df = 63; p = 0.37).

Additional (secondary) outcomes showed that there was a significant reduction in the number of health-care visits after the start of the intervention (mean of 1.73 appointments in the 6 months prior to baseline compared with 0.96 appointments between baseline and 6-month follow-up, Wilcoxon signed-rank test z = 3.663; p < 0.001).

The economic evaluation found that the key difference between models in determining costs was the staff grade (NHS band) and the overall length of time practitioners spent with the women (hours). The baseline HRQoL score was already high in this population, which may have affected our ability to improve this. There were no significant differences in HRQoL between women seen by specialist physiotherapists and women seen by other practitioners. The average EQ-5D-5L scores did not indicate a deterioration in HRQoL for any model of delivery. The limited participant numbers precluded assessment of cost-effectiveness for the models of delivery.

The headline costs indicated that, generally, services delivered using higher-band physiotherapists only were more costly. In common with other clinical studies, the introduction of the PROPEL intervention was accompanied in some clinical sites by an increase in staff for the duration of the project. Changes in resources may provide the opportunity to reduce waiting lists, as was the case for site D during the PROPEL intervention, or to enable access to treatment for more patients. Future research into sustainable changes to resources is advised.

In an analysis of the long-term follow-up data of 293 of the original POPPY trial participants (both intervention and control participants), the main outcome measure was any related secondary care treatment during follow-up using routinely collected individually linked hospital episode data. This showed (using a mixed-effects logistic regression on any treatment during follow-up with random effect of centre) that the effect of the intervention (PFMT) is statistically significant and is associated with a reduction in the log odds of any treatment during follow-up. People in the intervention group had a lower hazard for any treatment.

Longitudinal follow-up data of the POPPY trial participants did not indicate a difference in the use of health-care resources between the groups. Interpretation of outpatient health-care resource use between groups is limited by the absence of coding reference information. On the basis of the available follow-up data obtained from the administrative data sets, there is no observed difference in resource use between the groups at 11 years post PFMT.

The dissemination and implementation workshops were attended by 120 participants (72 in London and 48 in Glasgow) and identified key barriers (and facilitators) in line with those observed in the realist evaluation. The presentation of the PROPEL intervention findings allowed for the discussion of the possibility of different ways to deliver PFMT. This ‘evidence base’ helped the majority of participants to shift into thinking that ‘change’ was possible and to begin developing action plans to discuss with their local services.

Limitations of the study

This study allowed for local services in three areas (one in England and two in Scotland) to determine how they would implement/deliver PFMT. It was originally anticipated that all sites might implement PFMT delivered by a mix of specialists and other HCPs, and that there would be more higher-level co-ordination of referral pathways to fit seamlessly with referrals (from primary care), triage and treatment with PFMT. The reluctance of one site to train non-specialist physiotherapists meant that we had to find other sites that were, or were willing to start, delivering alternative models. However, this did then provide a strong specialist service against which to evaluate the outcomes compared with other staff delivery.

The negotiation of implementation of different models in other sites and the availability of staff for training incurred delays in implementation, which inevitably affected the recruitment of participants to the outcome study and, therefore, long-term follow-up of women beyond 6 months. The ‘failure’ to pursue some proposed models meant that we needed to re-think how PFMT could be delivered in the face of these barriers; this, in turn, has contributed to our understanding of such barriers. This study has also been conducted in the everyday world of the NHS, where staff absences and shortages are common. The fact that staff could be identified and trained, and that PFMT was implemented with successful outcomes, bodes well for other services being able to attempt similar changes.

The number of women recruited to the PROMs study was lower than anticipated (target: 120 women) and the recruitment rates varied across sites, which hampered cross-site comparisons. The slow rate of recruitment also affected our ability to continue with 12-month follow-up for a large proportion of the women. However, there were sufficient women recruited to enable the key research questions to be answered in relation to the effectiveness of PFMT in improving prolapse symptoms and in determining whether or not the use of non-specialist physiotherapists would result in significantly reduced beneficial outcomes for women.

