Proceedings from the 2nd Annual UK Implementation Science Research Conference, 'Advancing the science of scaling up: Improving efficiency and effectiveness of implementation strategies in healthcare': meeting abstracts

P1 What makes innovations both ‘stick’ and ‘spread’? A multidisciplinary systematic review to understand implementation depth and scale-up of innovations in healthcare Alexandra Ziemann, Yaru Chen, Yiannis Kyratsis, Charitini Stavropoulou, Harry Scarbrough Centre for Healthcare Innovation Research (CHIR), City, University of London, London, United Kingdom; 2 VU Amsterdam, Amsterdam 1081 HV, Netherlands; School of Health Sciences, City, University of London, London, United Kingdom; CASS Business School, City, University of London, London, United Kingdom Implementation Science 2019, 14(Suppl 2):P1

Correspondence: Alexandra Ziemann (alexandra.ziemann@city.ac.uk) Background Often, healthcare innovations are not sustained after adoption and vary in their effectiveness when scaled-up. [1] Diverging research strands focus on either implementation or scale-up/diffusion. These strands typically have different analysis levels, focus on different implementation phases and are conducted in different research fields (e.g., health research (implementation), organisation studies (diffusion)). This systematic review aimed at synthesising evidence on implementation depth and innovation diffusion published in these diverging research fields.

Method
We systematically searched eleven health and organisation/management studies databases for theoretical and empirical studies published in English language, including Medline and Business Source Complete. The search strategy combined terms for innovation with terms for implementation depth or diffusion. Further we handsearched key textbooks, references and citations of included studies and relevant journals in the two fields. Two reviewers screened for inclusion and extracted data. Conflicts were resolved in the team. Data are analysed applying qualitative thematic analysis and summarised in a narrative synthesis.

Results
Preliminary results show that conceptualisations of implementation depth and innovation diffusion vary according to the theoretical perspective. They can be conceptualised along dimensions of scale (extensiveness, completeness, prevalence), scope (degree of fit, fidelity and adaptation), process (incremental and continuous or disruptive and episodic) and outcome (decoupling and ephemeral or sustained change).

Conclusion
The review synthesises concepts from different levels of analysis and diverse literatures. It sheds light on the blind spot in our understanding of how to achieve both, the widespread and sustainable implementation of innovations. Background Dementia is a progressive illness and in the later stages, the person will have difficulty communicating their needs in relation to thirst, hunger, pain or discomfort. In Ireland, over one third of people with dementia reside in Long Term Care settings. Evidence based guidance documents for dementia palliative were developed by the authors with the support of the Irish Hospice Foundation. The aim of this study is to implement the guidance intervention in three Long Term Care (LTC) facilities in Ireland.

Method
The Consolidated Framework for Implementation Science (CFIR) underpinned the 3 phases to promote implementation. Phase 1 consisted of situational analysis, interviews with residents and carers and audit of current practice to identify factors that for the up scaling of the intervention and implementation process. The implementation process is guided by Participatory Action Research (PAR) methodology. These data informed the design of the educational intervention within work based learning groups (WBLG).

Results
Work based learning groups with virtual and physical resources were developed using combined blended approach to learning. Four face-to-face sessions with active and creative learning opportunities were designed based on findings from the situational analysis results in phase 1.

Conclusion
This dynamic approach to implementing the guidance within WBLG enables learners to have greater control over how the learning takes place. Provisions within the work place allow for 'in employment' education to overcome any disconnect associated with traditional education methods. Using this Participatory Action Research process within an implementation framework created learning and embracing of change in practice from within including the identification of strategies to overcome difficulties and maximise benefits of implementation.
Correspondence: Andrew Boilson (andrew.boilson2@mail.dcu.ie) Background The multi-national MIDAS (Meaningful Integration of Data Analytics and Services) project is developing a big data platform to facilitate utilisation of a wide range of health and social care data to enable integration of heterogeneous data sources, providing analytics, forecasting tools and bespoke visualisations of actionable epidemiological data. Method An evaluation framework starting with a logic model and semi-structured interviews using the principles of realist evaluation was developed working with 6 developers, 5 health service managers, 2 allied health professionals and 3 policy makers. Parallel case studies were used to address the requirements of stakeholders at critical time points during the project.

Results
At this early stage of the MIDAS project's development, the logic model represents an accurate representation of the project's key outputs, outcomes and impacts identified through the logic model transcripts coding process. The logic model was developed through a number of iterations of consultation with the MIDAS consortium as an initial process of evaluating, planning and developing the project. As the project progresses, key indicators (outputs, outcomes and impacts) that were not initially included in the logic model may be identified through interviews with end users of the MIDAS platform tools and the technical development teams.

Conclusion
The objective of this study is to narrow the gap between end user requirements of the platform tools and technical developers' expectations of the end user needs, an ongoing process of refining the logic model is required at critical stages of the project's development. To narrow the gap an ongoing process of refining the logic model is required at critical stages of the project's development. Overall, the early stage interviews indicated the logic model is an effective framework for the evaluation of the project.

O4
Scaling up with structure: Impact, feasibility and theoretical reflections on a practice-informed scale up template to improve patient safety Andrew Sibley, Tracy Broom Wessex Academic Health Science Network, Innovation Centre, Southampton, UK. Implementation Science 2019, 14(Suppl 2):O4 Correspondence: Andrew Sibley (andrew.sibley@wessexahsn.net) Background In 2017/18 Wessex Patient Safety Collaborative (PSC) led a project to support local NHS teams expand established patient safety initiatives. The aim was to develop a 'scale up template' (SUT) and test its use in practice. The SUT undertook "deliberate efforts to increase the impact of innovation to benefit more people and to foster policy and program development on a lasting basis." [1] The SUT aimed to support "scale up to at least 60% of a target population that could potentially benefit from the program." [2] The SUT developed a 10stage process to scale up and used pragmatic techniques of engagement, face-to-face, with staff to work through the complexities of scale up. Method A mixed methods investigation, of the impact and feasibility of the SUT, collected quantitative scale up data on each initiative and undertook qualitative focus groups with staff.

Results
Quantitative findings indicated all initiatives reached their target of 60% (55 of 92 possible) scaled up units (hospital wards, care homes etc). Sixteen themes from a synthesised thematic analysis of four focus groups were identified. Themes included: Overwhelmingly positive perceptions of SUT; SUT changed the previous scale up approach; Train the trainer potential. Facilitator themes included: Project manager role vital; First meeting with AHSN PSC lead vital; SUT developed staff awareness of implementation science/scale up; Executive sponsorship vital. A key barrier theme was the systemic underestimation, by staff, of scale up complexity and challenges.

Conclusion
The SUT was perceived as a positive force for change as teams pragmatically worked with complexity to scale up their initiatives. The SUT was a catalyst for changing approach after previous failed attempts to scale up had occurred. Active ingredients of its value were identified. Reflections on how the SUT mapped to implementation theories was undertaken to further develop the SUT.

Background
Implementing and scaling-up innovations is a collective social process [1,2]. However, there is limited analysis about how knowledge exchange operates in practice to support scaling-up [3]. ESCAPE-pain is an evidencebased complex intervention for osteoarthritis currently in >100 sites in England. The study investigated the role of knowledge exchange in scaling-up ESCAPE-pain. Method A qualitative case study approach was used. Forty clinicians and managers participated in 50 interviews across 14 organisations in England. Four NHS providers were selected as case studies, each comprising 5-7 participants. A fifth case study was undertaken of a clinical-academic network' s role in supporting the scale-up of ESCAPE-pain. This used an ethnographic approach comprising multiple interviews with project and senior managers, 8 months of observations and documentation.

