Preparing for patient partnership: A scoping review of patient partner engagement and evaluation in research

Abstract Background Realizing patient partnership in research requires a shift from patient participation in ancillary roles to engagement as contributing members of research teams. While engaging patient partners is often discussed, impact is rarely measured. Objective Our primary aim was to conduct a scoping review of the impact of patient partnership on research outcomes. The secondary aim was to describe barriers and facilitators to realizing effective partnerships. Search Strategy A comprehensive bibliographic search was undertaken in EBSCO CINAHL, and Embase, MEDLINE and PsycINFO via Ovid. Reference lists of included articles were hand‐searched. Inclusion Criteria Included studies were: (a) related to health care; (b) involved patients or proxies in the research process; and (c) reported results related to impact/evaluation of patient partnership on research outcomes. Data Extraction and Synthesis Data were extracted from 14 studies meeting inclusion criteria using a narrative synthesis approach. Main Results Patient partners were involved in a range of research activities. Results highlight critical barriers and facilitators for researchers seeking to undertake patient partnerships to be aware of, such as power imbalances between patient partners and researchers, as well as valuing of patient partner roles. Discussion Addressing power dynamics in patient partner‐researcher relationships and mitigating risks to patient partners through inclusive recruitment and training strategies may contribute towards effective engagement. Further guidance is needed to address evaluation strategies for patient partnerships across the continuum of patient partner involvement in research. Conclusions Research teams can employ preparation strategies outlined in this review to support patient partnerships in their work.


| INTRODUC TI ON
Including patient partners in research holds promise for targeting patient-important research questions, creating meaningful change in patient outcomes and health systems, and realigning both research processes and outcomes to be patient-centred. 1 Internationally, efforts to grow inclusion of patient representatives on research teams have been driven by funding bodies, government and calls to action by patient communities. 2 A primary driver in the field of patient engagement is active collaboration with patients on research teams -including family, caregivers and friends -in governance, priority setting, conduct of research, and knowledge translation. 9 However, many researchers continue to struggle with how to operationalize research partnerships with patients, both realistically and effectively. We propose that partnerships require a shift in focus from engaging research participants in ancillary roles to more active roles, wherein patients engage as collaborative team members throughout the research process. While patient partnerships are often discussed, their impacts are rarely measured. 10 Previous systematic reviews 2,3,11,12 on the involvement and engagement of patients in research have not differentiated patient partners from consenting research participants who are engaged in research. A knowledge gap exists related to whether patient partnerships -as a subset of patient engagement -have an impact on research outcomes.

| Study aims
The primary aims of this work were as follows: (a) to identify the body of research where patients were included as patient partners in the research process; and (b) to determine whether the impact of patient partnership was examined and, if so, what impact the partnership had on research outcomes. The secondary aim was to describe specific partnership methods, barriers and facilitators across included studies in order to make practical recommendations about operationalizing patient partnerships.

| Guiding framework
The scoping review was guided by the Arksey and O'Malley Scoping Review Framework 13 which lends structure for identifying the research question and relevant studies, selecting studies, charting the data, and collating, summarizing and reporting results.

| Research question
Within the existing body of health research literature, how is the impact of patient partnership being examined and, when examined, what impact does partnership have on research outcomes?

| Search process
The search strategy was developed with an experienced information scientist in consultation with the research team. The search was undertaken in Ebsco CINAHL, and EMBASE, MEDLINE and PsycINFO via Ovid. No limits for language, date or publication type were applied. Using a combination of keywords and database-specific subject headings, the concept of community-engaged/patient partnership research (encompassing community-institution relationships and patient or consumer participation) was constructed as a concept for this search. This concept was then combined with language describing health services research or research design to form the basis of the search strategy. All databases were searched from database inception to October 2019. For a sample MEDLINE search, see Appendix 1 (other search strategies are available on request). Once duplicates were removed, citations were uploaded to a web-based program (Distiller SR) 14

| Inclusion criteria
The population of interest was defined as patients or their proxies including informal caregivers, family and/or friends who were considered patient partners during any stage of health research. Only studies where patient partners had not signed consent were included in this review as we interpreted the practice of consent to imply a research participant role versus a partner in the research process. All quantitative, qualitative or mixed methods peer-reviewed research articles from any health-care setting were included. Included articles described an evaluation of the impact of patient partnerships on research outcomes using validated tools for quantitative studies, or patient partner interviews or focus groups for qualitative studies. This applied examination of the impact of patient partnerships necessarily excluded studies where the objective was to develop a measure or tool to evaluate patient partnership, for example (see Appendix 2 for Screening Criteria).