Logistically, it was not possible to make widespread use of the POP-Q in this study. The use of this measure requires a high level of skill in conducting a physical examination and may also require additional visits by women in studies using the POP-Q. The use of non-specialist physiotherapists working in different settings and geographical locations meant that POP-Q assessments required the additional availability of someone trained in its use to travel to different locations, often to assess one woman. This provision was not possible for most sites.

We have endeavoured to follow up women in the longer term using routinely available and linked health data. Whereas inpatient records are robust indicators of treatment outcomes, outpatient records are more variable in their availability and their accuracy of recording of treatments and procedures. The majority of the POPPY trial participants were resident in Scotland, which made accessing and linking hospital-based activity possible, but did not make accessing and linking primary care data possible. However, the number of original POPPY trial participants (n = 293) available for long-term follow-up was sufficient to enable a robust analysis.

Implications for policy

An NHS review has revealed that there are currently hundreds of women now undergoing surgery to have mesh implants removed. Although a repair using mesh is not the only surgical procedure for prolapse, it has constituted a common procedure in favour of more traditional procedures. Most mesh devices were launched without clinical trials to support their effectiveness, meaning that complication rates were never established prospectively. The National Institute for Health and Care Excellence banned the use of vaginal mesh to treat prolapse in December 2017. Recognising the potential scale of the problem, and in response to lobbying, the UK government announced a review into how the NHS addresses concerns about medical treatments, including vaginal mesh devices. This review is led by Baroness Cumberlege. 40 Following presentation of the PROPEL intervention findings at our dissemination and implementation workshops, the Chartered Society of Physiotherapists approached the PROPEL intervention team to input its findings to this review. These were submitted in February 2019. The PROPEL study was discussed at an oral hearing. 41 Informed by the PROPEL study, the Chartered Society of Physiotherapists made the following recommendations to the review:

• Women who have had a prolapse or stress urinary incontinence should be referred for PFMT as a first line of treatment.

• Increase the specialist physiotherapy workforce and train non-specialists to provide PFMT.

Similarly, the Scottish government has initiated its own response to the mesh concerns. A Scottish independent review42 of the use, safety and efficacy of transvaginal mesh implants in the treatment of stress urinary incontinence and POP in women was initiated in response to growing public concern about the number of women experiencing serious complications. At this time, the Cabinet Secretary for Health and Wellbeing asked NHS Scotland to consider suspension of transvaginal mesh procedures pending the outcome of this review. 42 Concluding this review in 2017, its chairperson expressed the hope that their report would go some way to ensuring that the voices of patients continue to be valued, and that they could be assured that future NHS prolapse interventions would be evidence based and audited in order to deliver the best possible outcome with minimum adverse effect. 42

The PROPEL intervention began by listening to women in our three case study sites and gave this information to the local teams to consider in their development and delivery of services. We have ensured that the implementation of PFMT using different models of delivery was subject to a robust outcomes evaluation (using the same primary and secondary outcomes as the original POPPY trial). This has provided a strong evidence base, on which other NHS sites can build, to deliver an evidence-based intervention for women with prolapse that does indeed minimise the potential for adverse effects.

The independent review (Scotland)42 found that current research studies on safety and effectiveness have not provided sufficient evidence on the long-term impact of mesh surgery. They recommended that the lack of long-term follow-up and related outcome data, including information on quality of life and activities of daily living, should be addressed.

The POPPY trial and the PROPEL intervention have both focused on robust outcomes data and the primary outcome has been based on a measure that captures women’s views of their prolapse symptoms and quality of life. This study has also conducted more long-term follow-up of the original POPPY trial participants. We believe that this adds to the strength of evidence for PFMT.

Following an invitation to the Chief Medical Officer for Scotland to our dissemination and implementation workshop in Glasgow, the Scottish government responded with an invitation to present our findings at a meeting that took place in February 2019. These findings were shared with the Consultant in Public Health Medicine, the Government’s Oversight Group and the Senior Medical Officer. The level of interest in our findings at national policy level indicates the relevance of our work for both policy and practice.

