Patient and public involvement in dementia research in the European Union: a scoping review

Background Internationally, there is a drive to involve patients and the public in health research, due to recognition that patient and public involvement (PPI) may increase the impact and relevance of health research. This scoping review describes the extent and nature of PPI in dementia research in the European Union (EU) and summarises: (i) how PPI is carried out; and (ii) the impact of PPI on people living with dementia and the public, researchers, and the research process. Methods Relevant studies were identified by searches in electronic reference databases and then filtered by two reviewers independently. Eligibility criteria for included studies were: (i) people living with dementia and/or care partners; (ii) PPI activity in dementia research conducted in the European Union (EU); and (iii) published between 2000 and 2018. An adapted version of the Guidance for Reporting Involvement of Patients and the Public (GRIPP2 SF) was used to collate the data. There was no language restriction other than the abstract needed to be available in English. Results We found 19 studies from the UK and one from the Netherlands meeting inclusion criteria. No studies from other EU countries met inclusion criteria. Studies reported various methods of PPI including workshops, drop-in sessions, meetings, consensus conference, reader consultation and participatory approach. The reported aims of PPI included identifying and prioritising research questions (n = 4), research design (n = 5), undertaking and managing research (n = 8), and data analysis and interpretation (n = 3). All PPI related to design and implementation of non-pharmacological studies. One study described two pharmacological studies as case studies incorporating PPI. Seventeen studies reported anecdotal impacts of PPI. Conclusions Further development of PPI in dementia research in the EU and in pharmacological dementia research is required. Given the wide range of objectives of PPI in dementia research, PPI methods should be flexible and appropriate for the research context. Researchers should also formally evaluate and report the impacts of PPI for researchers, patients and the general public using good quality research designs to foster development of the field and enable the benefits and challenges of PPI to be better understood. Trial registration PROSPERO 2017: CRD42017053260. Electronic supplementary material The online version of this article (10.1186/s12877-019-1217-9) contains supplementary material, which is available to authorized users.


Background
Patient and Public Involvement or PPI in research is described as "doing research with or by the public, rather than to, about, or for them" [1]. PPI involves recognising the importance of patients and public's viewpoints and concerns, and that the view of patients and the public may differ from those of researchers [1].
PPI in health research has become increasingly common due to the growing recognition that it may increase the relevance and utility of research outputs to patients and the general public and in turn make the research more cost-effective [2]. It is also considered to be ethically desirable to involve people in research that pertains to them [2,3]. PPI may also increase the effectiveness of research by making recruitment materials easy to understand, advising on appropriate recruitment strategies, and suggesting implementation and dissemination strategies [4,5].
Across Europe, healthcare policies recognise the importance of PPI in research [6,7], and the European Union Clinical Trials Regulation (due to come into effect in 2019) recommends PPI as a quality standard for clinical trial design [8]. European charities and patient groups including the European Lung Foundation, the European Patient Ambassador Programme, the European Patient Forum, and the European Patients' Academy on Therapeutic Innovation, also advocate PPI in research to ensure patients and the public can influence the development and delivery of health research [9].
Following the growing interest in PPI in health research, within dementia research specifically, there is increasing interest and awareness of the need for PPI. Alzheimer Europe [10,11] has recommended involving people living with dementia in research, and national governments and charities have emphasised that people living with dementia and their care partners have the right to contribute to research [12,13]. To help us understand the extent of PPI within dementia research in Europe, we undertook a scoping review to map out this area which has not been previously reviewed. This review summarises: (i) how PPI with people living with dementia is being carried out; and (ii) the impact of PPI on people living with dementia and the public, dementia researchers, and the research process within the European Union. The focus is on studies in European Union due to unique research, medical and social care arrangements within the European context.

Methods
We followed the guidance on Preferred Reporting Items for Systematic reviews and Meta-Analyses extension for Scoping Reviews (PRISMA-ScR) Checklist for this scoping review [14] (See Additional file 1).
For the purposes of this review, PPI is defined as involvement in research being carried out 'with' or 'by' members of the public rather than 'to', 'about' or 'for' them [1], the term 'care partner' is denoted as a spouse, family member or professional caring for a person living with dementia. Additionally, for the purpose of summarising the data, the term 'impact' refers to the outcome as a result of the changes made from PPI input in the papers identified for this review (Table 1).