| Selection of studies
At the initial screening level, both titles and abstracts of citations were screened by a single reviewer. Full-text screening was conducted independently by two reviewers, with agreement necessary for inclusion. Disagreements were reconciled through discussion prior to exclusion of studies.

| Data extraction
Data extraction from included studies was independently completed by two reviewers. All disagreements were resolved through consensus between reviewers. Study characteristics were extracted, as well as process for patient partner involvement, patient partner role/contribution, barriers and facilitators to partnership, approach

Sexuality and intimacy
Two community representatives were invited to participate in all aspects of the research that did not require additional training (eg recruitment, data collection, and data analysis). Community representatives were included in all study correspondence and meeting information, and took part in study discussions. To describe the process and impact of patient and public involvement and engagement in a systematic review of factors affecting shared decision making around prescribing analgesia for musculoskeletal pain in primary care consultations.

Musculoskeletal conditions
Five members of a patient Research User Group (RUG) collaborated with researchers in the review process. This was facilitated by a patient partner support team. RUG members attended workshops at three key points and were involved in discussion related to the research questions; factors important to patients; findings; and planning for dissemination. Patient partners also reviewed abstracts, presentations and publications, and gave presentations and contributed to discussions at conferences.

Rhodes et al, (2002), UK
To understand the experience of service users with diabetes who were part of a Users' Advisory Group for a project evaluating diabetes services in a city in the UK.

Diabetes
A service user advisory group met every 2-3 months over 2 years and provided input regarding the research process. Group members also participated in the larger steering committee that met twice a year. Feedback about participation was received through taped discussions with advisory group members.

Vale et al, (2012), UK
To evaluate the involvement of patient research partners in the conduct of a systematic review and meta-analysis from the perspective of patient partners and researchers.
Cervical cancer Six patient research partners were involved in providing feedback, locating study investigators, interpreting results, contributing to a study newsletter, providing input into a lay summary, and co-authoring an editorial. Patient partners and researchers completed short surveys with open-ended questions about their involvement. These responses were coded, and a summary report was sent to all involved. A final meeting was held to discuss the analysis and revise the summary report.

Patient partner role or contribution Level of engagement 7 Impact evaluator
Participants provided insights into the types of roles that patients and carers took on during the implementation process. This included sharing of experience, offering suggestions for change, and implementing possible solutions. Study highlighted that impact was on a smaller scale; however, partners provided ideas for change and possible solutions as well as motivated staff for possible changes.

Consult, Involve Unclear
Community representatives were included in discussion and decision making regarding plans for recruitment, thematic analysis, changes to study plan, broader dissemination of study findings, and future grant work. Stemming from a presentation at a half-day conference by community representatives, a sense of authenticity of study results was described.

Collaborate Research team
Participants contributed to the inclusion of a lay perspective for the research, offered practical viewpoints on the research, and acquired new knowledge and skills, confidence and personal support for their illness experience.

Participate, Consult
Research team and patient partners All patient partner representatives and researchers expressed an increase in their confidence in all described roles over time.
Co-applicants: Involve, Support, Collaborate Advisory Group Members: Collaborate, Involve Patient Groups: Collaborate, Involve External, independent Impact of patient partnership included establishing importance of review question, facilitating funding application, identifying additional important factors (leading to amendment of the search strategy and data extraction forms), developing a framework for narrative synthesis based on patient-identified categories, translating patient concerns into practice recommendations, prioritizing options for dissemination, identifying limitations in the review literature, and informing the next phase of research.

Consult, Involve Research team
Advisory group members felt that they had impacted the research process as well as had personally gained from the experience of being involved. Advisory group members appreciated being able to connect with other people with diabetes as well as being able to contribute to the community, eg provision of information about services to other people in their community with diabetes. Users contributed to the research process by reviewing research documents (interview guides, surveys) and also gave input regarding new topics for research. The advisory group provided local credibility and access to community networks.