Conclusions

It is possible to train different staff types/skill mixes to effectively deliver PFMT to women. Women’s self-reported outcomes significantly improved across all service models delivering PFMT. The PROPEL intervention has demonstrated how different models and service contexts can affect implementation, but, if these are addressed in advance, then there will be a greater likelihood of successful integration with existing health-care pathways and greater opportunities for ensuring that (non-specialist) staff delivering PFMT are supported to deliver PFMT with appropriately triaged patients and appropriate clinical supervision. Services delivered by higher-band physiotherapists only were more costly than services delivered by other staff mixes. No statistically significant difference in the improvement of HRQoL between baseline and 6 months within any site, or from baseline to 12 months between participants treated by specialist physiotherapists and participants treated by other clinical staff, was found. Small participant numbers prevented an assessment of cost-effectiveness.

The realist evaluation combined with the robust outcomes data confirms that PFMT can be successfully delivered using a range of staff/skill mixes and in different NHS settings, and that outcomes are not compromised by different delivery models. This study provides examples of delivery of PFMT (beyond delivery by specialist physiotherapists) by clinicians (i.e. nurses, other physiotherapists) who have an interest in women’s health.

Implications for practice

The evidence supports training a broader range of HCPs with an interest in women’s health and/or with a knowledge of physiology/body muscles to deliver PFMT to women.

Delivery of PFMT by other clinicians was supported by specialist physiotherapists undertaking triage of women to determine their suitability for PFMT. The role of the specialist physiotherapist could then be enhanced, to provide education and support to other HCPs to enable them to safely deliver PFMT to women while managing more complex cases of POP themselves.

Increasing the use of and referral to PFMT as a first-line treatment was facilitated by improved multidisciplinary team working across urogynaecology services and improved communication with primary care. A review of existing pathways to PFMT could identify areas for improvement.

Primary care referrals to PFMT as a first-line treatment for POP would probably increase with more awareness-raising (of POP and PFMT) and education for GPs and other primary HCPs.

The following activities could follow on from this study to begin to address some of the concerns above and raise the potential for change on a wider scale across the NHS:

• Develop awareness-raising training for GPs and information for urologists and urogynaecologists about the benefits of PFMT as a first-line treatment.

• Develop new information for practice nurses, midwives and continence services to capitalise on existing opportunities for early identification and intervention, for example by advising women about PFMT during routine smear tests and perinatal appointments.

• Develop public health campaigns for women of all ages (including teenagers), including, for example, a DVD or animation, to encourage understanding of pelvic health and exercises.

• Engage with gynaecological service leads to reconfigure care pathways that include PFMT as a first-line option.

• Undertake work to improve care pathways for women with prolapse, including self-referral pathways for PFMT.

Research recommendations

This evaluation provides a toolkit of outcomes that could be used in routine service delivery, especially the POP-SS outcome measure, which is a relatively easy-to-use, self-complete measure that has now been widely evaluated. Routine evaluation of outcomes could provide more widespread evidence for PFMT delivery by HCPs.

Logistically, it was not possible to make widespread use of the POP-Q in this study. A study comparing the ability of the self-complete POP-SS tool to correlate with the (physical) changes in POP severity (as measured by a clinical assessment using the POP-Q) could help further studies to determine whether the POP-SS is sufficient or we need to input further resources in pelvic health research to ensure that there is more robust completion of the POP-Q examination.

The follow-up study included inpatient and outpatient attendances for a large proportion of original POPPY trial participants. Further studies might explore the information available in primary care and whether or not this has a bearing on assessing the impact of PFMT in the longer term.

The ‘activities’ outlined in Implications for practice would benefit from being grounded in ‘intervention development’ research, which includes input from HCPs (across sectors) and women with POP to ensure that the content of interventions and the formats of delivery are optimised for maximum impact and reach. This would include ‘implementation’ research associated with intervention delivery.