Patient and public involvement in the review
We involved public contributors in this review to help us understand what the review findings mean for people with dementia and care partners and to inform our learning from the review findings. Three public contributors to a dementia focused research advisory group (including one person with early onset of dementia and two care partners, all aged over 65 and based in the UK) was consulted via a face-to-face meeting to discuss their perspectives on the emerging findings from this scoping review.
The public contributors were informed about the purpose of the meeting, given an overview of what a literature review is, and what the aims of the review were. The group was then presented with a short version of Table 4 printed in large fonts. The PPI group was asked to comment on whether they agreed that PPI is important in dementia research, and what they thought about the different PPI methods identified in this review. We asked for the group's views on what would be a good outcome if they were involved in dementia research and Table 1 Possible outcomes pertaining to patient and public involvement in research [2] Patient and public involved

Research questions
This scoping review addressed the following questions in relation to PPI in dementia research: 1) How is PPI being carried out with people living with dementia within the European Union? 2) What are the reported impacts of PPI on people living with dementia and the public dementia researchers, and the research process within the European Union?

Inclusion criteria
The population, intervention, comparison, outcomes and study design (PICOS) structure guided the study inclusion criteria for this review. Inclusion criteria were: (i) people living with dementia and/or care partners; (ii) PPI activity in dementia research conducted in the European Union (EU); and (iii) published between 2000 and 2018. The time period from 2000 to 2018 was selected to ensure that recent practice was included. There was no language restriction other than the abstract needed to be available in English. There were no restrictions on study design and grey literature was included. We excluded studies reporting public engagement activities which focused on dissemination of research [1]. Book reviews, opinion pieces, unpublished theses and literature reviews were also excluded.

Search strategy
We used the term 'patient and public involvement' as a starting point to develop a search string and identified additional keywords that were used in articles to refer to PPI; this enabled us to build a free text search strategy for PPI (Table 2). We then combined the search string for PPI with the Medical Subject Heading (MeSH) keywords for 'dementia' (

Study selection
JM and SE independently reviewed the titles and abstracts against the eligibility criteria of 9203 studies. Full texts of potentially eligible studies were then manually filtered by JM and SE, a third reviewer (SP) was consulted for any disagreements and reconciled through discussion.

Data extraction
The review protocol was used to develop a data extraction form (See Additional file 3). Form fields included study characteristics (author, country and year of the publication, study aims, study population, study design), the PPI term used, PPI approach, whether PPI impact was evaluated, how PPI was evaluated and evaluation findings.