Results
Two themes are explored. 1) Creating spaces to exchange formal and tacit knowledge: a. Physical spaceslarge stakeholder knowledge sharing events and partnership working with early adopters were used to share learning about implementing ESCAPE-pain. b. Digital spacesa website and app were developed to expand geographical reach and provide practical knowledge about ESCAPE-pain and its implementation.
2) Types of knowledge created and its purposethe knowledge exchange process developed understanding about: a. ESCAPE-pain and how to manage osteoarthritis better through manualising it, developing a training course, and creating resources (e.g. return on investment tool). b. How to implement ESCAPE-pain to expand reach across different settings, widening target populations, and broadening the workforce delivering it (e.g. clinician/non-clinicians).

Conclusion
The study provides empirical evidence that builds on reframing the idea of the evidence-practice gap as one of a "space" of interaction [4]. The clinical-academic network actively facilitated better connectivity across systems to support the exchange and production of knowledge to drive spread. It was a non-linear and emergent process that utilised formal and tacit knowledge. This has implications for planning system-wide scale-up.
Correspondence: Anja Sommer (a.sommer@maastrichtuniversity.nl) & Cassandra Rehbock (c.rehbock@maastrichtuniversity.nl) Background We assessed the usefulness of systems thinking methods to guide the implementation of interventions reducing demand in a German emergency medical service (EMS) system with the aim of a. analysing intended and unintended consequences of implementing innovations in a complex system; and b. enhancing stakeholder involvement in the implementation process.

Method
Within a stakeholder group (EMS managers, frontline staff, health insurance representatives), causes and consequences of 'rising EMS demand' were discussed and incorporated into a causal loop diagram [1]. Potential interventions to reduce demand were identified by the group and incorporated into the diagram. A pathway analysis was conducted identifying intended or unintended consequences of prioritised interventions in the system [2]. Next steps include assessing time delays, a cross-comparison of the interventions' impact and a stakeholder evaluation of the methodology. Results 36 variables in four thematic groups were identified as interlinked causes/consequences of demand rise (staff, patients, other services, policy level). Eight feedback loops were identified. Four of 13 interventions were included in the path analysis showing no unintended consequences in the system ( Figure 1). Results of the next analysis steps will be presented at the conference.

Conclusion
Applying systems thinking methods helped understanding the complex interactions in the EMS system around the specific problem of demand rise and assess the impact of potential interventions. The methodology allowed for the interactive involvement of the stakeholders from the beginning of the implementation process.

P7
Can a short course increase the skills of early career researchers in initiating and sustaining partnerships with policy? Findings from a pilot Researchers internationally are increasingly motivated to work in partnership with policy agencies on research projects. Such partnerships can be challenging to initiate and sustain, however, and there are few resources available to guide researchers wishing to build skills in this space. The current study presents pilot data from a short course designed to increase the capacity of early career researchers to initiate and sustain research partnerships with policy agencies. The specific aims of the study are to assess the: a) Uptake and acceptability of the course; b) Impact of the course on intentions to work with policy makers and confidence in doing so; and c) Extent to which the impacts (if any) of the course were sustained six months post attendance. Method A pilot evaluation of the first two rounds of this course, developed by the Centre for Informing Policy in Health with Evidence from Research (CIPHER) and the Sax Institute, was conducted. Process measures were collected to assess feasibility, a post-course feedback form was administered to assess acceptability. Participants completed a short questionnaire rating their skills and confidence in working with policy makers, pre, immediately post and six months post course attendance.

Results
The course was feasible to run. Ten participants attended and completed each course (as planned) and feedback data suggested the course was highly acceptable to participants (average ratings of ≥'satisfied' for all items). Significant improvement in self-reported skills, confidence and intentions regarding working with policy agencies were noted amongst participants post course and almost all were maintained at six month follow up.

Conclusion
Early data suggests that this short course is a feasible and effective way of improving early career researchers self-reported skills, confidence and intentions to work in partnership with policy agencies on research projects.

O8
Implementing the children and young people's IAPT quality improvement initiative in Cambridgeshire: findings from a process evaluation Anne-Marie Burn 1  In 2011, the Children and Young People's Improving Access to Psychological Therapies (CYP-IAPT) programme was introduced nationally to improve outcomes and experience of care for children and young people.
The aim was to support services to embed key CYP-IAPT principles into everyday clinical practice so as to deliver evidence-based practice, demonstrate accountability through outcome monitoring, increase access to effective services and engage children and young people to participate in decisions about treatment and services. The current study explored the process of implementation and the experiences of professionals in specialist mental health settings in Cambridgeshire. Methods i) An audit and analysis of national and local documents issued between 2011 and 2015 and mapped along a five-year timeline; ii) Thematic analysis of in-depth interviews with 20 staff working in three specialist mental health settings in the Cambridgeshire and Peterborough Foundation Trust, conducted at two time points along the implementation trajectory.

Results
While there was high investment from frontline staff, there was a lack of clarity about the scope and aims of the CYP-IAPT programme during the early implementation phase. Programme developers emphasised some CYP-IAPT principles more than others and there was variation in perceptions of the extent to which principles were embedded. The creation of dedicated staff posts were a key driver to implementing and sustaining the CYP-IAPT model at a local level; specialist training and enhanced supervision facilitated evidencebased practice and outcome monitoring. Barriers to implementation included inadequate and inflexible IT systems, time limited funding and a lack of support from senior management. Organisational differences between partner agencies led to ineffectual collaborative working and high staff turnover prevented knowledge continuity.

Conclusion
This study provides valuable insights into local implementation of a complex national quality improvement programme. Recommendations will inform other local quality improvement initiatives relating to children and young people's mental health.

P9
Developing an implementation strategy for the use of objective adherence data in routine clinical practice: a case study in cystic fibrosis  In this paper, we used the theoretical domains framework (TDF) to identify implementation barriers to health professionals accessing objective adherence data through CFHealthHub. In 2019, scale up of implementation is required in a further 19 CF centres.

Method
Qualitative data were collected through semi-structured interviews with health professionals in three participating UK CF centres. The participants (n=13) were purposively sampled based on location and professional category. A topic guide was created exploring all theoretical domains. Transcripts were analysed by two researchers using framework analysis.

Results
Interviews demonstrated that participants did not have routine habits for using adherence data in clinical care. Analysis indicated that an implementation strategy should address all 14 domains to successfully support implementation scale up. Participants reported insufficient training and low confidence in using adherence data. As a result, participants reported negative beliefs that adherence data would be used to "tell patients off". In addition, participants frequently justified lack of engagement because they believed adherence was primarily a physiotherapist responsibility.
Environmental barriers, such as computer access and pressures on staff time were common. Participants thought environmental and social influence barriers could be addressed by dedicated senior management.

Conclusion
The identified barriers supported the development of an implementation strategy using the behaviour change wheel. The strategy includes support for habit formation and other barriers using intervention functions, such as environmental restructuring, enablement, education and modelling. The success of this strategy will be evaluated as the project opens in new CF centres.