Contributions of authors

Margaret Maxwell (https://orcid.org/0000-0003-3318-9500) (Professor of Health Services and Mental Health Research, Director of the Nursing, Midwifery and Allied Health Professions Research Unit) is lead author of the report and, as such, contributed substantially to writing the report (specifically all abstracts and summaries, and Chapters 1, 8 and 9), as well as editing contributions to all other chapters. She was PI of the study.

Karen Berry (https://orcid.org/0000-0001-6795-4149) (Project Manager and Research Fellow) was the lead researcher who led all of the data collection activity and contributed to its writing (specifically Chapters 2 and 3, and all of the appendices) and was editorial lead for the entire report.

Sarah Wane (https://orcid.org/0000-0002-1766-882X) (Research Associate and Lead Research Fellow during maternity cover for KB) contributed to all data collection activity and, specifically, the qualitative data collection and analysis. She contributed to editing the final report, specifically Chapters 2 and 3.

Suzanne Hagen (https://orcid.org/0000-0002-9741-9160) (Professor of Health Services Research, Deputy Director of the Nursing, Midwifery and Allied Health Professions Research Unit) was the PI of the original POPPY trial and contributed substantially to the PROMs study methods and the long-term follow-up study using data linkage to routine NHS data. She edited the final report, with specific attention to the Scientific summary and Chapters 8 and 9 and the intervention description in Appendix 1.

Doreen McClurg (https://orcid.org/0000-0002-2872-1702) (Professor of Pelvic Floor Physiotherapy) led the adaptation and implementation of training in PFMT for non-specialists. She contributed to editing the final report with specific contributions to Chapters 8 and 9 and sections related to PFMT training and its delivery.

Edward Duncan (https://orcid.org/0000-0002-3400-905X) (Associate Professor of Applied Health Research) contributed expertise to the realist evaluation and the planning, running and data collection and analysis of the dissemination and implementation workshops. He contributed to editing the final report and, specifically, Chapters 8 and 9.

Purva Abhyankar (https://orcid.org/0000-0002-0779-6588) (Lecturer, Faculty of Health and Sports Science) led the realist evaluation and conducted the data analysis for its evaluation. She wrote Chapters 4 and 5 and edited other sections of the final report, with specific attention to reporting of the realist evaluation in the Abstract, Scientific summary, and the Discussion and Conclusions in Chapter 9.

Andrew Elders (https://orcid.org/0000-0003-4172-4702) (Statistician, Nursing, Midwifery and Allied Health Professions Research Unit) was the lead statistician on the PROPEL intervention and contributed to the PROMs study and the POPPY data linkage follow-up study in all aspects, from inception and design to data management, analysis and write-up. He co-wrote Chapters 5 and 6, and edited other sections of the final report, with specific attention to reporting of the outcomes and data linkage studies in the Abstract, Scientific summary, and the Discussion and Conclusions in Chapter 9.

Catherine Best (https://orcid.org/0000-0002-3652-2498) (Lecturer/Statistician, Nursing, Midwifery and Allied Health Professions Research Unit) was a statistician in the PROPEL intervention who contributed to data management and analysis for the PROMs study and the POPPY data linkage follow-up study. She co-wrote Chapters 5 and 6, and edited other sections of the final report, with specific attention to reporting of the outcomes and data linkage studies in the Abstract, Scientific summary, and the Discussion and Conclusions in Chapter 9.

Joyce Wilkinson (https://orcid.org/0000-0002-1008-1882) (Lecturer, Faculty of Health and Sports Science) contributed expertise to the realist evaluation and conducted data analysis for the realist evaluation alongside PA and SW. She contributed to Chapters 4 and 5 and edited other sections of the final report, with specific attention to reporting of the realist evaluation in the Abstract, Scientific summary, and the Discussion and Conclusions in Chapter 9.