Collating and summarising data
Due to the wide variation in PPI approaches in the included studies and the variability of the quality of reporting of PPI, a descriptive approach was used to summarise the results of the review. This involved descriptive summary tables with headings based on the adapted Guidance for Reporting Involvement of Patients and the Public, short form version 2 (GRIPP2-SF) [15] guidelines (Table 3) to describe and summarise the data. Adaptation was made to the GRIPP2-SF [15] by adding the following additional fields (i) which EU countries the study had taken place (ii) the population involved and (iii) what PPI terms were used and how this was reflected in the study methodology. Additionally, because our review aimed to identify the impact of PPI, we added fields to capture whether studies incorporated systematic evaluation methodology describing how the impact of PPI was assessed on PPI members, researchers and on the research process and what the conclusions of the evaluation were. We used the NIHR's research process model [16] to describe which stage of the research process researchoriented PPI was focussed on, including; identifying and Results 9,203 studies were identified through database searching, eight were identified through other sources (hand searching of references and recommendation from colleagues), yielding 155 studies after removal of duplicates. Fifty-one full text papers were assessed, and twenty studies were included (Fig. 1). Thirty-one studies were excluded: ten were research studies with people living with dementia and care partners as participants; nine studies reported PPI in clinical service development/assistive technology; seven reported PPI but the PPI activity was not in relation to dementia research; five articles were commentaries citing dementia and PPI. Nineteen of the included studies were conducted in the UK and one was undertaken in the Netherlands. They were all published between 2005 to 2018 (see Table 4).
How PPI with people living with dementia is being carried out The key areas of PPI were in identifying and prioritising research questions (n = 4), research design (n = 5), undertaking and managing research (n = 8), and in data analysis and interpretation (n = 3).
A variety of terms were used to describe PPI activity, including workshops, drop-in sessions and meetings [27], individual meetings [22], modified Delphi process combined with a consensus conference and anonymous reader consultation [30], participatory approach [23], interviews, focus groups, questionnaire, voiceover group meetings [26] and a patient and public involvement event [36].
The impact of PPI on people living with dementia, the public, dementia researchers, and the research process within the European Union Only three studies formally evaluated the impact of PPI in dementia research [23,25,33]. Most studies reported impacts of PPI anecdotally. Of the three studies that included a formal evaluation of impact, Stevenson et al. [25] used a paper questionnaire asking what members liked most and least about the session, and also tried to capture their perspective on the benefits of being a coresearcher in the exercise. Littlechild et al. [23] used semi-structured interviews and focus groups to evaluate the impact of involvement in all stages of the research process from prioritisation and formulation of research questions, study design, recruitment, data analysis and interpretation to dissemination. Littlechild et al. [23] surveyed the viewpoints of co-researchers, statutory organisations, voluntary organisations and academic researchers [23]. Finally, Morgan et al. [33] employed an online survey, semi-structured interviews and focus groups to evaluate the impact of volunteers who provide PPI input in research projects funded by the UK Alzheimer's Society. These studies identified the impacts of PPI in dementia research as follows: Three studies described PPI in identifying and prioritising research questions [17,18,26]. Studies incorporating PPI in research design reported various impacts, particularly during development of non-pharmacological interventions for people living with dementia. For example Yates et al. [21] described how PPI informed development of a one-to-one, carer-led cognitive stimulation intervention for people living with dementia [21]. The authors [21] reported that PPI helped with the development of the drafts of the intervention manual and activity workbook in terms of language clarity and generation of ideas for the materials used in the intervention. Practical issues with the intervention were also identified by PPI feedback, including prioritising time to complete intervention sessions, and suggestions of ideas of how care partners can overcome barriers to completing the intervention. They also [21] reported that PPI feedback helped the research team understand reasons for non-adherence to the intervention and an awareness of the support that care partners may need in carrying out the intervention.
The following were identified as impacts of PPI input in undertaking and managing research: identifying issues of importance to patients and care partners during development of a research proposal [26,32], assisting in the development of non-pharmacological interventions by making participant information sheets and consent documents appropriate for the target recipients [26,[30][31][32].
Interventions being re-named to make them sound more appealing and acceptable to potential participants, with a view to help with the recruitment rate for the intervention [30]. Two studies also involved PPI members in data collection as co-researchers [24,32].
Studies that reported PPI in analysis and interpretation reported that PPI generated new insights and endorsed researchers' interpretations of the findings [25,35,36]. For example, Stevenson et al. [25] described how researchers involved people living with dementia as co-     PPI group perspectives on review findings As described earlier, we consulted a PPI group of people living with dementia and care partners to discuss the findings of the scoping review. They commented on the various PPI approaches identified in the review and felt that standardising PPI approaches might be useful to improve the quality of PPI in research. They felt a more standard approach to PPI could help people living with dementia understand what was expected in terms of their PPI role. The PPI group suggested that PPI approaches should consider the needs of participants, and should reflect on how best to involve particular participants to facilitate meaningful input. They emphasised that PPI input should be obtained at an early stage in the research process to ensure that patients and care partners are involved in research priority setting and identifying research outcomes that are relevant to them. The PPI group felt that it was important to be involved in all stages of research to ensure that the relevance of the research is maintained throughout, including the design of the study materials, facilitating participation in the research and identifying dissemination methods to reach relevant audiences. The group expressed a preference for PPI approaches that involve small group meetings to ensure that the discussion is focussed on key issues and also so that individuals could be more closely supported to provide input. With respect to the general lack of formal evaluation of the impact of PPI, the PPI group suggested that this could be due to researchers not thinking far enough ahead in planning PPI input to their research. They thought it was important that PPI should be acknowledged in published materials to underpin the value of PPI input in research.

