Patients as partners in health research: A scoping review

Abstract Background The role of patient involvement in health research has evolved over the past decade. Despite efforts to engage patients as partners, the role is not well understood. We undertook this review to understand the engagement practices of patients who assume roles as partners in health research. Methods Using a recognized methodological approach, two academic databases (MEDLINE and EMBASE) and grey literature sources were searched. Findings were organized into one of the three higher levels of engagement, described by the Patient and Researcher Engagement framework developed by Manafo. We examined and quantified the supportive strategies used during involvement, used thematic analysis as described by Braun and Clarke and themed the purpose of engagement, and categorized the reported outcomes according to the CIHR Engagement Framework. Results Out of 6621 records, 119 sources were included in the review. Thematic analysis of the purpose of engagement revealed five themes: documenting and advancing PPI, relevance of research, co‐building, capacity building and impact on research. Improved research design was the most common reported outcome and the most common role for patient partners was as members of the research team, and the most commonly used strategy to support involvement was by meetings. Conclusion The evidence collected during this review advanced our understanding of the engagement of patients as research partners. As patient involvement becomes more mainstream, this knowledge will aid researchers and policy‐makers in the development of approaches and tools to support engagement. Patient/User Involvement Patients led and conducted the grey literature search, including the synthesis and interpretation of the findings.


| BACKG ROU N D
The involvement of patients in health research is an emerging phenomenon, with critical barriers still present for researchers seeking to involve patients as partners. 1  These organizations aim to support the involvement of patients, their caregivers and families in the research process by promoting and providing infrastructure to support PPI in health research.
As such, the role of patients and how they engage in research has evolved over the past decade. CIHR defines this new role as "patient partner" and describes when patients contribute to the research process and research-related activities, different from the traditional, more passive role, as research participant, where patients receive treatments or are under observation. 5 Patient partners are more involved in conducting research activities, at all stages of the research process, including supporting grant applications, assisting with participant recruitment and performing research dissemination activities. 6 In Canada, the SPOR patient engagement framework encompasses four principles to guide the involvement of patients as partners in research. These include the following: (a) ensuring inclusivity, so that perspectives of patients who may otherwise be excluded, marginalized, hard to reach or are members of minority groups are invited; (b) providing support through flexible environments with training opportunities that will meet patients' diverse needs; (c) promoting mutual respect by valuing patients', researchers' and clinicians' perspectives through regular discussions; and (d) co-building research projects by identifying priorities and gaps and working to implement solutions together. 7 Equally important in supporting patient involvement, this same framework emphasizes six outcomes for patient engagement including (a) multi-way capacity building, which ensures that the support needed for patients and researchers to be able to work together are in place; (b) inclusive mechanisms and processes, which supports an environment where patient engagement can occur at all levels; (c) multi-way communication and collaboration, which describes a safe environment of mutual respect; (d) patient-informed and directed research, which discuss supportive research approaches that engage a diversity of patients; (e) experiential knowledge is valued as evidence, recognizing the importance of lived experience and the ability of this knowledge to be mobilized and translated; and (f) a shared sense of purpose, where participants, both patient and researcher, work together towards a common goal. 7 Bird et al 1 conducted a scoping review exploring the impact of patient partners on research outcomes, acknowledging the results of this review primarily involved adults with chronic, long-term conditions, a number of critical barriers and facilitators for researcher seeking to undertake patient partnerships were identified. Hoekstra et al synthesized the research partnership literature, and recommended a four step guidance to support partnerships processes including (a) building and maintaining relationship; (b) determining the appropriate level of engagement for each phase of the research process; (c) selecting or adapting strategies to the desired level of engagement; and (d) communicate and report the principles, strategies, outcomes and impacts of the research partnerships. 8 Broadly, there is an understanding of researcher motivations to engage patients in the research process, such as improving the relevance of research, 9 and an understanding of the motivations of patients who are involved as partners, such as improving health care. 10 Systematic reviews on the involvement and engagement of patients in research have not differentiated between the evolving role of patient as partner in research and patient as research participant. 9,11 Currently, there is little peer-reviewed published literature that differentiates the role of patient as partner in research from the role of patients as research participant. 12 Although PPI has been shown to develop on-going, productive and mutually advantageous relationships between researchers and patients, 13,14 the role of patient as research partner is not well understood. While the evidence provided by Manafo et al 15 describes the characteristics of patient as partners, we conducted this review to understand the engagement of patients who assume roles as partners in health research projects. Since "patient as partner" is defined by individuals who assume roles with greater involvement, increased sharing of power and increased responsibility on research teams, 16 we focused the review at the higher levels of the engagement spectrum (involve, collaborate and lead/support) as defined by Manafo et al 15 The specific objectives of this review were to (a) report the purpose of engagement; (b) report the outcomes of engagement; (c) to capture the patient role as described by the study author; and (d) to explore the engagement strategies to support patient involvement.

| ME THODS
A scoping review was conducted because this body of literature has not yet been comprehensively reviewed, it is heterogeneous in nature and therefore not amenable to a more precise systematic review.