O10
What are the relationships between contexts, mechanisms and outcomes which underpin the 'dynamic sustainability' of a complex health care intervention across a system? Whilst there is agreement within the implementation science arena that sustainability is a key outcome, limited attention has been given to understanding how to sustain complex interventions in practice [1]. Historically, the emphasis has been on evaluating and exploring the initial implementation, with time and budgetary constraints often meaning that longer term sustainability has not been a focus. To understand what affects the determinants of sustainability, we need to understand the relationships between determinants. To address this gap, we present a review of the evidence base which aims to contribute to our understanding of these determinants and how they interrelate in different contexts. Method A systematic rapid realist review of the literature was undertaken, to examine the relationships between contexts, mechanisms and outcomes which underpin the 'dynamic sustainability' of a complex health care intervention across a system. Evidence was submitted to thematic content analysis and synthesised to create a series of Context-Mechanism-Outcome Configurations (CMOCs). Testable hypotheses based on these configurations were derived for further research.
Results 539 studies were screened, of which 17 studies were included in the review. Projects were initiated in a variety of healthcare settings and represented a range of complex interventions. The findings include a series of 17 CMOCs each illustrating how, triggered by specific contextual factors a combination of programme resource and stakeholder reasoning led to specific outcomes.

Conclusion
The review identified for the first time a set of CMOCs that determine sustainability of complex healthcare interventions. These CMOCs are empirically testable. Our further research focuses on studying how these CMOCs impact on sustainability outcomes.

P11
Implementation of non-specialist delivered counselling for depression and harmful alcohol use in

Method
This counselling has been implemented in 16 PHC facilities. We aimed to conduct three focus group discussions with the managers from these sites. Two focus groups were conducted; however, participation was low (n=7 total). We followed up with indepth individual interviews with 8 Facility managers and 1 Operational manager. Interviews were conducted in English and transcribed verbatim. Thematic analysis in NVivo 12 was conducted by two researchers independently.

Results
Three main themes emerged.
1. Perceived benefits underpin managers' support for implementation. All managers were supportive of the wider implementation of the counselling. They described a range of benefits including (i) prioritisation of mental disorders in the facility (ii) addressing social determinants not addressed by clinical care. 2. Improving facility communication. Most managers felt implementation would benefit from additional and ongoing communication (e.g. in staff meetings) about the counselling, due to constant turnover of staff and complexity of the intervention.

Strategies for improving integration into current services.
These included (i) linking with mental health nurses for counsellors to attend to less severe cases (ii) ensuring feedback to the treating clinician (e.g. HIV treatment) regarding outcome of counselling.

Conclusion
Facility managers in South Africa are a difficult to reach group given the pace and nature of their work [1]. This study indicates their support for implementation of MIND counselling despite health system constraints. Participants focused on facilitators to enable the counselling to be integrated into their facilities, which should be incorporated into future implementation studies.

Results
Phase 1 saw the academic context constraining researchers' ability to concentrate on developing KM products. Toolkits were perceived as ill-defined, practical resources, helping users put knowledge into practice. Participants reported overwhelming cynicism towards them yet felt obliged to produce them. Phase 1 themes supported Phase 2, where the impact agenda perpetuates KM product development as stand-alone interventions instead of a broader KM strategy. Cross case analysis revealed themes influencing the potential to move KM products from 'designated boundary objects' to 'boundary objects in use': need (ranging from unclear to identified gap), audience (ill-defined to clear), team dynamic (dysfunctional to collaborative), leadership ((dis)engaged), project management (chaotic to clear) stakeholder engagement (tokenism to co-production), and product (ill-defined to boundary object-in-use). KM products were characterised as symbolic boundary objects, used as bargaining chips with funders or research subjects.

Conclusion
To optimise the potential application of KM products, researchers and funders should consider the motivation for their development.
Where developing such a product is felt, potentially, to make a meaningful impact in practice, carefully planning stakeholder engagement, collaboration and project management might enhance their impact.

O13
Of In the NHS there are challenges regarding the implementation, sustainability and spread of quality improvement innovations [1]. Technology in practice offers a useful theoretical perspective to explore the implementation of quality-improvement tools. By applying this perspective, tools are viewed as everyday technologies that play an active part in healthcare as opposed to being inanimate objects in the form of paper-based or computer-based tools [2,3].

Method
This study sought to systematically and critically examine the role quality-improvement tools play in healthcare organisations. Using sepsis as an example, the study explored two comparative case studies of implementation strategies used to improve the recognition and treatment of sepsis. Ethnographic observation in the Emergency Department (ED), ward settings and relevant meetings (200 hours), semi-structured interviews (60) and documentary analysis were used. The data were analysed using thematic analysis.

Results
The empirical findings showed that when implemented there was a gap between tools and practice which was crossed by a chain of rerepresentations. Each re-representation contained a multitude of actions and decisions that were; highly distributed across a range of heterogeneous entities such as the tool itself, healthcare workers and equipment. This resulted in workarounds and various elements drifting from the original intention of the tools. For example, using the tool in the ED was different compared to the acute inpatient setting.
In the ED work practices were directed towards the recognition and treatment of acute illness. Conversely, in acute inpatient settings work practices were directed towards a multitude of other activities.

Conclusion
This research provides valuable insights into where distribution and drift occur. The findings contribute towards an understanding of the mechanisms by which improvement programmes work. An understanding of the role that tools play when used in practice can lead to more efficient and effective implementation, sustainability and spread of quality improvement innovations. The Implementation science Research development (ImpRes) tool and supplementary guide have been developed to provide guidelines on how to design high-quality implementation science research [1]. ImpRes did not include quantitative assessment criteria for evaluating the quality of implementation science protocols: these were developed in this study to identify strengths and weaknesses in implementation science research protocols and make recommendations to address current deficits in implementation research design.

Method
This study developed 'Implementation science Research development tool Protocol Assessment Criteria' (ImpResPAC), a protocol scoring system, to quantitatively evaluate implementation science protocols for information provided on five out of the 10 ImpRes domains: research characteristics, implementation theory frameworks and models, implementation strategies, implementation outcomes, and unintended consequences. ImpResPAC was applied to 16 implementation science protocols, published in Implementation Science, to identify strengths and weaknesses of planned implementation science research.

Results
The majority of protocols scored highly on describing the research gap, but few differentiated between the implementation activities associated with each implementation stage. Although most protocols provided strong rationale for the choice of theory, framework and model underpinning the research, many lacked thorough application throughout the implementation research. In most cases, the implementation strategy was explicitly described and justified. Many protocols described the intention to measure fidelity, but few intended to measure appropriateness, acceptability, penetration and sustainability. Only two of the protocols intended to measure unintended consequences. The reliability of ImpResPAC was evaluated by conducting intra-class correlation coefficient (ICC) tests on a subset of protocols and demonstrated agreement of 0.85 (ICC).

Conclusion
The application of ImpResPAC to 16 protocols identified strengths and weaknesses in research design. Furthermore, ImpResPAC presents a reliable scoring system that researchers, funders and decision-makers can use as a learning tool to self-assess implementation science research protocol quality. In line with national policy, to address the issues associated with inappropriate polypharmacy, the Centre for Implementation Science are evaluating the implementation of a Wessex Academic Health Science Network led Action Learning Sets (ALS) tool, to improve healthcare practitioners' confidence, perceptions and experiences of stopping inappropriate medicines safely.