Helen Mason (https://orcid.org/0000-0002-9303-2794) (Professor of Health Economics and Deputy Director Yunus Centre for Social Business and Health at Glasgow Caledonian University) led the economic evaluations within the PROPEL intervention in relation to the PROMs study and the POPPY data linkage follow-up study. She co-wrote Chapter 7 and edited other sections of the final report, with specific attention to reporting of the economic evaluations in the Abstract, Scientific summary, and the Discussion and Conclusions in Chapter 9.

Linda Fenocchi (https://orcid.org/0000-0003-2536-8234) (Researcher in Heath Economics at Glasgow Caledonian University) conducted the analysis for the economic evaluation of the models of service delivery and their outcomes and for the economic evaluation of the longer-term follow-up of the POPPY trial participants. She co-wrote Chapter 7 and edited other sections of the final report, with specific attention to reporting of the economic evaluations in the Abstract, Scientific summary, and the Discussion and Conclusions in Chapter 9.

Eileen Calveley (https://orcid.org/0000-0001-8731-2338) (Research Fellow) contributed to the inception of the PROPEL intervention and the overall design and planning of the study and its early set-up stages. She contributed ongoing expertise to data collection and was specifically involved in the planning, organisation and analysis of the dissemination and implementation workshops. She contributed to editing the final report.

Karen Guerrero (https://orcid.org/0000-0001-9591-059X) (Lead Consultant in Urogynaecology, Queen Elizabeth University Hospital) was the lead clinician (Scotland) and contributed to the inception of the PROPEL intervention and the overall design and planning of the study and its implementation in one of the realist evaluation sites. She contributed to editing the final report, specifically its conclusions and recommendations.

Douglas Tincello (https://orcid.org/0000-0002-6385-851X) (Professor of Female Pelvic Medicine and Reconstructive Surgery, University Hospitals of Leicester NHS Trust) was the lead clinician (England) and contributed to the overall planning of the study and its implementation in one of the realist evaluation sites. He specifically contributed to the dissemination and implementation workshops, facilitating data collection at these events and contributing to analysis. He contributed to editing the final report, specifically its conclusions and recommendations.

Publication

Abhyankar P, Wilkinson J, Berry K, Wane S, Uny I, Aitchison P, et al. Implementing pelvic floor muscle training for women with pelvic organ prolapse: a realist evaluation of different delivery models. BMC Health Serv Res 2020;20:910.

Data-sharing statement

All data requests should be submitted to the corresponding author for consideration. Access to anonymised data may be granted following review.

Patient data

This work uses data provided by patients and collected by the NHS as part of their care and support. Using patient data is vital to improve health and care for everyone. There is huge potential to make better use of information from people’s patient records, to understand more about disease, develop new treatments, monitor safety, and plan NHS services. Patient data should be kept safe and secure, to protect everyone’s privacy, and it’s important that there are safeguards to make sure that it is stored and used responsibly. Everyone should be able to find out about how patient data are used. #datasaveslives You can find out more about the background to this citation here: https://understandingpatientdata.org.uk/data-citation.

Disclaimers

This report presents independent research funded by the National Institute for Health 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, NETSCC, the HS&DR 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, NETSCC, the HS&DR programme or the Department of Health and Social Care.

Pelvic floor muscle training

Control

There was no control as this was an implementation study.

Programme

Recommended pre-attendance learning

Professional issues learning tasks:

1. Familiarise yourself with the consent policy used within your organisation.

   Ensure that you are aware of the key components of valid consent, including capacity, the requirements relating to information-sharing, as well as the documentation requirements of your organisation.

   Reflect on the process you use when seeking consent for a vaginal examination and consider whether or not this fulfils your statutory and professional obligations.

2. Familiarise yourself with your organisation’s chaperone policy.

   Consider the process used by you/your organisation in order to establish whether or not a chaperone is required for vaginal examination and reflect on whether or not this fulfils your professional/organisational obligations.