Inform, Consult, Involve, Collaborate
Research team and patient partners The inclusion of patient partners led to researchers taking on another project related to this topic and an editorial on patient perspectives. There was consensus that patient partners brought a voice that would have been otherwise absent and helped to provide insights into the impact of cervical cancer on women's lives. There was also a sense that the issue of late effects would be explored in future trials given its prominence in the systematic review. To assess the impact of public involvement in the co-development of an assistive technology for people experiencing foot drop (using functional electrical stimulation (FES)) as a consequence of stroke.
Stroke (with foot drop as a residual side-effect) Co-design process within an assistive technology design study. A lay advisory group of ten people included those with experience with FES, family members, and community members. Activities of the patient partners included activities from conceptualization of the study to end stage dissemination. Advisory group met 9 times. Evaluation of lay advisor experience took place through audio recorded interviews at the beginning, middle and end of the project. A public involvement model based on INVOLVE was used.

Coser et al, (2014), Canada
To examine the process and personal impact of youth co-researchers who were involved in a participatory research project about factors that promote resiliency and prevent use of injection drugs for street-involved youth.

Street-Involved Youth with Injection Drug Use
Youth with first-hand experience with street involvement were contracted part-time for 12 months. Youth co-researchers were involved with facilitating focus groups, analysis, and dissemination of findings at academic conferences and community meetings, and were paid $15/hour for taking part in the project. Six youth co-researchers were interviewed at 3 and 7 months into a 12 month project about their experience. Field notes, meeting minutes, and debriefing sessions were also examined.

Revenäs et al, (2018), Sweden
To describe the experiences of stakeholders (people with Parkinson's disease, health-care professionals, facilitators) with co-designing an eHealth service intervention.

Parkinson's disease
Four co-design workshops were held to explore co-care needs of people with Parkinson's disease and health-care professionals. Participants included 7 people with Parkinson's disease, 9 health-care professionals, and 7 facilitators. Participants' feedback on what worked well and what could be done differently was collected on note cards. Facilitators' feedback was provided verbally while a researcher took notes. Researchers also wrote reflections in a diary. After the final workshop, a Web-based questionnaire was sent to participants to collect data on experiences with the workshops. To evaluate the impact of patient involvement on the process and outcomes of designing a new primary care clinic service in a large integrated delivery system.

Primary Care settings
Patients contributed to co-designing a new service in which a lay staff person connects patients with community resources. Twelve patient co-designers participated in a four-day design event and eight patient co-designers participated in a three-day 'check-and-adjust' event 15 months post-implementation. An interactive orientation session was held prior to the initial design event. Data sources included interviews, event observation and surveys.

McDonald et al (2016), USA
To explore the experiences of scientists and community members within a Community-Based Participatory Research-focused project related to violence victimization and health for those with developmental disabilities.
People with disabilities and violence victimization and health The project included a steering committee which provided leadership (5 scientists and 4 community members with developmental disabilities) and a community advisory board which also included 4 people with developmental disabilities. Interviews and focus groups were conducted to understand participant experiences. Interviews were completed in-person, over the phone, or written depending on the needs of the participant.

Patient partner role or contribution Level of engagement 7 Impact evaluator
The lay advisory group provided input into FES design as well as input into the forthcoming clinical trial of the assistive technology. Participants also reported feeling that they had made a meaningful contribution and a few had also gone on to take part in other research studies as a result of their involvement.

Involve, Collaborate, Lead Research team
Youth co-researchers identified feeling that participation positively influenced their identity, self-esteem, and sense of meaning for doing work. They felt that they had acquired knowledge and skills that would be transferable beyond the project. Some challenges were related to varying learning abilities of the youth and the need to adapt training and support to accommodate these differences.
Researchers noted that they needed to provide both training regarding research as well as support for personal lives given that difficulties that youth had experienced and continued to experience. Partners were engaged across eight possible study phases, from identifying research topics to disseminating results. Outcomes and measurement identification were the most common phase of engagement. Partner engagement influenced the selection of research questions, interventions and outcomes. Partner engagement also contributed to changes to recruitment strategies, enhanced enrolment rates, improved participant retention, more efficient data collection, and more patient-centred study processes and outcomes. Partners also participated in study conduct (recruitment, data collection, dissemination). More than two-thirds of investigators indicated that partners had at least a moderate influence.