Discussion
How PPI with people living with dementia is being carried out in dementia research in the European Union To our knowledge, this is the first review to examine the types and impact of PPI in dementia research in the EU. There were increased reports of PPI activity in dementia research within the UK from 2012 onwards, although there were only a few published reports of PPI in dementia research from the EU. The range of terms used to refer to PPI by authors in the selected studies in this review was varied. The terminology used to refer to PPI may be based on different understandings and different objectives for PPI in each study. Use of different terminology for PPI may also reflect debate about the meaning and types of PPI appropriate for different research contexts [37][38][39]. The studies identified in this review also used various different approaches to PPI (see Table 4). Approaches may vary according to the aims of the specific PPI. Conversely, the various PPI approaches posed challenges for this review in terms of (i) systematically describing PPI approaches and (ii) identifying which approaches are effective or appropriate for particular research objectives. Despite variation in terminology and approaches, all of the PPI reported in this review related to research involvement 'with' or 'by' members of the public rather than 'to', 'about' or 'for' them [1]. The key principal of working 'with' or work 'by' members of the public [1] may serve as a starting point for future PPI in dementia research. Anecdotal evidence reported in this review suggests that a variety of PPI approaches are effective in facilitating involvement and impact on research. PPI is a rapidly developing area, and a variety of PPI approaches is likely to be appropriate due to variation in research study context, patient population and research objectives [5,[40][41][42][43]. Previous suggestions [44][45][46][47] for effective PPI included utilising PPI from the early planning stages of research by assisting researchers to identify different research topics, to adapt their research questions and subsequently also helps researchers design and conduct their research in a way that potential participants deem to be ethically satisfactory. PPI input was reported at all stages of research, including the initial proposal for funding [31], design [20-22, 28, 29] and data analysis and interpretation [25,36]. Whilst most of the studies incorporated PPI input into at least one single stage of the research process, 8 studies used PPI in all stages of undertaking and managing the research. Theoretically, PPI could occur at all stages of research. However there are challenges around maintaining long term involvement of members of the public in research projects [32]. The challenges in maintaining the involvement of people with progressive medical conditions such as dementia are particularly acute. Giebel et al. [32] suggested on-going recruitment of PPI representatives to ensure continued PPI in long term research projects.
The impacts of PPI on people living with dementia and the public, dementia researchers, and the research process within the European Union PPI is suggested to increase the cost effectiveness of research by ensuring that research outputs are appropriate to the patient group of interest [48]. In this review, just three studies included a formal evaluation of the impact of PPI, with 17 studies reporting anecdotal impacts of PPI. The reported impacts of PPI depended on the objective of the PPI and stage of research that PPI was conducted (e.g. in setting research priorities versus research design).
The lack of formal evaluation and the different approaches that were used in studies that did attempt to formally evaluate the impact of PPI made it difficult to establish whether one approach may be more effective than another, or even whether PPI in dementia research does deliver consistent benefits. Other reviews of PPI in health research have previously identified a lack of evidence for the impact and benefits of PPI, and where studies have reported the impact of PPI, the quality of the evidence is low [2,5,49].
A review by Boote et al. [50] identified that inadequate resources are allocated for monitoring and evaluation of PPI impact and PPI impacts tend not to be systematically recorded during the process of PPI. There is a need to substantially improve the evidence for the impact and cost effectiveness of PPI [51,52].

Strengths and limitations of the review
As only abstracts written in English were included, some relevant EU studies may have been missed. PPI activities may be under-reported in general [23,25], so some relevant PPI activities in the EU may not have been identified in this review. With the exception of the British Medical Journal (BMJ), most journals do not request information about PPI, so PPI activities may go unreported. Although BMJ have introduced the requirements for specific information about PPI, a study [53] found that only 11% of studies actually report PPI activity. No studies in this review used a standardised format for reporting PPI such as the GRIPP [15,54] which may have reduced the quality of reports of PPI. Lack of standards for reporting PPI made it difficult to extract relevant information from the papers.
A strength of the review was the involvement of a PPI group in interpreting the review findings. A very wide range of search terms was used to capture variation in terminology used to describe PPI. A large section of databases, including those indexing grey literature were systematically searched.