Methods
Up to 125 healthcare practitioners attending ALS across three localities in the South of England are being invited to participate in the study. A mixed method approach involving pre-post ALS questionnaires, evaluation forms and observational methods, will be used to evaluate the study.
To date, interim findings have been analysed from one locality (n=33), and methodological insights recorded which relate to the prospective scale up and evaluation of the ALS across two further localities. Quantitative paired pre-post ALS questionnaire responses were analysed using the Wilcoxon signed-rank test, whilst qualitative data were analysed using thematic analysis and ethnographic methods. Overall impact of the ALS will be evaluated using Kirkpatrick's evaluation model.

Results
Interim findings show the ALS has contributed to improving perceptions, confidence and experiences in addressing polypharmacy and more specifically perceptions of; stopping inappropriate medicines, use of knowledge, information and tools, shared decision making with patients/clinicians and the role of institutional factors. Methodological insights relating to the scale-up of the ALS across two further localities include the impact of locality upon participation, as well as the process of engagement of participants in the evaluation.

Conclusion
Interim findings and methodological insights in scaling up the study may be of relevance to national policy to understand how to address polypharmacy issues in different localities.

P16
The Life Link Clinic: Meeting the socio-economic needs of patients and families within complex rehabilitation in-patient services Claire Hendry

Background
Feedback from patients and families in complex rehabilitation services highlighted that socio-economic issues increased their stress and anxiety levels. In response, an integrated provision is being piloted within complex rehabilitation inpatient settings under the umbrella of the Cheshire and Merseyside rehabilitation network. This service aims to offer a multi-agency approach to support patients with any socioeconomic issues, to help reduce stress and anxiety, to promote integrated accessible support.

Method
We conducted sixteen Semi-structured interviews with staff and service users to evaluate the existing three services and improve the implementation of the two new sites. Using the DASS 21 and the NHS family and friends test, service users are asked to provide feedback on the service and rate their stress and anxiety before and after accessing the service. Staff and service users' perception of the service is also collected through semi-structured interviews. Using a mixed method analysis, the semi-structured interviews were thematically analysed and paired t-test utilised to compare changes in preand post-data.

Results
Interviews indicated that this support was considered essential, which helped to improve the two new sites implemented since December 2018. The service evaluation is on-going. Preliminary analysis suggests that stress levels in families are above the ranges which we see in patients on admission to the acute rehabilitation setting. Using information from the staff questionnaire, socio-economic issues encountered by patients, carers, and families were mapping across the main themes raised when accessing the service. This reflected that the service was meeting the needs of patients and families.

Conclusion
Findings from the evaluation of this implementation of two new sites and five in total across the North West Coast region can help to identify the benefits of this service to people with complex brain injuries and their families in terms of their social needs. The education in quality improvement programme (EQUIP) for urology aims to address the need for an evidence-based curriculum in quality improvement (QI) methods training in surgery. Key goal of EQUIP is to be nationally scalable in the UK. Here we present the development of the EQUIP scale-up strategy.

Methods
The strategy was developed using evidence triangulation from mixed-methods studies: evidence-based curriculum development (evidence reviews), national gap analysis (qualitative data), and programme piloting at two UK sites for feasibility and effectiveness (quantitative data). The collected data involved a range of stakeholders: training programme directors and trainers as well urology trainees, and a steering group which included educational practitioners, patient representatives, trainees and consultant urologists. Further, we framed the implementation model of the scale up plan based on the Exploration, Preparation, Implementation, Sustainment (EPIS) framework (Fig 1).

Results
We identified prioritization challenges, Trust resources, and time commitments as the greatest barriers towards the scale-up of EQUIP, while structured approaches to QI work as the drivers towards developing an educational platform in QI techniques, a network to support undertaking QI work and creating a QI infrastructure.

Conclusion
This study offers an empirical and theoretical understanding of how to scale-up the EQUIP programme nationally in the UK. Based on the study, the scale-up strategy involves national stakeholder support; an IT platform; nationally shared training materials; and a train-thetrainers programme. The scale-up strategy will be evaluated further in the coming 2 years of EQUIP implementation and will be used to guide potential diffusion across other specialities.

O19
The One of the biggest challenges facing health-care professionals who care for women in labour and birth are decisions about the appropriate use of clinical interventions. Interventions for example caesareansections or instrumental births are necessary when problems arise, however routine use increases mortality and morbidity. We undertook a systematic review of studies to explore facilitators and barriers to the implementation of evidence-based practices to support physiological labour and birth, an important initiative, to reduce routine intervention use. We reviewed studies that explored practices in obstetric setting where routine intervention use is higher compared to midwife-led settings.

Method
Using PRISMA guidelines [1], databases was searched from 1990 to September 2018 and 31 original studies were included for thematic synthesis [2]. Analytic themes that were theoretically informed enabled us to explore facilitators and barrier at a micro level (obstetricians, midwives and women) and meso level (organisation) to implementing EBPs to support physiological labour and birth.

Results
The synthesis showed that prevalent risk perceptions of birth are an important barrier. This informed an approach based on risk surveillance and active management of labour. Obstetricians who hold strong risk perceptions of birth exert control over other professionals to apply a risk-based approach. An important barrier is their reluctance to relinquish this power. Approaches cognisant with EBPs to support physiological labour and birth is muted. Midwifery acquiesces, obstetric and midwifery preoccupation with risk surveillance and rationalisation of intervention use are important barriers. Women expect interventions to shape birth experiences. Creating tools to support people with SMI to assume more control of their physical health may help to ameliorate some of these problems. This study evaluates the use of a service user-held Physical health plan (PHP) for service users to determine whether its use increases uptake of physical health services.

Method
We have undertaken a pilot quality improvement study to test the use of the PHP [1]. This included testing a Theory of Change using focus groups with service users and staff and evaluating the use of the PHP in two community mental health teams using the RE-AIM implementation framework to guide the evaluation using qualitative and quantitative measures to do this (eg collecting data on how many people used the PHP, if they required support, if there was a change to uptake in physical health services, measuring Patient Activation)

Results
Focus groups were run with staff and service users before the pilot started which led to some change to the PHP and how it was used. 27 service users have completed the PHP. Participants have found it easy to use and were able to identify physical health needs. These were shared with clinical staff. Follow-up is ongoing which will include interviews with service users and clinical staff to elicit their views on using the PHP. Analysis of the implementation outcomes using the RE-aim framework is ongoing.

Conclusion
This study used an implementation framework to test a novel intervention for people with SMI. Participants found the PHP easy to use and analysis will help us to evaluate the success of the intervention and whether a larger-scale definitive RCT is warranted.

P22
Measuring Quality Improvement (QI) in healthcare is widespread [1] with diverse projects and target outcomes [2]. Success rates are highly variable, often with disappointing results [3]. Existing guidance on QI evaluation [4] does not aid organisational reporting on the impact of QI [5]. We aim to develop a framework for measuring success of QI to aid organisational learning and shed light on capacity building needs and resource allocation priorities in South London and Maudsley (SLaM) NHS Foundation Trust. Method A scoping exercise was conducted by an academic faculty (KC) who is embedded as a researcher-in-residence in SLaM. With combined inputs from QI team in SLaM and academic colleagues in Centre for Implementation Science as well as a rapid evidence scan, we designed a survey for retrospective summative assessment. A generic input-process-output framework was used for the survey to overcome the difficulty of making comparisons when project-specific aims and outcomes were heterogeneous and highly localised. Each QI team member then identified 3-5 "successful" and "unsuccessful" QI projects to create a gradient of contrast for small but rapid gains in learning about pertinent factors of QI success. We did not formally operationalise the definition of success in the survey so as to understand "work-as-done" rather than imposing a "work-as-imagined" criteria in the absence of an established definition of QI success [6].
Logistic regression was used to compare the impact of contextual, input and process factors on whether projects achieved their aims.

