Skip to main content

REsearch into implementation STrategies to support patients of different ORigins and language background in a variety of European primary care settings

Final Report Summary - RESTORE (REsearch into implementation STrategies to support patients of different ORigins and language background in a variety of European primary care settings)

Executive Summary:
There is significant migration to and within the European Union due to the relative economic prosperity and political stability of the region. In 2011, 6.6% of the total resident population of the EU-27 were foreign citizens, equivalent to 33.3 million. Of this number, just under 40% (12.8 million) were EU residents. The remaining 20.5 million were citizens of non-EU countries (Eurostat, Statistics in Focus, 31/2012). People migrate for a range of reasons, including economic migration, family reunification, retirement, study or asylum, resulting in a heterogeneous migrant population.
RESTORE aims to optimise the delivery of primary healthcare to migrants in Europe, who are vulnerable given their legal or financial situations and who are experiencing language and cultural barriers in their new host countries. Our research has been concerned with the “know-do” gap in relation to the implementation of evidence-based solutions to language and cultural barriers into routine clinical practice. In addition, the project focused on the translational gap between health services research and wider health related policy in migrant health care.
The RESTORE consortium has carried out research into the implementation of evidence-based health interventions which address language and cultural barriers in primary care settings. We investigated and supported implementation processes for evidence-based health interventions using a unique combination of contemporary social theory (Normalisation Process Theory (NPT)) and Participatory Learning and Action (PLA) research methodology.

RESTORE has involved training in NPT and PLA for the consortium, qualitative participatory fieldwork in Ireland, England, the Netherlands, Austria and Greece, and a comprehensive policy analysis which was led by the Scottish team. We have generated a wealth of data that addresses all of our study objectives. We have:
• Objective 1: Compiled a portfolio of twenty G/TIs that are currently available in our partner countries and established that the majority were not generated with the involvement of migrants
• Objective 2: Ascertained the relevance of NPT as a conceptual framework for studying the translation of G/ TIs into practice and have established the scale and nature of implementation work for each of the four NPT constructs.
• Objective 3: Analysed the shared and differential features of the primary care settings in the RESTORE study countries at both policy and practical levels, elucidating how primary care can impede or promote implementation of G/TIs and revealing wider social political factors impacting on implementation work.
• Objective 4: Generated evidence about very specific challenges to sustaining the use of G/TIs in routine practice. The new ways of working recommended by the G/TIs being implemented did offer benefits for migrants and their primary care providers, but certainly disrupted organisational routines and established dyadic interactions in consultations
• Objective 5: Established that it is possible to combine NPT and PLA and that they act as powerful heuristic devices to investigate and support implementation processes across our European settings.
RESTORE will have multiple impacts at the EU level. We expect the work of the project to impact on the research community, providers of primary care services for migrants, medical educators, policy makers and on migrant service users.

Project Context and Objectives:
There is significant migration to and within the European Union due to the relative economic prosperity and political stability of the region. In 2011, 6.6% of the total resident population of the EU-27 were foreign citizens, equivalent to 33.3 million individuals. Of this number, just under 40% (12.8 million) were EU residents. The remaining 20.5 million were citizens of non-EU countries (Eurostat, Statistics in Focus, 31/2012). People migrate for a range of reasons, including economic migration, family reunification, retirement, study or asylum, resulting in a heterogeneous migrant population. However, migrants share a number of common health issues, particularly among those who may be more vulnerable given their legal or financial situations, such as asylum seekers, undocumented migrants or economic migrants with low incomes. These migrant populations have an increased vulnerability to infectious diseases and ‘diseases of affluence’, which are prevalent in migrant host countries. There is also a relationship between poverty, poor health and a lack of access to health care and significant mental health issues. These issues may be exacerbated by an inability to communicate effectively in relation to health care needs and access. Language and cultural differences create barriers that affect the delivery of primary healthcare to migrants.
RESTORE aims to optimise the delivery of primary healthcare to migrants in Europe, who are vulnerable given their legal or financial situations and who are experiencing language and cultural barriers in their new host countries. Our research has been concerned with the “know-do” gap between the implementation of evidence-based solutions to language and cultural barriers and routine clinical practice. In addition, the project focused on the translational gap between health services research and wider health-related policy in relation to migrant health care.
In order to optimize the delivery of migrant primary healthcare in Europe, the RESTORE consortium has carried out research into the implementation of evidence-based health interventions which address language and cultural barriers in primary care settings. We investigated and supported implementation processes for evidence-based health interventions using a unique combination of contemporary social theory (Normalisation Process Theory (NPT)) and Participatory Learning and Action (PLA) research methodology. In the past, Normalisation Process Theory (NPT) has been used as a retrospective heuristic device to make sense of implementation processes which have already happened.. PLA has been used to invite community members who are usually excluded from decision-making in implementation projects, to participate along with other stakeholders to share knowledge and learn from each other about problems and potential solutions. RESTORE breaks new ground in three specific ways; it is the first project to: (i) use NPT prospectively; (ii) guide and understand a set of implementation journeys across European settings; and (iii) combine NPT with PLA as complementary partners.
The five main objectives of RESTORE were to determine:
• What guidelines and/or training initiatives were available in our partner countries that had been generated by primary care research in a way that was inclusive of all key stakeholders?
• How were the guidelines and/or training initiatives translated into practice by primary care staff? What were the processes of implementation, ‘on the ground’ in routine practice?
• What was the capacity of primary care settings in different countries (and, therefore, different organizational contexts) to incorporate the implementation of these guidelines/training initiatives within their current organisational arrangement?
• Will the implementation work for guidelines and/or training initiatives be sustainable - leading to normalised use of these technologies in routine practice?
• What were the benefits (if any) of using a combination of NPT and PLA to investigate and support implementation processes?

Project Results:
Experienced researchers and primary care providers from social science and clinical backgrounds in Ireland, Scotland, England, Greece, the Netherlands and Austria participated in the RESTORE project. The health care settings and organizational contexts vary across the partner countries, as do the capacities to respond to the planned implementation work. This variation has been an important feature of our work: it has allowed us to address recommendations that research about implementation processes should be examined in a consistent fashion across different contexts or settings and that such research should be meticulously detailed to allow comparisons across contexts.
The RESTORE has resulted in a meticulous detailed comparative analysis of implementation processes across European contexts
Below we present a description of the activity in each of our work packages emphasising key results from each one.

4.1.3.1 WP2 Training
WP2 Introduction
RESTORE is based on a combination of theory (NPT) and method (PLA) which is unique in the field of implementation science. NPT is a contemporary social theory, which was developed as a response to multiple failures to implement innovations in complex healthcare contexts . There are four constructs in NPT: coherence (sense-making), cognitive participation (engagement), collective action (enactment) and reflexive monitoring (appraisal). Participatory Learning and Action (PLA) research methodology is based on the work of Robert Chambers . PLA is an adaptive strategy that enables diverse groups and individuals to learn, work and act together in a co-operative manner, to share, enhance and analyse their knowledge and to plan together for positive action .
A key project objective was to identify what benefits, if any, are derived from the combined use of NPT and PLA. Therefore, in order to test NPT and PLA, we were aware from the outset that it was important for all researchers in RESTORE to have robust knowledge about both. WP2 focused on the design and delivery of training in the application of Normalisation Process Theory (NPT) and Participatory Learning and Action (PLA) research to all members of our consortium.
The objectives of WP2 were to:
1. Train consortium partners and researchers about NPT and in the use of PLA. This capacity-building was necessary to ensure that all consortium partners and researchers shared a common knowledge base, gained appropriate skills for our work together and assisted us to establish and maintain cross-setting and cross-cultural consistency within the project, while engaging in high quality, qualitative fieldwork in diverse local settings (task 2.1)
2. Conduct a comprehensive evaluation of the PLA training process across various European settings. This was important to inform the development of a flexible and reflexive model for the practical application of PLA methodology in primary care settings across the EU (task 2.2)