Consult, Involve
Research team and patient/stakeholder partners Patient partners contributed to a more patient-centred service design, broader perspective in priority setting, contributed their thoughts and experiences of how intervention would affect patient lives outside of clinic, clarified where service should be located in clinic and contributed diverse community needs that may have been overlooked. Patient partners brought their own expertise and skills to the design activities and described a sense of personal growth (ie learning where to access care, learning new skills).

Consult, Collaborate Research team and patient partners
Involved with promoting accessibility and review of project findings. Project team members described developing skills, meeting new people, earning money, and contributing to the study process. Contributes from patient partners improved recruitment and knowledge translation efforts, and enhanced the communityacademic partnership.
Involve, Collaborate External, independent to evaluation, and 'stage' of patient engagement (Manafo et al 7 ).
Data were extracted using a standardized data extraction form. See

| Data analyses
Results were summarized using a narrative synthesis process consistent with Arksey and O'Malley's Framework. 13 Our approach to narrative synthesis followed guidance provided by Popay and colleagues 16

| PATIENT PARTNER CONTRIBUTI ON S
Across studies, 17-30 patient partner contributions were described with respect to roles enacted and direct involvement in study-related activities. The impact of partnership was largely reflected in terms of personal experience and gains.

| Roles and direct involvement
Patient partners in included studies enacted various roles within the research team. Table 2 provides an overview of these roles.

Patient partner role or contribution Level of engagement 7 Impact evaluator
Partners contributed to initial project idea development from a patient perspective, suggested changes for simplifying a patient survey, and presented results at research meetings. Caregiver advocates contributed their perspective in study meetings, assisted in data analysis and summarizing themes in study transcripts, advocated for policy changes through membership in wider groups, shared information through personal social media accounts. Research participants trained in research ethics certification, monitored and facilitated study calls between researchers and study sites, advocated for addressing school calendar, flu, and allergy season in asthma visits. Patient advisory board clarified materials for dissemination to patients, contributed to dissemination strategies.
Consult, Involve, and Collaborate Research team and patient partners development of research questions, 28 interview guides and informational materials, 19 as well as identification of research priorities, 21,28 and contributions to data analysis 28 and to regular study briefings. 22 In one study using a participatory approach, youth co-researchers interfaced directly with study participants by facilitating focus groups with street-involved youth. 21

| Impact on personal experience
In addition to the contributions to research processes described above, personal benefits of patient partnership were also noted. For example, patient partners described acquiring practical skills 19,21,25,29 (eg learning to use a computer 17 ) and gaining knowledge about research processes and various topics. 19,[21][22][23]29 Some patient partners reported that the process of collaborating with researchers helped them to gain confidence in identifying themselves as experts and advocates. 17,20,23 Patient partners also shared that participation in research afforded them a social network of supportive peers; some noted that relationships forged -both personal and professionallasted well beyond the study period 22,23 and that participation was a source of positivity (eg laughter) 24 Finally, patient partners articulated that adding the 'patient voice' to research projects and advocating for change was an empowering experience. 19,22,23 Because of these personal impacts, some studies reported that patient partners sought further opportunities to become involved in research as patient partners beyond study completion. 20,21

| PR AC TIC AL CONS IDER ATIONS
Practical considerations were addressed largely in terms of barriers and facilitators to patient partnership. A summary of lessons learned, barriers and facilitators to patient partnership, as reported by authors of included studies, can be found in Table 3.

| Barriers
Eleven studies cited specific barriers to partnership encountered during their work. 17,19,[21][22][23][24][25][26][27][28][29] These barriers were varied and encompassed study logistics, team characteristics and perceptions of patient partner roles by researchers, as well as partners themselves. One of the most common barriers identified was the use of jargon. 19,[23][24][25]29 In one study, both researchers and patient partners acknowledged that the nature of some discussion topics, such as prospective funding sources or ethics applications, made it difficult for patient partners to understand and follow what was being discussed. 24 Researchers in this study reported feeling pressured to limit their conversations to 'nonacademic' topics, in order to limit the use of jargon. 24 Other common barriers described included power imbalances between the researcher and the patient partners, 17,24,26,27 and the impact of time pressures on the research process. 19,[21][22][23][24]26,27 Other less frequently reported barriers included logistical hurdles, such as meeting disability accommodations, 17 as well as challenges with retention of patient partners in studies where patient partners experienced changing life circumstances or disease recurrence, limiting their ability to continue their participation in the project. 19,26 In some instances, partnerships were found to burden both patients and researchers with emotional tolls and burnout. For example, in one study patient partners experienced an emotional burden if confronted with the possibility of a recurrence of their disease while participating in patient partner activities. 19 With respect to additional challenges, two studies cited financial resource shortage issues, 21,26 one study referred to fearing patient 'tokenism', 17 and one study indicated group conflict 21 as key barriers to forging and maintaining partnerships.