3. Familiarise yourself with the confidentiality requirements of your organisation, including those relating to recently enacted legislation such as the Female Genital Mutilation Act (2003)44 or the Prohibition of Female Genital Mutilation (Scotland) Act (2005). 45

   Ensure that you know how to disclose or escalate any concerns you have relating to patient confidentiality issues.

4. Familiarise yourself with the relevant sections of your organisation’s infection prevention/control policy.

   Ensure that you know how to apply the infection prevention/control policy to ensure safe and effective practice in relation to patients attending as part of the PROPEL intervention.

Additional background learning/knowledge update

The tutor team also recommends that you review the anatomy of the pelvic floor prior to attending the study day. The following YouTube (YouTube, LLC, San Bruno, CA, USA) link is a useful 10-minute tutorial showing the three-dimensional arrangement of the pelvic floor complex: www.youtube.com/watch?v=P3BBAMWm2Eo (accessed 30 January 2020).

Other useful resources relating to POP include:

• www.nhs.uk/conditions/prolapse-of-the-uterus/Pages/Introduction.aspx (accessed 30 January 2020).

• www.rcog.org.uk/en/patients/menopause/pelvic-organ-prolapse/ (accessed 30 January 2020).

Relevant pages of the Royal College of Obstetricians and Gynaecologists, with good links to other sites and surgical options:

• http://pogp.csp.org.uk/publications/pelvic-organ-prolapse-physiotherapy-guide-women (accessed 30 January 2020).

The POP patient information leaflet of the POGP, which will be given to all women in the PROPEL intervention:

• www.yourpelvicfloor.org/conditions/pelvic-organ-prolapse/ (accessed 30 January 2020) – The pelvic organ prolapse patient information leaflet of the International Urogynaecology Association (IUGA), providing a concise overview of the condition.

• www.csp.org.uk/videos/customized-pelvic-organ-prolapse-quantification-exercise-quick-score-entry (accessed 30 January 2020).

Demonstration of a POP-Q measurement; note: this is for information only.

Important information about the practical session on vaginal examination:

This training day includes a practical session on vaginal examination. The aim of the session is for you (as a participant) to practise the examination that you will perform on patients as part of the PROPEL intervention assessment.

Format of the teaching session: one tutor will demonstrate the vaginal examination, using the other tutor as a model. Following this you will be able to practise the examination by peer or tutor examination in small groups. There will be no more than four participants per group.

Tutor examination will involve a tutor acting as the model. Peer examination involves one or more participants acting as the model. Each participant makes their own decision whether or not they act as a model. If you choose to act as a model, you will decide whether this is for one or more examinations and you have the right to opt in or out of the practical sessions at any time.

It is recognised that peer examination can be advantageous. Participants on similar courses who have acted as a model often recommend it as a valuable learning experience, but it is not a requirement. The advantages of acting as a model include you developing a personal insight into the patient experience, including comfort and/or dignity issues. This may allow better communication with your patients as to what they may experience when you examine them in clinical practice. The disadvantages are likely to vary on an individual basis. There may be no disadvantages to you personally; however, such disadvantages may include feelings of discomfort, embarrassment or lack of modesty, as well as relationship boundary issues with fellow learners and/or an exacerbation of legacy of prior sexual abuse and/or trauma. There should be no perception of coercion to participate as a model and you will not be asked to justify your decision.

If you choose to act as a model, you have the right to decline to be examined by any specific course participants owing to embarrassment or other factors. In the case of a participant declining to complete at least one vaginal examination on a model, it will not be possible to complete the skills checklist and the participant will not be able to complete the programme.

You will be provided with a consent form, which must be printed and brought with you to the PROPEL intervention training day. Further information will be given at the training day and you will have the opportunity to observe a demonstration of the examination before you decide on the extent of your involvement in the practical session. Once you have a better understanding of what is required and have had the opportunity to ask any questions you may have about the session, you will be asked to complete the form.

If you would like the opportunity to talk to one of the tutors prior to the training, regarding any aspect of the practical session (including any health issues), then please contact the PROPEL intervention team who will put you in touch with one of the tutors.