WP2 in Period 1: Months 1-18
During the first six months of the RESTORE project, PLA training was developed for RESTORE researchers. In terms of training materials and resources, a hard copy PLA training manual was compiled. The manual included material on the history of PLA and a description of its uses. The manual also included protocol documents for specific PLA strategies for data generation and analysis and relevant reading from academic journals and books about PLA. We planned to add material about PLA over time, as the training content was developed in response to emergent plans for fieldwork. In relation to NPT, there was less preparation of this kind as our plans were to utilise the on-line NPT tool-kit (http://www.normalizationprocess.org/npt-toolkit.aspx). We did, however, collect a small bibliography of relevant papers illustrating the development and use of NPT and developed RESTORE-specific training materials.
The delivery of training sessions coincided with face-to-face plenary meetings. The first training was for one week in Galway at month 7 (see D2.1). The PLA training was highly interactive and was designed to model a participatory mode of engagement, i.e. an attitudinal disposition of inclusion and equality that underlies all PLA work. The training also introduced researchers to some specific PLA techniques, for example Flexible Brainstorming and Timelines, which were relevant to the planned fieldwork with migrants and other stakeholders.
The NPT training also started in Galway. The NPT trainers delivered short interactive talks on the origins of NPT and its applications in health services research. The training introduced NPT’s four constructs but concentrated on Coherence - can stakeholders make sense of available guidelines and training initiatives? and Cognitive Participation - will stakeholders ‘buy into’ the implementation project and seek to organise themselves to drive it forward?
The second face-to-face training was in Vienna (April 2012). There was intensive work between PLA and NPT trainers to collaboratively design the training programme which was deliberative very “applied” in nature. For example we delivered training about the scope of designing a PLA options assessment matrix, which is suitable for comparative assessment of G/TIs, using questions informed by NPT constructs.
In terms of training resources, the PLA trainers developed a series of PLA protocol documents and qualitative fieldwork resources to standardize the researchers’ early qualitative fieldwork (April-August 2012). They also created PLA training DVDs which researchers could watch in their own time (individually or as a team). This was to allow them scope to watch and re-watch specific aspects of a technique in order to enhance their opportunities for practicing. At this stage of RESTORE, we agreed as a group that the idea of a hard copy manual would be replaced with the digital documentation and DVDs.
WP2 in Period 2: Months 18-36
There were three further face-to-face training sessions at plenary meetings during months 18-36. We met in Nijmegen (November 2012) and concentrated on developing researchers’ PLA facilitation skills in general terms but also specifically for facilitation of a PLA direct ranking technique. The content of the PLA training was tailored to match the progress of fieldwork: during SASI, stakeholders were moving to the point of selecting a guideline or training initiative as the implementation project for their local setting. The PLA direct ranking process was appropriate to ensure that this selection process was inclusive and democratic. Researchers received PLA training DVDs about direct ranking to take home and a series of relevant PLA fieldwork protocol documents.
For NPT, the training further de-mystified the NPT constructs and explored queries or concerns that researchers had about the use of NPT in RESTORE. The NPT training included interactive sessions to develop researchers’ understanding about the NPT constructs of collective action and reflexive monitoring: can stakeholders enact the new way of working recommended in their selected guideline or training initiative? Can stakeholders appraise the impacts of this and sustain it in day to day practice? These interactive sessions allowed researchers to develop their skills in identifying and “labelling” qualitative interactions to the different constructs of NPT.
In Liverpool (April 2013) we reached an important milestone: delivery of the final PLA face-to-face training session. As before, a PLA training DVD was provided. Training also focused on developing researchers’ skills for sequencing PLA techniques, for example to link a PLA flexible brainstorming with a PLA Seasonal calendar, so that tasks from brainstorming could be plotted over time to create a clear and concrete work-plan for the implementation work.
By the end of PLA face-to-face training in Liverpool, researchers had received training in numerous PLA techniques: flexible brainstorming, timelines/historical profiles, commentary charts, options assessment matrix, mapping, direct ranking and seasonal calendars. Their ‘basket of PLA techniques’ had expanded considerably, enabling them to move forward during Stage 3 (CAPES) with a wide range of options for designing PLA fieldwork which would be sensitive to the specifics of their local setting. Importantly, while the face to face training was ending, researchers were reminded and assured that there would be on-going support for PLA fieldwork by email and telephone contact the PLA training team.
At the plenary meeting in Crete (October 2013) we discussed task 2.2 a comprehensive evaluation of the PLA training process across various European settings. Plans were made to progress the collation of existing data and to develop a topic guide for a final round of interviews about research teams’ experiences of PLA training and its application in RESTORE fieldwork.
The NPT training concentrated on sharing information with each other about the use of NPT across settings and clarification of strategies to enhance quality and rigour in our NPT qualitative analysis. For example, we clarified that the work constituted a framework analysis informed by NPT. We examined a preliminary coding frame based on Irish data and, also, agreed ways to expand this coding frame using data from all settings. At the end of the NPT training in Crete we had clear and agreed goals for the coming months:
• Establish an NPT lead for each team who will take primary responsibility for data analysis
• Convene frequent teleconferences for the NPT leads and trainers to discuss coding developments.
• Circulate a first draft of coding framework in January 2014.
• Test and refine this coding framework on CAPES transcripts
Following our meeting in Crete, we provided our Data Monitoring Committee with an update on progress and samples of our data. They provided positive feedback on both.
WP2 in Period 3: Months 36-48
The PLA trainer remained available for consultation and support until the end of fieldwork. Individual teams did take up this opportunity during the final months of the project. It was noted at the plenary meeting in Dublin how valuable the consultation and support was. Specifically, researchers valued having the ability to discuss PLA experiences and arising queries from their fieldwork whenever needed, because it offered information as required but, also, confidence in the quality of our application of PLA in RESTORE fieldwork.
During months 36-48, we made significant progress with Task 2.2 Evaluation of PLA Training. We reviewed the available data for this evaluation and discussed a topic guide to generate a final round of relevant data. The final topic guide was ready for use in January 2015. In February and March 2015, we facilitated PLA evaluation group interviews with RESTORE researchers/teams in England, Austria, Netherlands, Greece and Scotland. In March 2015, individual interviews were completed. Interviews were between 50-100 minutes in length. During February and March 2015, we progressed the initial analysis of the completed interviews and indexing of all data resources for Task 2.2.
In terms of NPT training at the Dublin plenary, face-to-face sessions took place focusing on
• Sharing Fieldwork experiences – exploring NPT connections: this was a discussion based on a report from each team about the implementation journeys in each local setting. We reflected on the shared and differential nature of NPT data
• NPT experiences and coding – updates from NPT leads: this was a factual account from each NPT lead about the volume of coding that they had progressed
• NPT coding – focus on sub themes: this was an intensive analytic discussion about what sub themes were emerging for each of the four NPT constructs and involved debates about the naming of and boundaries between these sub-themes.
• NPT coding – exploring coding dilemmas: this was an opportunity for each team to present specific dilemmas about the coding process in terms of how detailed it should be (coding by line or by paragraph) and how to code issues relating to PLA and policy
All of these discussions fed into a series of revisions for the RESTORE coding frame. A revised version was circulated after the Dublin meeting for further use and testing by all teams.
Professor Carl May, Member of our International Advisory Board and lead academic in the development of NPT, was present at the meeting. He offered some key reflections on the analysis and interpretation of data. He emphasised that RESTORE does not prioritise the number of interpreted consultations that took place in a given setting because it was designed as a very detailed case study in five European settings that focuses on implementation processes.
From the Dublin training on, WP2 activity relating to NPT was less about training per se and more about co-producing an effective coding system. In this way the WP2 activities were very interconnected with WP5 and WP6.
The next training was before the plenary meeting at Granada (October 2014); this training was prepared using telephone meetings and email communication. We reviewed and developed the NPT coding frame, with particular attention to clarifying our knowledge and understanding of sub themes for each NPT construct. We discussed which sub-themes were generic across settings and could be used by all teams and which ones were truly ‘local’ codes that needed to be recorded in one setting only, so as to capture and represent a local issue. This kind of attention to generic and local coding was essential to fulfil our commitment to generating rich descriptions rather than ‘flattening’ the qualitative data. Interestingly, we noted that most of the data that was considered ‘local’ at the outset of the meeting was, in fact, relevant in other settings as well. Finally, we completed some dedicated exercises about coding data about the fourth NPT construct, Reflexive Monitoring.

WP2 Summary
RESTORE applied a unique combination of NPT and PLA to advance knowledge in the field of implementation science. Given the significance of this, WP2 was designed to support their application across national settings. Standardised training ensured a consistent, robust application of PLA combined with NPT during our research. We designed and delivered six face-to-face training programmes, led by experts in NPT and PLA from within the consortium. In addition, we engaged in an extensive series of interim training activities using digital documentation (e.g. standardized fieldwork protocols for PLA sessions), virtual learning tools (e.g. DVDs) as well as telephone conferences. Evaluations of the NPT and PLA training have shown that the training has been very responsive to expected and emergent issues in the qualitative fieldwork in RESTORE.
WP2 has resulted in new knowledge about training processes for primary care researchers in PLA and NPT.
4.1.3.2 WP3 & WP4 – Coherence and Cognitive Participation

WP3 & 4 Introduction
WP3 focused on identifying G/TIs in each RESTORE setting and initiating a PLA dialogue with stakeholders about their implementation. The objective of WP3 was to work with stakeholders using this PLA process to explore the NPT construct Coherence:
• the extent to which stakeholders discussed a particular G/TI as being a new way of working;
• the extent to which stakeholders had a shared view and understanding of each G/TI;
• the extent to which stakeholders could identify potential benefits arising from the implementation of the G/TI
WP3 is very intertwined with WP4, which relates to the NPT construct Cognitive Participation and the extent to which stakeholders engage with G/TIs. Specifically, to see if they
• were able and willing to get other key stakeholders involved in the G or TI;
• believe that it is right for them to be involved with the G or TI and if they can make a contribution to its implementation;
• believe that they have the capacity to organise themselves to collectively contribute to the work involved in implementation of the G or TI;
• believe they have the capacity to collectively define actions and procedures that are needed for the implementation work.
WP3 & 4 in Reporting period 1
The focus for period one, April 2011 – September 2012 was on WP3. The specific tasks in WP3 were two-fold. First, to conduct a mapping exercise to determine what G/TIs designed to support communication in cross-cultural consultations were available in each setting. Second, to initiate a PLA-brokered dialogue with stakeholders, to explore the extent to which they could make sense of these resources.
We worked together to develop standardised templates to guide this process. We contacted a range of statutory and non-statutory organisations by telephone and e-mail and conducted a search of electronic databases for G/TIs that had been reported in our country languages in peer-reviewed articles. We examined a large number of reports, training programmes (based on booklets and DVDs) and guidelines (n=230). Applying a set of agreed filtering criteria, we identified G/TIs of relevance to RESTORE. A final set of 20 G/TIs was compiled (see Table 1).