Results
Among 52 QI projects across five boroughs of London, 12 achieved their aims. Contextual and input factors showed very little impact. In contrast, process factors like stakeholder engagement and measurement plan showed sizable impact on whether projects achieved their aims.

Conclusion
Measurement plan and data quality matters more than contextual and input factors in determining success of QI. Schools need mental health interventions that are deliverable, scalable and effective. However, rates of successful implementation of evidence-based practices in schools is low. An underpinning factor is lack of understanding, and attention, by commissioners, providers and schools, to evidence about what works to bolster effective and sustained implementation. Knowledge brokering is one way to improve use of implementation science in program delivery in complex settings. To date, evidence for the use of KBs in public health settings is scarce, and their use for brokering implementation related knowledge is non-existent. The aim of this study was to examine whether it was feasible and acceptable for knowledge about implementation to be brokered to a steering group responsible for a wide scale, public mental health, school-based intervention, and what impact this might have on their implementation decisions.

Methods
The primary researcher attended 13 (3 hrs each) monthly steering group meetings during the initial implementation phase of the public mental health program and shared implementation knowledge with them. With consent, meeting notes were taken by the primary researcher, and a journal kept of 12 meetings; 5 meetings were also audio recorded and SG minutes referred to. All data were amalgamated and ordered into month-by-month summaries. The analysis attempted to identify when implementation knowledge was shared with the group (coded a 'key moment') and what the steering group did with this knowledge (coded a 'key outcome').

Results
Over the 13 meetings, 15 key moments led to 14 key outcomes, 10 of which involved implementation decisions being made based on  In Autumn 2018, Newham Clinical Commissioning (CCG) group piloted two service enhancements to groups of six general practices under the national Primary Care Home (PCH) programme: 1) managing in-hours demand for doctors' appointments through online triage, care navigation, and additional GP hours shared between practices, and 2) a complex case (CC) management team for patients with multiple co-morbidities. The evaluation aimed to assess the feasibility and acceptability of the PCH model and identify factors that could affect the roll-out of services across the Borough. Method A mixed-methods, formative process evaluation was carried out between December 2018 and May 2019: 1) routinely collected data, 2) interviews and focus groups with practitioners, deliverers and a patient participation group, 3) anonymised patient questionnaires, 4) web analytics, 5) document analysis. Normalization Process Theory [1] informed the development of research tools and analysis.

Results
In relation to managing in-hours demand, practices utilised to capacity the additional GP hours provided by locum doctors; this service enhancement was well-received by practices and patients alike. The implementation of online triage and care navigation was mixed, working most successfully in practices that were fully committed to embedding the approach and tackling challenges that arose. Although the CC management service was viewed positively by attending patients, particularly in addressing unmet social care needs, there was low uptake of the service by patients and carers, and practices did not 'buy-in' to the service within the pilot timeframe.

Conclusion
Within six months, the tested PCH models helped to widen access to GPs and stimulated partnership working ahead of the implementation of mandated Primary Care Networks, as envisaged in the NHS Long Term Plan. Securing the commitment of practices to joint working and ongoing communication work between all stakeholder groups was crucial to the success of the model.

P28
Developing and scaling up a strategy for patient and public involvement in the Centre for Implementation Science and King's Improvement Our achievements showcase how we have implemented our PPI strategy from 'ideas' to actionable achievable results. These include:

Figure 1 (abstract 24). Key moments and Key Outcomes from sharing implementation science knowledge with the SG
Setting up the Public Involvement Advisory Panel to offer expertise and advise from public members to staff on projects within the Centre Two public members recruited to our working group which oversees the operational running of the strategy A monthly Involvement Bulletin to keep in contact with public members, build positive working relationships and offer opportunities to get involved A guidance document being co-developed to offer advice and information for staff in the Centre on how best to carry out PPI activities Conclusion We have developed multiple ways to include public voices throughout the work of the Centre, striving for parity between academic and public members and we continue to move forward with public involvement that is not tokenistic, but trying to reach 'beyond the usual suspects' [1]. The school-based supervised toothbrushing program with fluoride toothpaste twice a day medical demand in the most deprived cohort of children. Health inequalities were decreased as well [1]. However, the implementation of a school-based brushing program is not merely based on oral hygiene instructions. Knowledge transfer has shown little effect on behaviour change. Changing behaviour requires an understanding of the context which behaviours occur [2]. The multifactorial nature of human behaviour made it difficult to evaluate this program through traditional randomized controlled trials. Therefore, we used a qualitative study design to analyse the potential facilitators and barriers in this complex intervention program [3].

Method
Focus groups and semi-structure interviews were used to collect the ideas from the elementary school students, teachers, and staff in this program. The topic guides were developed according to the Theoretical domains framework combined with COM-B Model. Data were analyzed with a content analysis approach.

Results
A sustainable environment which could encourage brushing behaviour was important. Toothbrush and toothpaste need to be supplied, and a set brushing schedule should be arranged. Brushing became a daily routine. Students may supervise each other brushing without the teachers' help. Opportunity in COM-B model, instead of motivation or capability, was the key theme to implement the school-based brushing program.

Conclusion
Future studies will re-evaluate the program and develop a logic model in order to transfer the toothbrushing program to other schools. The WHO Surgical Safety Checklist (SSC) reduces surgical morbidity and mortality by 50% but is not widely used in low and middle income countries (LMICs). Evidence of the effectiveness of SSC implementation in LMICs is lacking. Therefore, we aimed to systematically identify all SSC implementation studies in LMICs, the implementation strategies used, and critically evaluate success based on implementation outcomes.

Method
We used the PICO framework to design a systematic review of SSC implementation strategies and reported outcomes in LMICs, registering the study on PROSPERO. 73 strategies and 8 outcomes were pre-defined according to the Expert Recommendations for Implementing Change and Proctor's implementation outcome framework respectively.

Conclusion
This systematic review identified wide variation in implementation strategies used and outcomes reported in LMICs. Successful implementation showed a non-statistically significant association with mean strategy importance rating, as well as an association with improved clinical outcomes. Mortality from hypertensives disorders is reducing [1] but morbidity is prevalent [2]. Uptake of evidence-based guidelines and the associated determinants is unknown. Method Convergent parallel multi-method evaluation of evidence uptake and integrated analysis to inform implementation strategy. National survey, case-notes review, observations and qualitative interviews at 3 NHS Trusts (2018).