| Facilitators
Studies also noted several facilitators to engaging in patient partnerships. One common facilitator cited was valuing the patient partner role, 17,22,23,29 which was accomplished when parties including the research leadership team, team members and patient partners themselves expressed the value and worth of patient partner contributions. Aligned with this facilitator, it was also identified that clear role descriptions and responsibilities for patient partners were important factors for facilitating effective partnerships. [25][26][27] For patient partners, this often meant the use of explicit policies or guiding documentation to support these descriptions. [25][26][27] Additionally, studies reported that meeting the personal needs of patient partners (ie disability accommodations, scheduling adjustments, provision of refreshments and transportation) helped to remove barriers to partnerships. 17,20,21,25,26 Other studies noted the critical role that compensation for time and work played in facilitating patient partnerships. 25,26,30 An atmosphere of camaraderie between researchers and patient partners, as well as among patient partners themselves, 21

| Evaluation of patient partnership
In included studies, patient partnership was predominantly evaluated by the research team itself. In four included studies, researchers themselves evaluated the impact of partnership, 20,21,24,26 and in seven studies, this was jointly evaluated with patient partners or other stakeholders. 19,22,23,[27][28][29][30] Only two of the included studies used an external, independent review process. 17,25 As this field continues to develop, there may be further opportunity to look to external and independent evaluation of partnership on both  • Involving and/or collaborating with patient partners required additional time, effort • Aspects of the research process may be upsetting for patient partners • Some decisions such as those related to outcomes in a systematic review are 'preset' based on that which is collected in the individual studies • Difficult to align the needs of patient partners and clinical/scientific collaborators • Recruitment via existing networks (eg through organizations or other volunteers) • Using a small group of patient partners • Providing information (eg a booklet) upfront that is aligned with patient partner training and support • Establishing and maintaining good working relationships among the team including on-going updates (eg delays in research progress) • Communicating the value of patient partners • Recommendations to facilitate patient partnership include: o Hosting an information session early in the research process that allows potential patient partners to attend and decide if they wish to be involved outcomes and process. This may reduce the risk for bias, as well as enhance those facilitators related to team cohesion, trust and roles/responsibilities. As others have articulated, there may also be a 'moral obligation' to include those impacted by research in its development. 31

| Barriers and facilitators to partnership
In our review, we found that a number of research teams were thoughtful about enfranchising patient partners in the research pro- Finally, our scoping review revealed one of the most common barriers across studies to be a lack of training for patient partners.
To resolve this issue, it is our contention that partnership requires thoughtful preparation of both researchers and patients to create an informed patient partner engagement experience. Aspects of preparation include clear education, role definition and guidance for both researchers and patient partners, 50 which may serve to maximize patient partner involvement and contributions to research outcomes, rather than limiting patient partners to only providing reflections of their personal experiences. 2

| LI M ITATI O N S
The literature included in this review largely reflects studies involving adults with chronic, long-term health conditions as patient partners in research. To this end, only one included study involved youth as patient partners in their work. Thus, findings may not be applicable to researchers seeking to undertake patient partnerships with youth, or outside of the area of chronic conditions.
Secondly, only studies that did not consent patient partners were included in this review. While this team acknowledges that in some instances, partnership and participant roles may overlap, our interest was specifically in those studies evaluating the impact of patient partnership when patient partners were part of the research team 'only', and not holding dual roles as a research partner and participant. As such, the considerable patient engagement and patient partnership literature that required patient partners to formally consent to be part of the research team was excluded. While it is likely that this literature would offer further insights regarding impact of patient partnerships, that would be a different focus than the present review.

ACK N OWLED G EM ENTS
We would like to thank Katherine Allan, Zil Nasir and Joanna Chu for their assistance in this study.

CO N FLI C T O F I NTE R E S T
The authors declare that there is no conflict of interest.