Table 1 Breakdown of G/Tis by Country
Breakdown of G/TIs by Country
Country Guidance Training Initiative Country Total
Ireland 4 2 6
The Netherlands 1 6 7
Greece 0 0 0
England 2 3 5
Austria 0 0 0
Scotland 1 1 2
Totals G/TI 8 12
Overall Total 20

RESTORE has resulted in a portfolio of 20 guidelines and training initiatives that have been designed to improve communication between migrants and their primary care providers.
We identified the need to create a limited set of these G/Tis for presentation to stakeholders during our planned qualitative fieldwork. This was because we wanted to present a manageable number of G/TIs to enable stakeholders to study and know them in a meaningful way. We also wanted to present G/Tis that showed some potential for implementation. Therefore, we designed a standardised process using NPT for selecting a limited set of G/TIs in each national setting. With this process, each team identified five-six G/TIs for their fieldwork setting. Our prospective use of NPT to appraise G/Tis for their implementability is novel in the field of research about communication in cross-cultural consultations.
RESTORE has resulted in the development of a method to employ NPT to appraise Guidelines and training initiatives designed to improve communication between migrants and their primary care providers for their implementability across six European settings; this method has potential for broad application in the analysis of policy and other documentation in many other domains.
To progress the second task for WP3: initiation of a PLA-brokered dialogue, partners in Ireland, England, the Netherlands, Austria and Greece obtained ethical approval and progressed sampling and recruitment of stakeholders from community groups, statutory and non-statutory organisations. This was conducted during Stage 1 of our fieldwork, referred to as “WIDE” because PLA sessions focused on Welcoming & Introducing stakeholders to the RESTORE team and to each other.

We used purposeful and networking sampling strategies to recruit stakeholders for PLA sessions in WIDE. The combined sample size across settings was N=189. The completion of WIDE occurred at the end of the first reporting period.
WP3 & 4 in Reporting period 2
During the second reporting period, the PLA dialogue continued during a second phase of qualitative fieldwork in RESTORE; this phase was referred to as SASI because the focus was for stakeholders to share (S), assess (A) and select(S) a G/TI (I) in each setting.
Stakeholders who had been recruited and attended PLA session in WIDE were invited to remain involved during SASI. As anticipated, the sample size from WIDE narrowed down for SASI as a smaller set of stakeholders decided to engage further with the project. The combined sample size across settings was n= 78 participants, with an average of 15 participants per national setting. Note the numbers are indicative of sample size because numbers did fluctuate slightly during this fieldwork stage depending on stakeholders’ availability and the changing context e.g. changing roles in the Health Service.
PLA sessions focused on the generation of qualitative data with stakeholders about the sense that they made of the G/TIs presented to them by their research team (coherence) and the factors that influence their engagement with the G/TIs (cognitive participation). In brief, the following steps were carried out in participating countries:
1. Stakeholders were provided with summaries about the G/TIs in the limited set for their country and given an opportunity to share ideas and comments about them. The discussions and ideas which ensued were recorded, including stakeholders’ ideas about what elements of the G/TIs may need to be adapted and fine-tuned for their local setting. The data were digitally recorded and captured on PLA Commentary Charts.

2. Research teams reviewed the emergent data from these discussions and charts to consider its resonance with NPT. This was a crucial step as researchers were able to identify any gaps in the research, for example if there was limited data on cognitive participation, which was the empirical focus of WP4. Figure 4 shows the guidance provided to researchers to colour-code data on the aforementioned PLA commentary charts according to the NPT construct it pertained to. A blue spot was used to indicate resonance of data with Coherence, a maroon spot indicated resonance of data with Cognitive Participation and so on.

3. Stakeholders were then given an opportunity to rank the G/TIs in order using a PLA direct ranking process and, from this, to democratically select one as their implementation project.

From this interactive, productive and democratic PLA process, one G/TI was selected for implementation by stakeholders in each setting. The results are shown in Table 2 below.

Table 2 Final set of Selected Implementation Projects
Country G or TI? Code Title
Ireland TI and G R/IRL 25a “Working with an Interpreter is Easy: Self-Directed Training Package for Health Professionals “
The Netherlands TI R/NL 24c “Did I explain it clearly?” How to communicate with migrants with lower education and less command of the Dutch language”
Greece G R/IRL 1b “Guidance for Communication in Cross-Cultural General Practice Consultations”
England TI R/NL 21c “Ears of Babel. Culturally sensitive primary health care”
Austria TI R/SCO 2e “New European Migrants and the NHS: Learning from each other, Manual for Trainers, First Edition February 2009’, NHS Lothian, Dermot Gorman”

RESTORE has demonstrated the effective involvement of migrants in health research: migrants worked with other relevant stakeholders in Participatory Learning and Action research focus groups to democratically select one guideline or training initiative for implementation in their local primary care setting

This step completed the fieldwork of WP3 and WP4. While analysis and write-up continued to the project end, the major focus of activity now became the work of adapting and fine-tuning them to local settings. This theme was continued in further implementation work relating to enactment (Collective Action) and appraisal (Reflexive Monitoring), which were the foci of WP5 and WP6 respectively.
WP3 & 4 Summary
In summary, from the second reporting period the goals for WP3 and WP4 were achieved;
a) Completion of mapping exercise to develop a portfolio of G/TIs
b) Sampling and recruitment for fieldwork to initiate a PLA dialogue
c) Presentation of the portfolio of G/TIs to stakeholders
d) Selection of G/TIs by stakeholders in each setting for their local implementation projects
e) Generation and analysis of data about Coherence and Cognitive Participation

The work completed in WP3 and WP4 underpinned the work that followed in Period 3 of the RESTORE project. WP3 and WP4 provided baseline data for WP5, which was focused on collective action work and its subcomponents: interactional workability, relational integration, skill-set workability and contextual integration. Discussions were initiated on how use the NPT sensitizing questions related to collective action and reflexive monitoring (WP6) for the qualitative fieldwork being carried out in each setting.
4.1.3.3 WP5 Collective Action
WP5 Introduction
WP5 continued the PLA-brokered dialogue with stakeholders, focusing on the progression of implementation journeys and aimed to determine with stakeholders how the intervention, which they had selected, co-designed and sought to implement would be enacted in their routine work. WP5 worked with stakeholders to explore Collective Action: the extent to which
• stakeholders can perform the tasks required by the guideline/training initiative;
• stakeholders can maintain their trust in each other’s work and expertise through their use of the guideline/training initiative;
• the work required for the implementation of the guideline/training initiative is appropriately allocated to stakeholders with the right mix of skills and training to do the work;
• the implementation work is supported by organizational context and resources

WP5 relates to stage 3 of our fieldwork, referred to as CAPES because stakeholders:
• Collaborate (C) by engaging in PLA research to
• Assess (A)
• Participants’ (P) experiences and
• Exploration(E) of implementation work and generation of potential
• Solutions (S) to challenges of implementation
The two principal tasks for WP5 (Task 5.1 - PLA-Brokered Dialogue with Stakeholders and Task 5.2 - Co-Design Potential Solutions) were carried out in parallel in each of our five fieldwork sites.
WP5 in Reporting period 2
In WP5, we continued to work closely with our stakeholder groups from Stage 2 SASI. In CAPES we worked with 65 stakeholders across all fieldwork sites, of which 20 joined the project during CAPES. Note as mentioned previously in relation to the sample for Stage 2 (SASI), the numbers here are indicative of sample size because numbers did fluctuate slightly during this fieldwork stage depending on stakeholders’ ability and the changing context e.g. changing roles in the Health Service.
The duration of CAPES fieldwork ranged from 15-19 months. This was longer than the estimated time frame of 12 months for this fieldwork component, reflecting the complexity of the implementation journey in each setting. A total of 62 PLA sessions were held in CAPES in all countries, with an average of 12 per country.
Prior to delivering the chosen G/TI in each setting, we supported stakeholders using a variety of PLA techniques in their efforts to refine, amend or co-design their chosen G/TI, taking into account identified problems. We worked extensively with stakeholders to find customised solutions that made sense at their local level. The work of adapting G/TIs in this way proved more substantial, iterative and organic than anticipated. In some instances, there were several cycles of adaptation, delivery and evaluation carried out by stakeholders.
The novel use of Participatory Learning and Action research in RESTORE enabled migrants and other relevant stakeholders in RESTORE to successfully adapt their selected guidelines and training initiatives to enhance their relevance for their local primary care setting
WP5 in reporting period 3
While the earlier CAPES reports for Stage 3 fieldwork contained data about the co-design processes, CAPES reports submitted during reporting period 3 (April 2014 and July 2014) related more to the translation of the modified G/TIs into real world primary care settings. From this we conducted preliminary comparative analysis of the five implementation journeys, highlighting factors that promoted and inhibited implementation work, discussing commonalities and divergences between fieldwork sites, and noting unresolved issues.
Factors promoting implementation
1. In all centres it was crucial to the success of the implementation journeys to have adequate time. Adaptation and delivery of the various G/TIs were not simple tasks. Time was needed for stakeholders in all centres to undertake substantial local revisions to the chosen G/TIs. Time was also needed for practitioners to take part in the variety of training initiatives on offer, and then to enact the knowledge and skills generated by each G/TI in the context of their routine, everyday practice. In all settings, there were challenges engaging stakeholders. However, the investment of time, while demanding, does seem to have enhanced stakeholders’ ownership of the implementation work. This dynamic between time as a demand and time as an investment, and how this impacts on stakeholders’ engagement with the implementation work, requires further analysis.