Figure 1 (abstract 030). Implementation strategies, implementation outcomes, and clinical outcomes
Results Survey (n=97) Self-reported compliance with guidelines was high. Absence of antihypertensive prescribing guidance was mimicked by variance in reported practice.
Case-notes (n=55) 100% cessation of teratogenic medication thereafter variance in first-line anti-hypertensive prescribing existed. Target setting documentation occurred infrequently (56.4%) however action was taken when bloodpressure exceeded 150/100mmHg (76.5%) resulting in internalised setting (20%) and absent setting (23.5%). 45.5% of women developed severe hypertension. Observations (n=23) Provision of correct amount and type of information to support shared decision-making (52%). Women's involvement in decisions (43%) and offered choice of anti-hypertensive (19%). Women's interviews (n=18) 14/18 experienced conflict about hypertension management, conflict resulted in 8/18 non-adhering (fig 1.a). Facilitators to concordant adherence were found to be determined by 'trust' mediated by management of side effects, information and professional's approach to women (fig 1.b). Clinician's interviews (n=13) Sub-optimal information was attributed to strength of antihypertensive evidence and lack of guidance. Clinicians also lacked SDM skills and tools. Optimal pathways and schedules of care are unknown.

Conclusion
Severe hypertension is prevalent. Strategies to reduce associated morbidity should engender 'trust' and improve target setting. Standardised pathways and schedules of care are required, accompanied by research into safety and effectiveness of anti-hypertensive. . In order to improve the quality of MCCD of home deaths, SmartVA auto-analyse, an electronic decision support tool for physicians was introduced to assist MHOs in the certification process. Method SmartVA Auto-analyse was implemented in a phased manner. A stakeholder consultation was followed by site visits and workshops with MHOs. A pretest was conducted to assess feasibility and acceptability by MHOs in 13 municipalities. The intervention included training MHOs in SmartVA Autoanalyse, and MCCD (Fig 1). Next, the intervention was scaled-up to 50 municipalities across 6 regions. A mixed-methods evaluation was performed using cause of death data and group discussions with the MHOs and policy meeting with the Department of Health.

Results
During the pilot 5649 home deaths were registered (54% male). SmartVA used to certify 81% of all deaths (4333), for the remaining 19%, doctors could assign a cause of death based on the availability of medical records. Physicians agreed with the SmartVA diagnosis in 65% cases. Group discussions identified two key themes 1) SmartVA was acceptable to MHOs, and 2) It standardised certification for home deaths and improved the quality of death certification. The Government plans to roll out the intervention in a phased manner over the next 2 years. Conclusion SmartVA can be implemented by MHOs and improve the quality of death certification of home deaths in Philippines.   [ 4] was used to select the implementation outcomes: acceptability, appropriateness, feasibility, adoption, penetration, implementation cost and sustainability. We used the COnsensus-based Standards for the selection of health Measurement INstruments (COSMIN) checklist to assess their methodological quality [ 5 ] and a newly developed Contemporary Psychometrics (ConPsy) checklist to assess their robustness.

Results
Two reviewers independently screened 6,586 titles and abstracts, and 313 full papers. More than 50 publications were included, among which over half measured acceptability. Less than 10 studies measured each of the other implementation outcomes (figures pending update). Most studies reported some measurement properties. Content and structural validity were the most reported properties for the validity domain; and internal consistency for the reliability domain. Studies had overall poor or fair methodological quality, and the ConPsy checklist indicated poor instruments' robustness. However, some acceptability measures, like the Ottawa acceptability of decision rules instrument [ 6 ], or the Impact of Health Information Technology Scale [ 7 ], had excellent methodological quality.

Conclusion
This review provides a much-needed repository of appraised implementation outcome instruments for physical healthcare settings. It highlights the efforts to quantify implementations outcomes, and the need for further psychometric testing of existing measures. It also encourages researchers to consider how to best measure feasibility, penetration, implementation cost and sustainability.

P35
The Similarly funders require Principal Investigators (PIs) to describe impact pathways and statements for grant applications to demonstrate how research may be expected to inform policy or practice, and ultimately provide benefit to society. Historically the primary role of the academic has been to carry out rigorous and original research to add to disciplinary knowledge. This study aims to examine the role of the PI to describe the role of PIs in driving the adoption and uptake of their own research findings in the healthcare setting.

Method
Purposive sampling of the 163 REF 2014 impact case studies in Unit of Assessment 2 (Public Health, Health Services and Primary Care) was used to identify case studies, which cited a definitive effectiveness trial in the impact summary. Trial PIs will be emailed a questionnaire to describe the roles and work in driving the uptake of their research findings (or intervention). They will also be asked about what perceived freedoms/enablers or disincentives/barriers meant they have been able to do more/less impact generation than they would have liked, the role of serendipity/luck, how their role in enabling implementation has changed over time, and whether they made use of specific tools or frameworks to guide their evidence implementation efforts.

Results
Data will be analysed and presented using descriptive statistics.

Conclusion
The study will characterise the variety of roles that lead researchers can play in promoting the successful uptake or application of their research findings in health policy and practice, together with outlining the perceived contextual factors that have enabled (or constrained) them to adopt such roles. Recent data shows Blackburn with Darwen (BwD) has a higher than average prevalence of both over-and under-weight in children starting school. Following local concern for the lack of pre-school child weight management provision, an 8-module online health weight toolkit was co-produced in partnership between local stakeholders and academic partners. All modules include guidance from the Royal College of Paediatrics and Child Health (RCPCH), National Institute for Health and Care Excellence (NICE), World Health Organisation (WHO), and NHS. As part of the Partner Priority Programme (PPP) of the Collaboration for Leadership in Applied Health Research and Care North West Coast (CLAHRC-NWC), an implementation strategy was developed that focuses on engaging healthcare professionals with the online healthy weight intervention tool.

Method
The implementation strategy draws on the Consolidated Framework for Implementation Research (CFIR) and involves multidisciplinary collaboration between academics, healthcare professionals, local authority commissioners and four public members. This presentation will reflect on the CFIR process and milestones in the implementation of the online tool. The shared learning will offer insight into how stakeholder engagement enhances the process of implementation.

Results
The process in engaging different stakeholders in the development of an implementation strategy consolidates the acknowledgment of local expertise. However, this process can be challenging when considering the inner and outer context in creating a strong partnership for implementation. For example it required ensuring shared vision, addressing team dynamics, alignment with local pathways, exploring individual role restrictions, and alignment with national policy/guidance.

Conclusion
The CFIR provided a structured framework that enabled exploration of different factors to support the process of engaging different stakeholders and public members in the development of an implementation strategy.

Acknowledgments
We would like to acknowledge the tremendous work of Dr Ruth Young in supporting the implementation of this project. We would also like to show gratitude to Shirley Goodhew for supporting the internship process of this project.

O38
Embedding implementation science in practice: learning from the Partner Priority

Background
Increasingly, practitioners and other professionals are expected to be involved in research and evaluations not only as advisers but also as researchers. The Collaboration for Leadership in Applied Health Research and Care North West Coast (CLAHRC NWC) Partner Priority Programme (PPP) has undertaken an approach of work that facilitated this type of co-working, which in its third iteration set out to build capacity in Implementation Science. This offered a Consolidated Framework for Implementation Research (CFIR) through which practitioners could consider the drivers associated with the implementation of their projects.

Method
Four reflective narratives explore the journey travelled by practitioners new to the framework, reflecting the inside and outside setting, characteristics of the intervention and the individuals involved.
Offering guidance as to how in the future their fellow professional could be engaged and enabled to deliver Implementation Science in their institutions.