Future challenges and recommendations
Methodology and terminology used to refer to PPI varied. In different contexts, different words are used for similar activities, or the same terms were used interchangeably to refer to different things. A lack of standard terminology for PPI was identified in previous reviews of PPI [5,39]. Telford, Boote and Cooper [49] recommended that detailed accounts of PPI should be included in research reports and publications. Subsequently, substantial work has been undertaken to develop reporting standards for PPI in the form of the GRIPP [54], and in turn the shorter, revised GRIPP2 [15]. More robust reporting of PPI could guide future research in PPI, develop best practice for PPI [55] and reduce research waste via reducing ineffective application of PPI [56,57].
Perhaps related to the issues of varying terminology and approaches, a key shortcoming identified in this review was the lack of good quality evidence for the impacts of PPI in dementia research. A better understanding of the benefits and impact of PPI in dementia research would encourage researchers to embed PPI within research culture and provide an incentive for patients and members of the public to be involved. As good practise, effective PPI could be appropriately costed in proportion to the overall budget. PPI costs include staff time and expenses, PPI members' costs and expenses, administration, training for staff and PPI members, transportation, venue hire, and monitoring and evaluation costs [2,58]. It is critical that there be good quality evidence for the benefits and impacts of PPI in dementia research in order to convince researchers and funders for the need to appropriately fund PPI activities. Evidence of impact of PPI is not the number of patients involved in PPI, but should relate to the useful difference the PPI made to the research. Establishing an evidence base for PPI in dementia research requires formal evaluation of relevant impacts and systematic reporting of PPI [41,43,54]. Although some guidance for reporting of PPI is available [15,[59][60][61], the GRIPP 2 [15] reporting checklist provides one possible framework for describing PPI contributions to research. The GRIPP 2 [15] checklist has a particular emphasis on describing PPI activities. But a shortcoming is that the GRIPP2 has no standards for evaluating and reporting the quality of the PPI or for systematically quantifying impact of PPI on research. In the present study, we adapted the GRIPP2 by adding two sections to describe the method of evaluation of the impact of PPI (if included) and what the reported impacts of PPI were. Future iterations of the GRIPP2 should include evaluation of PPI impact.
The great majority of the studies identified in this review were carried out in the UK. Lessons learned in UK could inform methodological development of PPI in dementia research in EU. However, with Brexit there remains uncertainty about the UK's capacity to engage and collaborate in future research in EU, in addition to challenges posed by more restricted mobility of health care workers and researchers between the UK and the rest of EU. These challenges may be addressed by accelerating the development of the European Research Area [62], including reducing barriers to movement, a shared UK and EU research agenda, and development of European policies to overcome exclusion and promote participation of people living with dementia in research. UK researchers must persevere and continue to contribute to the discourse concerning PPI in dementia research in EU.
European charity and governmental research policies are beginning to advocate inclusion of PPI in dementia research [10][11][12][13]. PPI in dementia research may be further facilitated if peer reviewed journals, sponsors and funding institutions were to require PPI. The British Medical Journal (BMJ) released a 2014 guide for authors to report PPI in research published in the BMJ [63]. If no PPI activity took place, authors are expected to clearly report this. Adoption of similar guidelines by dementia research funders and peer reviewed journals may promote PPI in dementia research.

Conclusion
This review describes the various approaches to as well as the strengths, weaknesses and opportunities for PPI in dementia research in EU. The number of published studies reporting PPI in dementia research is growing, reflecting recognition of the importance and the feasibility of PPI in dementia research. A variety of PPI methodologies were used at all stages of the research process, all of which may be appropriate in different contexts according to the research question and aim to be addressed, the characteristics of PPI members and the resources available.
Variation in the terminology used to describe PPI and variation in the quality of reporting limited our capacity to interpret and synthesise the research. Due to the lack of a standard definition of PPI, there is a need to identify universal principles of involvement that can be implemented in dementia research. PPI in dementia research should be evaluated in relation to both the effectiveness of implementation and in accomplishing its objectives in informing research. Evaluation of PPI should be based on good quality research designs with rigorous standards of reporting in order to justify the effort and cost associated with PPI in terms of benefits to researchers, patients and the general public.