2. In four of the five fieldwork settings there was a conscious decision to expand the focus of the G/TIs from an initial emphasis on one staff group to encompass a wider participant group; and in the remaining setting there is an aspiration to do so in the future. We consider that this was a direct consequence of the effective utilisation of PLA methodology, which not only encouraged local stakeholders to consider who would benefit from the G/TI but also provided clear mechanisms for effective expansion of stakeholder and training groups. Furthermore, PLA techniques (e.g. Seasonal Calendar, Flexible Brainstorming) provided valuable and powerful ways for inter-stakeholder groups to work together, share experiences and learn from each other.

We note, however, that the expansion of interest to wider professional and practice groups was not without limitations. For example, some practice assistants in the Netherlands, and some reception and GP staff in Ireland, did not fully engage. This may have been due to lack of interest in some cases, but was compounded by contextual factors, including family difficulties and local organisational pressures.
3. There was a creative set of decisions by local stakeholder groups to enhance their own expertise by involving technical experts in the CAPES implementation journeys. Local stakeholders identified tangible benefits from bringing in people who possessed requisite skills which were not readily available within the existing group. Training organisations were included by the Dutch and English teams, in the latter case becoming key members of the local stakeholder group. The Austrian team brought in international and national topic experts to help in delivery of their TI. In Greece, the introduction of a third sector organisation, the Hellenic Red Cross, was an essential element of guideline delivery. In three sites, the involvement of policy makers was helpful in ensuring that the G/TI was adapted to fit within the local policy context. In Ireland a dual approach was adopted, bringing outside expertise in the form of the training team, but thereafter drawing on the expertise already available within the stakeholder group. This worked as a strong means of affirming existing expertise, and was also a means of cementing bonds of trust that served well when difficulties were experienced.
4. Migrant stakeholders brought fresh perspectives to the implementation process. In the Netherlands, for example, they reminded the stakeholder group that it is not always a case of choosing between family and professional involvement in the interpreted consultation; family members can play a valuable role providing support and information, in addition to the work undertaken by professional interpreters. In England stakeholders from the Chinese community emphasised that cultural and linguistic barriers are not confined to recently migrating groups and communities. In Ireland, migrant stakeholders were instrumental in the production of attractive, culturally and politically appropriate recruitment materials.
5. Transnational co-operation was also in evidence. The English stakeholder group communicated directly with the Dutch training organisation Pharos, when considering methods for building evaluation processes into their TI. The Irish and Greek teams exchanged ideas regarding their plans for undertaking PLA ‘Walk-Throughs’ or mock runs of the interpretation services they were setting up, as part of implementing their selected guidelines, prior to commencing their delivery. The Scottish team supported all national teams to identify and capture stakeholders’ views on the existence of national and regional policies, and their impact on the implementation of their chosen G/TI.

Factors inhibiting implementation
1. The policy context impeded several CAPES implementation journeys. Drawing significantly on data from WP7 as well as fieldwork data it was clear that there was broad policy pressure, given Europe-wide economic austerity, with huge cuts to budgets (Ireland and Greece) and a tightening of budgets elsewhere. In England there was also a major restructuring of the health care system. In specific relation to migrant health care, the Dutch national decision to cut funding for professional interpreters in primary care had a major negative impact on availability of resources: this led stakeholders to adapt their selected training so that it included content about informal interpreters. The political context in Greece, giving priority to indigenous rather than migrant health care needs (a policy present though less overt in all other centres) put plans for local guidelines implementation at serious risk until the stakeholder group was able to involve the third sector organisation. The position and organization of primary care in certain countries also presented challenges to the implementation in study sites. For example, in Austria, family physicians tend to work in single practices, with little practice team support.
2. There was a combination of practical and socio-cultural difficulties in involving migrants as stakeholders in the planning, adaptation and delivery of G/TIs in several centres. In Ireland and Greece, the investment of time in adaptation work truncated time for the recruitment of migrants to use the new interpreting services. More time for fieldwork may have resolved this issue. In the Netherlands and Austria, the problem of involving migrants in the adaptation of training was addressed by holding separate stakeholder events for migrant service users.
3. Each fieldwork centre intended to focus its attention on a specific health condition: mental health in England, Austria and the Netherlands; diabetes in Ireland and Crete. This focus was maintained in England, Austria and Ireland, but not in the other two centres. In part this was a practical matter. In Crete, for example, the focus had to be on setting up a generic new interpretation service, and focus on a particular group would have unnecessarily limited the scope of the new service.
4. While there was agreement on the importance of PLA approaches across all field work sites, as identified in the methods, we note that there was variation in the practical application of PLA techniques between sites. Application was extensive in Ireland, moderate in England, the Netherlands and Greece, and more restricted in Austria. In planned papers we will consider the extent to which this variation may have impacted on the enactment of the implementation journey in each setting.

WP5 Summary
From our third reporting period, the goals for WP 5, which were to generate and analyse data about Collective Action, were achieved. Stakeholders were supported to
• perform the tasks required by the guideline/training initiative;
• explore their trust in each other’s work and expertise through their use of the guideline/training initiative;
• allocate the work required for the implementation of the guideline/training initiative to each other and additional technical experts
• explore the extent to which the implementation work is supported by organizational context and resources
In all five fieldwork centres, PLA dialogues were instrumental to the generation of these data. We highlight the significant involvement of migrants across centres in the overall, combined sample of stakeholders. Throughout WP5, we have been impressed with the on-going commitment to the RESTORE project across stakeholder groups and across settings as evidenced by their sustained involvement from Stage 2 SASI onwards.
The combination of NPT and PLA in RESTORE has comprehensively elucidated the complex implementation work involved in translating guidelines and training initiatives to improve communication in cross-cultural consultations into ‘real-world’ primary care settings.
4.1.3.4 WP6 Reflexive Monitoring
WP6 Introduction
WP6 continued the PLA-brokered dialogue with stakeholders described above in WP3, WP4 and WP5. WP6 focused on the last stage of the five RESTORE implementation journeys and aimed to determine with stakeholders how the intervention, which they had selected, co-designed, sought to implement and enact in their routine work, would be appraised. This relates to the NPT construct Reflexive monitoring (RM).
The appraisal of the interventions implemented in WP5 focused on whether stakeholders could
• determine how effective and useful the G/TIs were in their local settings, using formal or informal evaluation methods?
• collectively agree, or not, about the worth of the G/TIs in their local settings?
• agree that the G/TI was worthwhile?
• modify their work (individually or collectively) in response to the appraisal of the G/TI?
WP6 in Reporting Period (Months 1 to 18)
The NPT construct Reflexive Monitoring focuses on appraisal of a G/TI after a period of use. Therefore, we did not expect stakeholders to focus on this kind of implementation work in the early part of the implementation journeys. Thus, there were no pre-determined tasks for WP6 during this first reporting period.
WP6 in Reporting Period 2 (Months 19 to 36)
The principal tasks for WP6 were initiated after the implementation of the G/TI, which started across sites during our second reporting period. Given the close inter-connection between WP5 and WP6 the same sample of stakeholders were involved in each WP.
Stakeholders in each setting planned a variety of ways to appraise the implementation of the G/TI. We recorded plans for obtaining specific plans to document subjective data i.e. views and experiences of migrant service users, and objective data i.e. the registration of the (increase of) use of interpretation services in the UK, Ireland and Crete, the (increase of) registration of the language and literacy level of individual patients in The Netherlands, and waiting time at the front desk in the Netherlands).
Given the differences in chosen G/TI for implementation, there was a large variation in the amount and nature of data generated about reflexive monitoring, during the second reporting period. For example, in Ireland, the UK, Austria and the Netherlands, a training initiative was implemented relatively early in Stage 3 CAPES. The training initiatives were immediately evaluated with participants. In each of these settings, the training was evaluated using PLA techniques and was judged very positively by the stakeholders. The training was viewed as helpful by the professionals, as they learnt strategies for improving transcultural communication in their daily practice. In Ireland, some additional training needs (e.g. more information on culture and cultural competence in healthcare) were identified; these were addressed during subsequent PLA sessions. There was no equivalent data from Crete.
Another example was that, by April 2014, some data from the Netherlands and the UK were already available on the appraisal of the effect of the training in daily practice. Stakeholders in these settings subsequently reported important organisational changes as result of the training, such as gathering information about the use of interpretation services, the registration of language and literacy of patients and about the way to improve accessibility of the practice for migrants with low literacy. Also more consultations were scheduled on the half or whole hour. In contrast, no such data was available in Ireland about organisational changes, because recruitment of migrants to use the newly available interpreting services was low.
Finally, in line with NPT and PLA’s attention to macro-level issues and in keeping with WP7 work, stakeholders in each setting were encouraged to pay specific attention to policy impacts, and to consider when and where there is scope for RESTORE to impact upon policy configuration
In summary, at the end of the second reporting period, all national research teams had observed or encouraged stakeholders’ dialogues about appraisal and stakeholders at all sites were involved in gathering data on reflexive monitoring. This meant that the key foundation work for WP6 was achieved during the second reporting period and it was clear that data gathering was a priority going forward in the final months of RESTORE fieldwork.
WP6 in Reporting Period 3 (Months 37 to 48)
Focusing on findings about reflexive monitoring, we found that there were some commonalities in the documented plans for appraisal. There was a CAPES session planned for appraisal in each site – this was agreed by the consortium. All sites that delivered training initiatives planned to conduct participatory speed evaluations of the training. Exit interviews with service users, to appraise the impact of the G/TI on practice, were planned in Ireland, England and Greece. There were also examples of site-specific appraisal activities such as a plan to use an email message system from the Dutch research team to prompt the staff who had received training to reflect and report on its impact on their daily work.
Drawing on data from these completed appraisal activities, we learned that the implemented TIs in Ireland, England and the Netherlands were very positively appraised. In England, practice staff reported that they had enjoyed the drama format of the training in their setting and that the use of actors had made the training memorable.
In Austria most participants in the training were, in general, favourable about the e-learning mode they had experienced, but most reported that the e-learning component was "too long" and should be shortened.
Other completed activities, such as an email feedback system in the Netherlands, provided important data about organisational change. These data showed that stakeholders were more aware of the signs of low literacy and were more sensitised to communicate more effectively with poorly-literate migrants or with non-Dutch speaking migrants and of the disadvantages of informal interpreters. However, they reported that it was difficult for them to implement lessons learned in every consultation as it cost them more time. In particular, practice assistants were hesitant to spend more time on this communication as they felt the migrant him or herself has a duty to communicate better as well.
Some of the planned appraisal activities were not progressed. Focusing on this difference between planned and actual appraisal activities, we noted that some plans could not be progressed because of challenges implementing knowledge from training into routine practice. Specifically, in Ireland and Crete, there were significant problems involved in recruiting migrants to use the newly available interpreting service. This reduced the scope for the planned exit interviews for example. Other appraisal plans could not be progressed because of inter-agency challenges. In England the collection of quantitative data regarding the volume of interpreted consultations prior to and after the training was not pursued. This was because the interpreting company could not release such information to the practice because that would breach their confidentiality agreement with the practice.