Results
Being at the forefront of change is not easy, however, the CFIR offered the practitioners a means to discuss and understand the challenges they were facing in the implementation of their projects. No longer was it seen as a case of individually named people appearing to be obstructive. It became apparent that changes were challenging their knowledge and beliefs about the intervention as the new project did not sit easily with their understanding of the wider organisation. The characteristics of one project was found to be a major challenge not only in terms of delivery but also as a concept and practice within mainstream health care.

Conclusion
Practitioners need support to enable them to apply Implementation Science frameworks such as the CFIR. In particular, they need a learning environment in which to apply the methodology, reflect on the outcome, and assess the impact whilst continuing with the 'day job'.

O39
Experience of using mental health indicators in six low and middle-income countries where mental health is integrated in primary care: a qualitative study Shalini Ahuja 1 , Charlotte Correspondence: Shalini Ahuja (shalini.ahuja@kcl.ac.uk) Background Efforts to scale up integrated mental health care in low and middleincome country are underway. Some of these initiatives are focusing on strengthening information systems for mental health. In the previous publications, we described the need and the process of implementing key indictors covering mental health service delivery in lower and middle-income countries. These measures were introduced by a multi country consortium Emerald (Emerging Mental Health System) at a district level in six LMICs (Ethiopa, India, Nepal, Nigeria, South Africa and Uganda).

Methods
In this paper, through a qualitative study with 128 in depth interviews, we assessed the feasibility, acceptability and utility of a set of indicators for routine monitoring of mental health care and evaluated the performance of a set of indicators for routine monitoring of mental health care.

Results
Most mental health indicators were deemed relevant and potentially useful for improving care, and therefore acceptable to end users. Exceptions were indicators on functionality, cost and severity. The simplicity of the data capturing formats contributed to the feasibility of using forms to generate data on mental health indicators. Health workers reported increasing confidence in their capacity to record the mental health data and minimal additional cost to initiate mental health reporting. However, overstretched primary care staff and the time-consuming reporting process affected perceived sustainability.

Conclusion
This qualitative study provides a potential model for evaluating the use of mental health information systems across low and middle income countries and recommends key feasible indicators to be adapted for similar settings.

O40
Amalgamating theoretical frameworks to understand individual and organisational level implementation Stephanie Best 1,2 , Janet C Long 1 , Melissa Martyn 3,4 , Clara Gaff 3,4 , Jeffrey Braithwaite 1 , Natalie Taylor  Despite the rapid advances in genomic medicine, few studies incorporate implementation science theoretical frameworks. 1 When exploring implementation on multiple levels, for example the use of clinical genomics in Australia, it can be challenging to identify a single appropriate approach and amalgamating frameworks may be advantageous. We used two frameworks to learn from early adopters to design interventions to support initial implementation and sustainability of genomics in wider and less genomically specialised healthcare settings. Method Working with Australian Genomics and Melbourne Genomics, we interviewed 37 people with experience of clinical genomic testing. i) non-genetic medical specialists, using the Theoretical Domains Framework (TDF) to understand determinants of practice along the patient journey ii) service-level decision makers, using the Translation Science to Impact framework (TSci) to understand organisational and external factors affecting translation phases. Individual and organisational barriers and enablers were coded to the TDF and the TSci, respectively. Areas for overlapping themes across frameworks were explored.

Results
Clinician level barriers and enablers were identified across four key target areas (patient identification, test selection, communicating results, and mainstreaming) along the genomics pathway and represented all of the TDF domains. Service-level factors were identified to represent all the TSci Key Issues across the pre-adoption, adoption, implementation and sustainability phases relating to 'Needed Practice Supports'. Overlap between frameworks was found across all target areas, but predominantly for clinician-identified barriers and enablers to mainstreaming of genomics which covered all TSci phases.

Conclusion
Our study highlights the potential to combine different theoretical frameworks to maximise the value of theoretical approaches in highly complex settings. The next steps will be to explore the extent to which a theory driven approach to intervention design (e.g., behaviour change techniques) can be applied to address both individual and organisational level barriers.
Correspondence: Stephanie Best (stephanie.best@mq.edu.au) Background Genomic testing, reading the entire genetic information, provides an opportunity to end the diagnostic journey for many patients with rare disease. Typically, return of results takes around six monthsa timeline too slow for the management of acutely-unwell children. Whilst faster turnaround times can deliver more clinically useful results, it demands a restructure of the whole genomics pathway and causes disruption to traditional practices. Melbourne Genomics led a rapid-genomic testing (<21days) realworld project in two hospitals. We used implementation science principles to learn from this project and apply them in an Australian Genomics multi-centre project for ultra-rapid genomic testing (<5days) in acutely-unwell children.

Method
Data on the barriers and enablers to implementation were gathered (observations at multidisciplinary team meetings, research journals (February 2016-September 2017) from the project leads and participating genetic clinical and laboratory staff. The multicentre project data collection also included interviews with laboratory scientists, genetic clinicians and paediatric intensivists (April-June 2019). Analysis was undertaken using the Consolidated Framework for Implementation Research (CFIR).

Results
Several CFIR constructs were prominent (see table 1) across both hospitals. Enablers were identified, and outcomes noted. Conclusion Accounting for organisational culture and tailoring implementation science principles to real-world problems are key to success. Using implementation science theory has enabled structured learning from one project and facilitated the initiation of a multi-centre project that will inform the next steps of delivering genomic testing in the acute paediatric setting

Method
This project is part of the NIHR CLAHRC NWC Partner Priority Programme, enabling partner organisations to implement services collaboratively with academics and the public. The project team includes three public advisors, two with experience of caring for someone with MND, and one working for a third sector organisation providing information and advice services for individuals with neurological conditions. Recruitment is carried out by MND association visitors, outpatient and community therapists. To ensure the occupation is based on the person's interests, the Modified Interest Checklist is used [1]. Once suitable activities are identified, the project lead attends the activities with the plwMND, acting as a support for the plwMND and the activity provider for a total of six weeks. Referrers complete questionnaires prior to recruitment to establish their current use of wellbeing services, and quality of life measures are ad-ministered to the plwMND and their family carer pre and post intervention. Interviews will be carried out post intervention with plwMND and their family carer, the referrers, and community providers.

Results
Starting in February, two plwMND have utilised the service so far. Both have recently finished working and expressed concern that they had no structure or purpose. They have chosen varied activities including reading and history groups, art classes, bird watching, and accessible exercise.

Conclusion
Currently, we are at the beginning of this implementation project. Findings will have important implications on how a social prescribing service can be provided to plwMND, and if successful, how this service can be rolled out wider across the region and nationally.

P43
Stakeholder-driven, mixed-methods implementation evaluation of two psychoeducational programmes to reduce the incidence of severe hypoglycaemia in type 1 diabetes Tayana [1][2][3], we have set out to evaluate implementation of two novel psychoeducational courses for people with type 1 diabetes (T1D) and problematic hypoglycaemia. The courses are delivered within a hybrid type 2 effectiveness-implementation cluster randomised trial (4; NCT02940873). Here we report co-design of the implementation arm of the trial.
In a qualitative manner, the stakeholders reviewed for relevance, feasibility and clarity a selection of outcomes, tools [12], approaches, measurement time-points and participant groups. Following each Interviews for the multi-state trial are ongoing. Early findings indicate the importance of relationships between genetic and paediatric intensivist staff meeting, stakeholders' feedback was incorporated in an iterative manneruntil the final versions of the materials were produced for implementation within the trial.