WP6 Summary
From our third reporting period, the goals for WP6 - to generate and analyse data about Reflexive Monitoring - were achieved. Stakeholders were supported to work together to consider the effects of their selected G/TIs on practice, to explore their worth, and the scope to modify existing routines to sustain the new way of working in day-to-day practice.
In all five fieldwork settings, PLA dialogues were once again instrumental to the work of WP6. PLA sessions allowed stakeholders to continue to share their individual perspectives and to arrive at agreements about appraisal plans, appraisal activities and the meaning of the resultant data. Importantly, these data represented an ending point in the RESTORE implementation journeys. Stakeholders completed their work together in each site. Furthermore, these data provided the final set of results for our overarching NPT analysis of the processes involved in translating G/TIs into routine practice. Thus, the end of WP6 marked a significant point in the overall project.
The combination of NPT and PLA in RESTORE has revealed important insights into the significant challenges that arise when trying to sustain the use of knowledge from guidelines and training initiatives to improve communication in cross-cultural consultations into ‘real-world’ primary care settings
4.1.3.5 WP7 – Policy Analysis and Guidance
WP7 Introduction
The comparative analysis of migrant health policy in the RESTORE countries, and the development of policy guidance for the EU, were the principal foci of WP7. Led by partner 3 (University of Glasgow), this work involved all the Consortium partners - through the attention paid to the policy environment in each country during the empirical data collection described in WPs 3-6, through the identification and analysis of policy reports from each country and by interviewing elite stakeholders in each country.
WP7 had three tasks. Task 7.1 – conducted primarily in the first reporting period - was a desk-based analysis to identify and describe the macro-level policy context for migrant health in the partner countries, including a description of the migration patterns and health care structure in each setting. Task 7.2 – reported in our second period review period - identified and reviewed specific policy documents and guidance relating to the provision of health care for migrants in general, and primary care in particular, with a focus on issues relating to language and communication. Task 7.3 conducted over the last 12 months of RESTORE, comprised a qualitative meta-synthesis of empirical data collected within WPs 3-6, along with data generated in Tasks 7.1 and 7.2 and interviews conducted with elite stakeholders within the European Union and in each national setting.
WP7 in Reporting Period 1: Months 1-18
During this phase, the team focused primarily on Task 7.1. Initial work focused on identifying data sources which could be used to compare the health systems of the RESTORE countries and on developing migration profiles for each country. Given the focus of RESTORE on migrants in vulnerable situations, we particularly focused on the migration of asylum seekers to RESTORE countries, as well as undocumented migrants and those seeking employment, especially in lower paid work.
Health systems
The team commenced by reviewing the health care structures of each partner country and, importantly, the structure and location of primary care within each health system. Initial country profiles were circulated and the content agreed with each in-country research team. At this time, we drew on the comparative health system work conducted by Wendt and, in primary care by Kringos, to further develop our understanding of each country’s health care system and the place of primary care within that system. Our descriptions of the health care system focussed on variables likely to have an impact on migrants’ access to health care and on practitioners’ responses to migrant patients, e.g. the type of funding for health care (public vs private provision); the extent to which there is universal coverage; whether or not registration with a GP is required; and gatekeeping functions of primary care.
Migration profiles
Migration profiles were developed for each country, focusing on a brief history of migration in each country; a review of immigration, emigration and net migration for each country over the past decade; and an overview of vulnerable migrant groups (in particular asylum seekers, refugees, undocumented migrants and victims of human trafficking). Comparison between countries was challenging, due to the limitations and differences in datasets and the definitions used in each country. As a result, we used high-level data sources as much as possible, for example data obtained from the EU EUROSTAT database for migration data; the Organisation for Economic Co-operation and Development (OECD); the International Organization for Migration (IOM); the European Migration Network (EMN); and the United Nations High Commissioner for Refugees (UNHCR). RESTORE in-country teams reviewed the datasets ad identified and provided additional information on national sources where appropriate.
Period 1 Outputs
During this period, the team produced two substantive outputs. The first was a document outlining the key data sources in relation to migrants; this was made publically available on the RESTORE website (http://fp7restore.eu/pdf/web-link_resources.pdf). The second - describing the key international legal instruments relating to migrant health care – gave the in-country research teams reference information during empirical data collection.
RESTORE has resulted in the compilation of a compendium of resources which characerise key sources in relation to migrants
WP7 in Reporting Period 2: Months 19-36
During this time period, we completed Task 7.1 and developed and almost completed Task 7.2. Key preparatory work for Task 7.3 was also carried out.
Task 7.1 - Identification and Description of Macro-Level Policy Context
Task 7.1 was completed in March 2013, when the country-level migration profiles were finalised. These contained a brief overview of the country’s migration history; net migration over the past decade, including data on the number of asylum applications; and a synopsis of the wider political and policy contexts in which migrant health care is situated, particularly in relation to those in vulnerable situations. This work identified several key issues that became the increasing focus of work in Task 7.2: for example the increasingly restrictive nature of migration policy apparent in several RESTORE countries, with states moving to a points-based system for migrant entry. Further, the increasing demand for proof of language competency as a condition of entry for some migrant groups was also evident, e.g. in Austria, the Netherlands and the UK.
Task 7.2 - Identification and Description of Specific Policy Documents and Guidance
Here, we focused on the identification and review of policy documents relevant to the provision of health care for migrants generally and in general practice/primary care in particular. In keeping with the overall aim of RESTORE, the focus here was on policies relating to intercultural health and support for communication in cross-cultural consultations. Policies which operated at national, regional and local levels were all included. This allowed us to use locally-generated data from WP3-6, as described in the DOW, as well as European, national and regional policy.