Results
Stakeholder input has enabled relevant, feasible, and appropriate implementation outcomes, validated surveys, interview questions, participant groups, and measurement time-points to be identified. The final outcomes included acceptability, feasibility, appropriateness, fidelity, context, unintended consequences and implementation cost. Mixed-methods approach was deemed most acceptable.

Conclusion
Our study offers a stakeholder-driven methodological approach to co-designing an evaluation of complex intervention implementation within a hybrid trial. Input from HCPs and people with lived experience, who co-authored this work, continues to be instrumental in maximising relevance of this research.

O44
Looking into determinants and mechanisms related to the implementation of a pain management guideline in nursing Facilitators and barriers to the implementation of the programme.

Qualitative Interview Postintervention
HCPs, people with T1D and their relatives 11. Implementation strategies [11] Strategies used to deliver and implement the programme; they refer to methods or techniques to enhance and promote adoption, implementation and sustainability of the programme.

Qualitative Interview Postintervention HCPs
Note. *HCPs = Health Care Professionals incl. diabetes educator, physician, psychologist, and administrative support. †AIM = Acceptability of Intervention Measure [12]; ‡IAM = Intervention Appropriateness Measure [12]; §FIM = Feasibility of Intervention Measure [12] practice change [1]. This study uses a mixed-methods approach to explore the mechanisms and determinants related to the implementation of a pain management guideline in Swiss nursing homes.

Method
We conducted a process evaluation alongside an effectivenessimplementation study in four nursing homes. Based on an a priori contextual analysis in the participating homes, we developed a conceptual framework describing the hypothesized interactions of implementation strategy, contextual factors, mechanisms, and implementation outcomes. With a care worker questionnaire survey at baseline (n=136), after three (n= 99) and six months (n=83) we assessed our hypothesized central mechanism, selfefficacy in pain management, and self-reported guideline adoption. We computed linear mixed-effect models to assess changes over time in selfefficacy and logistic regressions to assess associations between self-efficacy and guideline adoption. Concurrently, we conducted focus groups with care workers (n=16) to gain a deeper understanding of the implementation process, interactions between potential mechanisms, determinants and outcomes of the implementation. Qualitative data was analysed using a deductive thematic analysis approach.

Results
Overall, there was a significant increase in self-efficacy after three and six months (p<0.001). Self-reported adoption of guideline components ranged between 44% and 73%. We found significant associations between selfefficacy and adoption of some guideline components, e.g., performing a comprehensive pain assessment.
Qualitative findings indicate a different response to implementation strategies by registered nurses and nursing aides, e.g., in role empowerment. Training workshops increased awareness towards pain, stimulating changes in pain assessment behaviour.

Conclusion
Understanding implementation processes in a multilevel context is challenging. Developing a conceptual framework can facilitate the exploration of potential mechanisms and influences of contextual factors.
indicated the importance of providing continuing education on EBHC for CMPs which may facilitate the integration of clinical research evidence into their routine practice. Timely and complete referrals for HIV clients in need of care within health facilities, facility-community and inter facility is crucial in closing the gaps for HIV care. Monitoring, tracking and routine interaction with HIV clients by health workers and case managers has been proven to reduce leakages in treatment cascades and improve retention in care. However, most health facilities in Nigeria are challenged by the non-existence of communication infrastructure or resources required to facilitate timely tracking and completion of referrals for HIV positive clients. The USAID-funded Strengthening integrated delivery of HIV/AIDS services (SIDHAS) project implemented the provision of closed user group (CUG) mobile technology as a strategy to closing the gap. Description SIDHAS procured 2,606 mobile phones and lines, distributed to 785 health facilities, laboratories and 28 community-based organizations across 13 states in Nigeria. Mobile networks were distributed according to the strength of network available in the locale. The CUG allowed health workers providing HIV care and treatment services to call each other within the network with zero charges and to follow up HIV clients referred within service delivery points and those referred to other health facilities to ensure complete referrals. A desk review was conducted on the CUG directory and communication logs for 4 states between June -December 2017.

Lessons learned
Review of the phone directory and call logs shows evidence of inter and intra facility communications, therefore contributing to improved referral and linkage rate from 39% to 74% within the period for some states. Data collected from 4 states also showed about 2240 defaulters returned within the period compared to 785 that returned to care before the CUG implementation. Overall, Improvement in completion rates for referrals and tracking and provision of adherence support to HIV clients was recorded.
Conclusions/Next steps Given the results demonstrated by removing communication barriers and bridging the gap through CUG innovation, expanding CUG distribution to support groups in communities can be explored to reducing stigma and closing communication gaps amongst infected individuals. HIV care and treatment in Nigeria is largely managed by a combination of public and private health facilities. Although these institutions have experienced declines in funding from government/private owners over the years, they still account for the majority of patients in care. To bridge the funding gap, the USAID-funded SIDHAS project introduced different types of grants financing approaches and accountability frameworks to support scale up of HIV prevention, care and treatment services in 13 states in Nigeria. We share experiences from the various accountability frameworks applied on the project. Description From inception of SIDHAS project in 2011, FHI360 Nigeria conducted thorough pre-award assessments, after which 76 sub-awards were executed in 13 states between SIDHAS and State government for public health, faithbased, civil society organization, and private health facilities. Capacity building through trainings and mentoring was conducted. Accounting practices such as system enhancement, compliance audits, quarterly grant modification, expenditure analysis, scorecard audit, fund request review, documentation and reporting was mandated for each implementing agency (IA) routinely. A qualitative content analysis of the IAs' accounting practices, frameworks, reports and HIV prevention, care and treatment services delivered was conducted between October 2016 and September 2017.

Lessons Learnt
From the monthly expenditure analysis, sub-award monitoring reports and HIV program data reported across the 76 IAs showed that 100% of the IAs were compliant with submission of financial and program reports for the period of assessment. HIV testing, initiation of client on treatment and provision of care services were not interrupted at those sites within the period. An analysis of this result further shows that the combination of capacity building and use of accounting frameworks was instrumental to the results. About 82% Improvement in quality of program and financial reports received was recorded within the year demonstrating improved technical, institutional and financial domains of sustainability. Conclusion A combination of accountability approaches resulted in improved ability of IAs to provide HIV services, manage, and account for donor resources. Providing a combination of accountability structures for IAs can expand their ability to better manage and sustain delivery of comprehensive HIV prevention, care and support services. While PEPFAR streamlined operational health facility costs in 2014, the GoN was unable to adequately fund operational HIV/AIDS related services. As a result, health workers were using their personal funds to pay for operational costs, and when they could not, services were not administered. In response to these gaps in care, the SIDHAS project developed the RFI funding model to better cover health facility operational costs. The RFI model reimburses health facilities through service providers, for operational cost incurred during service delivery. This abstract describes lessons learnt using the RFI funding model.

Description
The RFI funding model for 13 SIDHAS-supported states in Nigeria was developed to cover operational health facility costs such as telephone cards and home visits for HIV clients, internet costs for data entry, transport for sample logging to hub laboratories, and quality improvement measures. Funds reimbursed for operation costs to each health facility were calculated at 110 naira ($0.31 US) per client currently on ART in comprehensive treatment facilities and 300 naira ($0.83 US) per pregnant woman tested for HIV in PMTCT standalone facilities. The model reimbursed health workers monthly using standardized verifiable activity request and report forms.