EU-level directives were identified by searching the European Union database EUR-Lex, which gives open access to EU laws and directives. All RESTORE partners contributed to the identification of national, regional and local policies, using document pro-formas developed by the WP7 team. For each identified policy document, teams were asked to complete a more detailed policy document description. Once completed, the descriptors and a copy of each policy document were shared for analysis. Given the remit of the policy analysis, teams considered documents around:
• primary care and migrant health issues,
• language issues/cultural competence in migrant health,
• language issues/cultural competence in primary care.
These multiple sources of policy identification ensured that local knowledge generated with stakeholders during the WP3 mapping exercise, and the fieldwork conducted in each country during WPs 4 to 6 were captured and fed back to the policy work.
This led to the identification of 69 documents, mainly from the desk-based searches, the WP3 mapping process, and through team knowledge (Table 3). Where required, policy documents from Austria, Greece and the Netherlands were translated; translated documents were sent to the respective RESTORE teams to ensure the quality and accuracy of translation.

Table 3 Identification of policy documents across the RESTORE countries
Country Number of documents
Austria 6
England 13
Greece 12
Ireland 15
Netherlands 7
Scotland 16

Two frameworks were applied to the qualitative analyses of the identified policies: Mladovsky’s framework for the analysis of migrant health policies; and Normalisation Process Theory. The use of NPT was particularly notable because this represents the first use of NPT to analyse health policy as a high level complex intervention (as opposed to a trial or service delivery intervention). This work was fully developed during the final year of RESTORE, and is described, below.

RESTORE has resulted in the novel use of NPT to analyse health policy as a complex intervention

Task 7.3: Qualitative Meta-Synthesis
Work commenced on Task 7.3 during this second reporting period. Key activities included: preparing an application for ethical approval to allow key stakeholder interviews to be conducted; a videoconference with the International Advisory Board, in February 2013, to discuss the work; and the provision of support to researchers in the partner countries to plan for and pay attention to issues of policy arising during their CAPES sessions.
Period 2 Outputs
An ethics application was prepared in January 2014 and submitted to the Research Ethics Committee of the College of Medical, Veterinary and Life Sciences, University of Glasgow. Ethical approval was granted in April 2014.
During this phase of the project, the policy team produced a series of briefing papers to support the teams involved in WPs 4 to 6. These include a final set of migrant and health care system profiles in each country; an overview of the international and European migrant health policy context; and detailed plans for each country to ensure that issues relating to national, regional and local policy are not lost during data collection in WPs 4-6. In addition, we presented at several national and international conferences; published a letter in The Lancet; wrote a paper on the role of primary care to facilitate health care access and use by migrants in vulnerable situations which has been submitted for publication; and contributed to several other papers in RESTORE (see WP8 for more detail).
WP7 in Reporting Period 3: Months 37-48
This final period of work focused mainly on Task 7.3. There were two principal activities:
• Qualitative interviews with EU and national stakeholders and analysis of policy relevant data generated during WPs 3 to 6 fieldwork.
• Meta-synthesis of data generated across Tasks 7.1 to 7.3.

Qualitative interviews with EU and national stakeholders (elite stakeholders)
Policy makers and experts were identified from the organizations operating at European and country level in the development of migrant health policy, including WHO Europe and the International Organization for Migration; from the knowledge of in-country RESTORE teams and their stakeholder groups; and from the RESTORE International Advisory Group.
Named individuals were contacted and asked if they would consent to be interviewed. In general, these interviews were conducted by telephone, although a smaller number were conducted by Skype or face-to-face. Those interviewed were also asked if they could, in turn, recommend others. This snowball sampling allowed us to identify a wider group of potential interviewees and to deal with the ever-changing policy environment in some countries. This was a particular issue in Greece, as a result of health care changes due to austerity measures; and in England, due to NHS re-organisation. Interviews were recorded, with consent, and transcribed. Interviews were analysed thematically, using the interview topics as a guide; codes and emergent themes were then mapped broadly to NPT.
A total of 21 individuals were interviewed across the six RESTORE countries. Interviewees came from a variety of settings, including European institutions concerned with migrant health; policy makers working in national health systems; heads of health care professional organisations; members of non-governmental organisations.
Data analysis
Both policy documents identified in Task 7.2 and excerpts of data generated during in-country fieldwork (WPs 4 to 6) were analysed using two theoretical frameworks: the policy framework of Mladovsky et al and Normalisation Process Theory. Mladovsky’s framework was helpful in looking at migrant health policies in general and considering where language and cultural competence were situated within those policies. Normalization Process Theory (NPT) helped us to understand whether and how such policies were implemented in practice by drawing our attention to four main constructs.
For the policy analysis work required in RESTORE, we re-considered the questions that NPT would ask of policy. This work was developed following discussions with Professor Carl May, of the International Advisory Board.
Meta-synthesis of data generated across Tasks 7.1 to 7.3
The final task of WP7 – Task 7.3 – was to synthesise the data generated across the RESTORE consortium, particularly from WPs 5 and 6, in combination with the data generated throughout WP7. This was facilitated by considering each country as a case study, and informed by the approach of realist synthesis.
Case studies and realist synthesis
The diversity of health systems, policy environments, migration profiles and socioeconomic contexts across the RESTORE countries made direct comparisons of data challenging. Thus, a case study approach was taken, bringing together the data outlined above for each country.
Realist synthesis asks “what works, for whom, in what context, how and why”. Its focus, therefore, is not on the intervention itself, but on the interplay of contexts, approaches and the way in which they might work (referred to as mechanisms) and outcomes which leads to successful – or unsuccessful – implementation of complex interventions. This approach had value for RESTORE because it allowed us to consider what aspects of policies work best in certain contexts. This was achieved by generating a series of research questions and propositions to be tested using the range of data previously described.
Data synthesis in RESTORE
A range of data were brought into the research synthesis:
• Migration profiles (T7.1)
• Health care systems reports (T7.1)
• Policy document analysis (T 7.2)
• Empirical fieldwork from WPs 4 to 6, in particular the CAPES reports and coding extracts from in-country teams
• Project Deliverable reports (WP3-6)
• Elite stakeholder interviews (WP7)
Data synthesis was focused by the initial aim of exploring the question: What elements of the policy environment best enable the implementation of interpreted / culturally competent consultations in primary care?
This was discussed with the RESTORE team and with Professor Carl May of the International Advisory Board. This led to development of a series of key research questions designed to explore the impact of each national policy environment on interpreting and cross-cultural competence and how it affected the implementation of the selected interventions at the local level (shown in the box below).
Key research questions to guide research synthesis
Key questions
1. Where in the policy environment is cross-cultural communication situated? (E.g. Is it situated within equality and diversity policy; language and communication; integration; patient safety and risk?)
2. What does the policy environment consider in relation to cross-cultural communication? (E.g. Does it discuss interpreting provision; use of cultural mediators; other approaches to ensuring cross-cultural communication?)
3. At what level are these policies operating and how? (E.g. National, regional or local policies?)
4. Where policy does exist, what implementation monitoring is taking place? Are the results of monitoring being used within primary care?
5. Does the primary care structure support the implementation of such policies?
6. What resources are in place to support cross-cultural communication? (E.g. Legislation, funding, training)? At what level do these operate (i.e. national, regional, local)?
7. Do countries with established diverse communities have better policy support for cross-cultural communication?
8. Are there other wider factors which support/hinder the implementation of policy? (E.g. austerity measures; structural re-organisation; resourcing)

Using these questions we synthesised the data in two steps:
(i) We constructed case study summaries for each RESTORE country.
(ii) Drawing from across these case study summaries, we then developed a set of recommendations; these are directed at European Government, national Government or health systems; professional bodies involved in undergraduate and postgraduate teaching; academic researchers. Each recommendation is embedded in our empirical analysis from each case study.

Recommendations
A series of 13 recommendations were developed which are directed towards the European Union; national Government and health systems; educators; and the research community and funders. These were presented to an audience of researchers, policy makers, practitioners and migrants at the RESTORE Final Conference, at the University of Limerick in March 2015.

WP7 Summary
WP7 has explored and identified how previous migration histories, health care systems and health care policy interact to provide a set of conditions which promote or hinder cross-cultural communication in European primary care. Of particular importance is the identification of the negative impact that broader societal-level changes can bring about – in particular austerity measures, but also health system re-organisation and flux. In both cases, these wider contextual influences can render cross-cultural communication “invisible”, with the provision of interpreting and other culturally competent services not being prioritised.
In conducting this work, we have developed a set of recommendations as outlined above. Following the project, we continue to focus on developing and disseminating these recommendations to both EU-level organisations and within the participating RESTORE countries. For example, the RESTORE team are in dialogue with the Council of Europe and the WHO about dissemination opportunities and country teams are actively maintaining relationships with national stakeholders.

Potential Impact:
WP8 focused on impact, dissemination and exploitation of results. We expect the work of the RESTORE project to impact at an EU level on the research community, providers of primary care services for migrants, medical educators, policy makers and on migrant service users.
We aimed to achieve scientific impact through our publications about
(i) Implementation of G/TIs to support communication in primary care settings. This area is under explored in the existing literature
(ii) Involvement of migrants in health research. There is a lack of health policy and practice to promote and support migrants involvement in research which is counter to the existing international imperatives around Public and Patient Involvement in research more generally
(iii) The operationalization of NPT and PLA in a prospective implementation project and the impacts of their use in RESTORE. This issue is very novel in the field of implementation science.
The main dissemination activities and exploitation of results are described below per task in WP8.

Task 8.1 - Project Website, Logo and Newsletters
The website was launched at www.fp7restore.eu in May 2011. A second web domain www.health-of-migrants.com was also launched in September 2012 which is cross-linked with the domain www.fp7restore.eu to ensure that migrant communities also have access to information the RESTORE-website prepares for them. The official logo was finalized and has been used for the website and for all RESTORE dissemination activities. We prepared and disseminated three project newsletters . The language of the newsletter was English, with an option to, in addition, produce versions for all partners in their respective languages.

Task 8.2 – Education: Dissemination of training processes
The focus of Task 8.2 was on education and dissemination of the training methods and training materials distributed to the RESTORE consortium. There have been numerous conference presentations, which have focused on dissemination of training methods for NPT and PLA. These have been valuable for the education of other primary care researchers.

Task 8.3 - Creative and Innovative Dissemination
New venues and channels for dissemination were actively discussed and explored at our consortium meetings and in fieldwork session with RESTORE stakeholders. As a result we set up a RESTORE twitter account which was used, along with RESTORE members’ individual accounts, to tweet about the project milestones and events.
In addition, we discussed the merits of identifying existing community events (e.g. science weeks, science festivals, immigrant-community festivals) that we can collaborate with. Over time ideas about creative and innovative dissemination involving stakeholders were streamlined into the RESTORE final conference. We invited and resourced stakeholders from the English, Irish, Greek and Austria sites to attend the conference. We highlighted migrants’ involvement in RESTORE at the opening plenary by interviewing migrants from the English and Irish setting about their experiences of the research project. This was highly commended by conference delegates who noted how rare it was to see this kind of priority attention to involving migrants in conferences and dissemination.
In addition to the innovative plenary session, stakeholders from the Irish setting participated in a workshop by demonstrating PLA techniques. Stakeholders from the English setting participated in an oral presentation by getting involved in the Q&A session. Stakeholders from the English site, who had led the training programme for cultural competence, presented a video summary of their training as a plenary at the conference.
Other relevant activities by the Irish partners included:
• Production of a video presentation for the International Implementation Research Network in Primary Health Care (IIRNPC) Australian Primary Health Care Research Institute (APHCRI) conference in Canberra, July 2014. This is available on the RESTORE website and is titled: “How do we better involve stakeholders in primary health care research? An example of the use of PLA in a six-country European project.” The purpose of the video was to display the innovative use of PLA in RESTORE as an excellent method for involving migrants in research. The video was produced by the Irish academic team and two migrants who had been involved in RESTORE fieldwork in Galway city. The video explains what RESTORE and PLA are about and describes migrants’ positive experiences of PLA. It illustrates that this is a very valuable approach for meaningful involvement of migrants (and other) patients or service users in primary healthcare research.
• Support provided to migrant stakeholders to present RESTORE at a national community interpreting conference and general practice stakeholders to present at a national general practice conference. Both these events were June 2014.

Task 8.4 - Educational Materials for medical educators
A task force was established during the April 2012 Vienna project meeting to identify relevant educational materials over the remainder of the project and to consider appropriate forms of dissemination to educators of healthcare professionals.
Information about the portfolio of guidelines and training initiatives, identified during the mapping process in WP3, was posted on our website for medical educators’ use during period 2. This includes information about the newly designed Austrian training initiative which was developed during RESTORE by adapting a Scottish initiative.
In Ireland, material about RESTORE was included in lectures to the UL Graduate Entry Medical School lectures for Year 3 students from 2012.
In the Netherlands the ‘Walk in the Park’ material, an exercise utilised during the initial PLA training (WP2) to raise awareness of cultural diversity was translated into Dutch and implemented in a session on cultural diversity for undergraduate medical students in Nijmegen and in the national master class for GPs working in deprived areas in the Netherlands. It was also used during the EACH meeting in Madrid in April 2014 and is used for training activities by Pharos (Dutch centre of expertise on health disparities) in general practices to train GPs’ and practice nurses' cultural competencies.
In Ireland in 2012, UL joined as a partner in a new EU Lifelong Learning project (C2ME) about cultural competence in medical schools (led by the University of Amsterdam) . One of the Dutch team was also involved as an International Advisor. The aforementioned Walk in the Park’ material is being shared with EU Lifelong Learning C2ME colleagues from Edinburgh University who are interested in its potential for their undergraduate programme. This demonstrates the impact of RESTORE PLA training in wider educational fora.
Finally, a collaboration with a Dublin based GP training programme resulted in the development of a module about migrant health for the Irish College of General Practitioners. The material was informed by RESTORE and C2ME. This will become part of the national curriculum for post graduate general practice training in Ireland.
Task 8.5 - Media Activity
RESTORE was highlighted in a number of media venues during the first 18 months. A project Press Release was printed in the local Rethymno, Crete Newspaper highlighting the RESTORE consortium visit to the Regional Council of Rethymno on October 25, 2013. The Dutch research team was interviewed by the local university magazine about RESTORE. One of the Dutch team was also interviewed by Mednet, an online magazine for medical professionals. RESTORE was highlighted in Health Matters, the publication of the Irish national Health Services Executive, in issues 91 and 92 (2013). Finally, as a result of a press release for the RESTORE final conference, an article appeared in The Irish Times, a national broadsheet, on March 27th 2015 about the need for professional interpreters in the healthcare system. There was an interview on local radio about these issues as well. Two migrants who had participated in RESTORE were interviewed as well as the project co-ordinator.
Task 8.6 - Scientific Dissemination
The RESTORE consortium has been highly active and productive with regard to scientific dissemination. In this section, we present a chronological record.
April 2014
− Cross –cultural communication: a key dimension in migrant health policy? at European Public Health 5th International Conference on Migrant and Ethnic Minority Health in Granada
May 2014
− Tackling language and cultural barriers in primary care: experiences gained in Greece from a European collaborative project at 1st Well-Med Conference in Alexandroupolis
June 2014
− Irish College of General Practitioners Annual research meeting in Kilkenny– stakeholders from the general practice involved in the fieldwork presented the research
− National Community Interpreting Conference in Dublin- community interpreter stakeholders involved in fieldwork presented the research
− ‘Language Matters: exploring the place of cross-cultural communication in migrant health policy in Europe’, Medical Research Council seminar, in Glasgow


July 2014
− ‘Supporting communication in cross-cultural primary care consultations: how to implement guidelines and training initiatives across Europe’, SAPC Workshop in Edinburgh.
− ‘Supporting communication in cross-cultural primary care consultations: how to implement guidelines and training initiatives across Europe’ Workshop (c. 50 participants) at 19th European WONCA-conference in Lisbon, Portugal,
September 2014
− ‘Austrian training initiative 'Cross-cultural competencies for General Practitioners', a poster presentation at the Royal College of General Practitioners (RCGP)-conference in Liverpool
− ‘RESTORE: Using Normalisation Process Theory and Participatory Learning and Action research to investigate and support the implementation of guidelines and training initiatives to improve communication between migrants and their primary care providers’ – Participatory Research at McGill seminar series, Montreal, Canada.
October 2014
− ‘Improving cross-cultural communication in primary health care settings in Greece: The RESTORE project’, at the EGPRN ‘Economic Crisis and Research in Primary Care’ Conference in Heraklion, Crete.
− ‘The healthcare of migrants in Greece and language barriers they encounter at the primary health care level”, at the 17th National Medical Conference of Greece in Heraklion, Crete.
November 2014
− At the Comunaid Workshop on Migrant Health at the Technological Institute of Crete, Department of Social Work in Heraklion, the Greek RESTORE-team presented the RESTORE project and conducted a PLA exercise with the participants.
− ‘The RESTORE Project; Results and findings in light of the economic crisis of Greece’ at the 8th National Conference of Public Health and Social Medicine in Thessaloniki, Greece.
− "The RESTORE Project; Results and findings in light of the economic crisis of Greece", at the 8th National Conference of Public Health and Social Medicine in Thessaloniki.
− "Methods to engage the most difficult to reach: developing care for migrants in Europe" at the NAPCRG Preconference in New York.
− "Cross-cultural communication: a key dimension in migrant health policy at the NAPCRG in New York.
− Two workshops were delivered at the NAPCRG in New York: "Using Normalization Process Theory to understand why complex interventions are (or are not) implemented" and "Supporting communication in cross-cultural primary care consultations: how to implement guidelines and training initiatives across Europe".
March 2015 2014
− ‘Prospective use of theory in implementation research: Normalization Process Theory in the EU RESTORE project’ at the National Primary Care Conference in Ireland
− “Implementing cross-cultural communication guidelines and training initiatives on the ground: A participatory study exploring implementation work with stakeholders in five European primary care settings on the ground” at the National Primary Care Conference in Ireland

In addition to the many conference and workshop activities outlined above, the consortium also had the following journal publications:

− MacFarlane A, O'Donnell C, Mair F, O'Reilly-de Brún M, de Brún T, Spiegel W, van den Muijsenbergh M, van Weel-Baumgarten E, Lionis C, Burns N, Gravenhorst K, Princz C, Teunissen E, van den Driessen Mareeuw F, Saridaki A, Papadakaki M, Vlahadi M, Dowrick C; Research into implementations strategies to support patients of different origins and language background in a variety of European primary care settings. Implement Sci. 2012 Nov
− van den Muijsenbergh M, van Weel-Baumgarten E, Burns N, O'Donnell C, Mair F, Spiegel W, Lionis C, Dowrick C, O'Reilly-de Brún M, de Brun T, Macfarlane A.; Communication in cross-cultural consultations in primary care in Europe: the case for improvement. Prim Health Care Res Dev. 2013 April.
− O'Donnell C, Burns N, Dowrick C, Lionis C, MacFarlane A; RESTORE team. Health-care access for migrants in Europe. Lancet. 2013 Aug 3;382(9890):393.
− MacFarlane A et al. Healthcare for migrants, participatory health research and implementation science. Eur J Gen Pract2014 Jan 27. [Epub].
− de Brún et al Guidelines and Training Initiatives that support communication cross-cultural primary care settings: Appraising their implementability using Normalisation Process Theory Family Practice 2015 Apr 27. pii: cmv022. [Epub ahead of print]

The consortium developed a publication strategy and has a number of other academic articles drafted and due for submission in the coming months.
Task 8.7 - Dissemination to Policy Makers
Outreach to policy makers gathered momentum from 2013. Service planners or policy makers were involved in fieldwork in many case study sites. There was a strong focus on connections with policy makers in WP7. In addition, there was engagement during fieldwork with
• the Austrian Chamber of Physicians (particularly with the Center for General Medicine at the Austrian Chamber of Physicians)
• the Viennese Society of General and Family Medicine
• 7th Regional Health Authority of Crete
• The Regional Council of Crete
• The Social Inclusion Office in the Health Service Executive, Ireland
In the final reporting period, we focused on WP7 and the Final Conference (see below) as an important part of our dissemination to policy makers. Representatives of the Council of Europe, the International Organisation for Migration, and the World Health Organisation attended and presented at the final conference. The final plenary of the conference involved presentations about WP7 policy work and key findings. Speakers from the IOM, WHO and CoE then responded with insights and advice about taking the findings forward.
Examples of our Recommendations, as presented at the Conference, are provided below.
European Union
Statement 1 Legal frameworks and Acts help to frame migrants rights, but often do not give them any legal rights to language support
The evidence The Charter of Fundamental Rights of the European Union makes clear that “Everyone has the right of access to preventive health care and the right to benefit from medical treatment”; however the legal requirement to ensure this lies with individual Member States. Again, this implies the right to be able to communicate, but does not enshrine that right in EU law.
The Council of Europe Recommendation of 2006 recognised the existence of language barriers and recommended that professional interpreters should be made available and used on a regular basis for ethnic minority patients
There was little evidence that national legal frameworks have recognised and enacted this Council of Europe recommendation. For example, while the 2010 UK Equality Act (covering both England and Scotland) provides a legal framework for public bodies to pursue equality, there is no legal right to interpreting support for patients. This means that interpreting and translating services are not always prioritised.
In Ireland, there was a focus on intercultural health, e.g. through the Intercultural Health Strategy and the reduction of health inequalities. Language and communication support was often linked to other policy areas such as violence again women or child protection, in order to protect interpreting services.
In Scotland, the policy driver to reduce inequalities in health resulted in a key policy documents Equally Well and later Fair for All, with a clear strategy to reduce inequalities in access for BME and migrant groups by providing access to interpreting services and addressing cultural competence. There was, however, no legal requirement to provide access to interpreters.
Recommendation 1 The Council of Europe recommendation that professional interpreters be used to ensure effective communication for migrants should be upheld and enshrined as a legal requirement across Member States.

Statement 2 All patients have a right to be able to give informed consent.
The evidence Legal rulings in several RESTORE countries, e.g. Austria, enshrine the rights of patients to be able to give informed consent. This can only happen if the patient fully understands the consultation with a health care practitioner.
Informed consent was also raised by the General Medical Council in the UK, in its Good Medical Practice guidance. This places responsibility for ensuring that a patient fully understands the outcome of the consultation (in terms of advice, treatment planning and/or decision-making) with the health care professional.
Professional guidance in Ireland also emphasizes the right to ensure effective doctor-patient communication, but did not explicitly discuss how this could be ensured.
In the Netherlands, a policy shift and retraction of funding had pushed the responsibility for language and communication away from the health system to migrant patients. The Royal Dutch Medical Association (KNMG) opposed this, citing the need to ensure effective doctor-patient communication and, by implication, the need to ensure that through interpreting services.
Recommendation 2 In order to address the right to informed consent and effective doctor-patient communication, health systems will have to address interpreting rights. Member States should consider making access to an interpreter a legal right.

National Government and Health Systems
Statement 5 The location of policies to support cross-cultural communication may be situated in several different policy arenas.
The evidence Several countries had situated language and cross-cultural communication within the policy areas of patient safety (e.g. Austria), acknowledging the need to ensure that patients are able to given informed consent.
In the UK, the need for doctors to communicate effectively with all patients was explicitly raised by the General Medical Council’s Good Medical Practice guidance as a professional requirement.
Greece did not focus on cross-cultural communication, instead focusing on migrants’ entitlement to access health care services; while this might implicitly include the right to effective communication across languages, this was not explicitly stated.
In Ireland, the focus was on intercultural health and on the reduction of health inequalities
In Scotland, cross-cultural communication policies were located in the inequalities agenda and supported by work of an NHS Scotland agency charged with addressing the inequalities agenda. The Translation Interpretation and Communications Support (TICS) strategy and action plan documented the need for effective communication within health care settings.
Recommendation 5 National health systems need to ensure that patients can communicate adequately to ensure that patient safety is maintained at all times. Situating support and enabling cross-culture communication may be better placed within patient safety legislation in each country.

Educators
Statement 10 Cross-cultural competence needs to be embedded at all levels of the health system
The evidence In Austria, there were legislative changes to Acts governing the training of health care professionals to embed cross-cultural training in undergraduate curricula.
The training needs of all staff must be addressed e.g. in the UK, general practice administrative staff are the very first staff group to engage with migrant patients, yet their training needs are often overlooked.
Recruitment and retention of staff from diverse backgrounds, coupled to training and support for staff to deliver a culturally competent service, were recognised as key aspects of the Irish intercultural strategy.
Recommendation 11 Cultural competence, and working with interpreters, needs to become a key component of undergraduate and professional training across all health care professions. This requires resources, including protected time and funding, and its impact needs to be assessed.

Research Community and Funders
Statement 13 The provision of interpreters is a cost-effective use of resources
The evidence To date, there has been little work assessing the cost-effectiveness of interpreted consultations, although the costs of interpreting services was raised as an issue in some RESTORE countries.
There is also little work into different models of interpreting services and the impact on outcomes, for example face-to-face versus telephone interpreting.
Overall there is a weak evidence base about cost-effectiveness of interpreting.
Recommendation 13 Fund research into the cost-effectiveness of providing interpreters and compare different models of interpreting support, from a health system and societal perspective.

These Recommendations are currently being transformed into a policy briefing document aimed at [i] the Commission; [ii] National Governments; [iii] the International Organisation for Migration – IOM; and [iv] the World Health Organisation - WHO. They will be further developed and submitted to a high-impact Journal for publication.
The RESTORE Closing Conference
As a final dissemination activity, to present and discuss our work, and to increase the impact of the project, the RESTORE Closing Conference took place in Limerick, on 26th – 27th March 2015.
A conference management team was established in 2013(UL, PT, GU, MUW, UOL), and a flyer was produced. This was posted on our website and disseminated by the consortium at relevant meetings and events (e.g. at the EUPHA migrant health conference, Granada, April 2014)
Abstracts were reviewed during autumn 2014 and a programme of plenary and parallel sessions as well as workshops was prepared. This featured the RESTORE consortium members but, also, included presentations from other academics as we had decided that it was important to put RESTORE findings ‘in conversation’ with other on-going research
In total there were almost eighty delegates from seven countries across Europe and North America. Excellent speakers included representatives of the Council of Europe, the World Health Organisation and the International Organisation on Migration.

The first plenary session focused on the involvement of migrants in health research and three migrant stakeholders were interviewed about their experiences of RESTORE.

The second plenary focused on lessons learned about implementing guidelines and training initiatives in primary care settings. The third plenary was based on a video about cultural competence training from the English site. The final plenary was focused on policy analysis and recommendations.
The conference was a major success and enabled sharing of experience and best practice, and very much validated the RESTORE belief that cultural sensitivity and language services are important for effect primary care delivery to migrant communities. The conference had good coverage in national media and social networks , and was covered by the WHO Public Health Bulletin Summary, the blogs of the Canadian Mental Association Journal. It was also highlighted in the report of the UL Governing Body, the highest governance level at UL.

The RESTORE team has been invited by representatives of WHO, IOM and COE to sustain collaboration beyond the project; planning is underway for meetings and new collaborations which continue our shared focus on migrant health and primary care.

List of Websites:

http://www.fp7restore.eu/
Prof. Anne MacFarlane (University of Limerick - www.ul.ie)