| Design
We followed the methodological framework proposed by Arksey

| Recruitment of patient partners
Posters were distributed among the first authors' personal networks, and other organizations including the Alberta SPOR Support Unit 19 and Albertans for Health Research. 20 The project description, proposed role and anticipated time commitment were included. Five individuals responded to the opportunity. All five individuals were interviewed and later selected. These individuals represented a diverse group in terms of diversity, age and sex and had varying experience working on research projects (from no experience to some experience).
Ethic approval was not needed as these patients were members of the research team rather than participants of the study. The details about the engagement process are described somewhere else. 21

| Stage 1: Identifying the research question
Our research team, including our patient partners and key stake- and was applied to the MEDLINE and EMBASE databases. Search terms were determined with input from the research team, research collaborators and our team of patient partners and included a broad combination of title and keyword search terms such as "patient", "consumer", "carer", "family" and "community" and mesh headings such as "Patient Participation" and "Health Services Research" (Appendix S1). A team of patient partners led and conducted the grey literature component of this review. Patient partners identified studies, reports and conference abstracts of relevance to this review. To ensure that all relevant information was captured, we modified the CADTH Grey Matters 24 tool for searching grey literature to include an exhaustive list of organizations with a mandate in the area of patient engagement such as INVOLVE in the United Kingdom and SPOR in Canada (Appendix S2). Two training sessions on developing and conducting a grey literature coupled with weekly discussions and support meetings were held with the patient partners. A custom Google search was conducted using a comprehensive list of search terms such as "patient participation", "public involvement" and "community integration" (Appendix S3). Finally, we handsearched reference lists of identified reviews to identify additional studies of relevance.

| Selection of sources of evidence
The review process consisted of two levels of screening: (a) a title and abstract review and (b) full-text screening to assign studies to a level of engagement. For the title and abstract review, the team calibrated on the first 100, until a high level of agreement between reviewers was reached. Four reviewers, in two teams of two, independently screened all title and abstracts. When determining eligibility at full-text screening, only studies that described the higher levels of Patient and Researcher Engagement Spectrum were retained such as at the involve level, which involved patients as members of an advisory group, the collaborate level, which described partnering as a team member; or the lead/support level which described patients leading research activities. 15 Thus, studies underwent an additional screening process where two independent reviewers assigned each study to a level of engagement according to the authors' report of the engagement activities. 15 Articles reporting more than one level of engagement method were assigned to the higher level. The level of engagement of included articles was confirmed during the data extraction process, and any disagreements were resolved by a third reviewer.

| Data charting
The same data collection instruments for both the grey and academic sources were developed a priori by the research team to confirm study relevance and to extract study characteristics. Study characteristics were recorded in tabular form (using frequencies/ Data extraction was completed by three reviewers, one of which was a patient partner who extracted data from all included sources from the grey literature search. All data were verified by a fourth reviewer. The data were compiled in Microsoft Excel. Weekly team meetings were conducted to discuss articles and maintain reliability and quality during the data extraction process.

| Stage 5: Data summary and synthesis
We used the six stages of thematic data analysis as described by a shared sense of purpose, which is described by participants working together towards a common goal. 7 Next, the patient role, as identified by the study author, was captured. Lastly, the strategies used to support patient partners during their engagement, such as focus groups and face-to-face meetings, were determined using the word frequency query in NVIVO 12.

| Stage 6: Consultation
According to the Scoping Review enhancements suggested by Levac, 18 two consultation meetings were held with the patient partners, members of the Alberta SPOR Patient Engagement Platform 26 such as the Patient Engagement Research Lead, and the broader research team including students and academics. The first meeting occurred at the beginning of the project where attendees reviewed and provided comments to the study question and the overall search strategy including the search terms. The second consultation meeting occurred to gather input from stakeholders and the research team members on preliminary findings to provide context and thoughts to inform the potential implications from the review. Participants were notified of both consultation meetings one week prior by email.   Table 1 shows the majority of the included records are examples of patients who assume roles as patient partners, (n = 61, 51%) with the most common occurring at the "Collaborate" level (n = 60, 50%). Patient partner characteristics such as age, gender and ethnicity were infrequently reported. See Appendix S5 for a detailed patient partner data including from included studies.

| Reported outcomes
Using the CIHR strategy for patient-oriented research patient engagement strategy as a lens, outcomes were categorized according to one of the six components of successful patient engagement. These included the following: (a) Multi-way Capacity Building (n = 25, 18%);

| Type of strategies used to support patient partner engagement
Included studies reported using multiple engagement strategies to support patient involvement. Using the word frequency query in NVIVO 12, the most common strategies used to support partner engagement were meetings (n = 37, 31%), followed by research activities where patients were involved in conducting or analysing data such as during focus groups (n = 29, 25%) and interviewing (n = 22, 19%). Regular communication strategies such as email (n = 8, 7%) and teleconferences (n = 8, 7%) were also used ( Table 3).

Frequency n (%) Description
Documenting and advancing PPI 59 (39) This theme described studies where researchers involved individuals in order to customize the engagement strategies to best support their projects and so to contribute to the evidence base by describing the process used such as Berg et al. 29 who used flexible methods to capture the knowledge and experience of participants Although Bird classified patients into one of the levels as defined by the Patient and Researcher Engagement framework, 15 we classified studies according to the "higher" levels of engagement (collaborate, involve and lead/support) a decision supported by the literature. 15,16 We recognize this could imply a hierarchy between the levels of engagement; however, we did not find a notable difference in the study outcomes or type of engagement activities outlined between these three levels. This is a novel finding because often the argument is made that in order to have the greatest impact on a study, patients should assume greater responsibility as co-researchers. While categorizing study outcomes, two additional outcomes, not described by  Other research activities 4 (3) engagement approaches in clinical trials 149,150 and data-intensive health research 151 require additional investment in training and capacity building for both patients and researchers.

| Strengths and limitations
This study has strengths and limitations. Due to the comprehensive scope of the search, the volume of studies required us to limit to two academic databases, limit the publication dates of the included studies and limit our search to English language peer-reviewed publications. These decisions were made to maintain the feasibility of the project. These limitations were balanced with a robust grey literature and hand-searching strategy. Despite this, it is possible that a search without these imposed limitations may have yielded additional studies of relevance to this review. We identified the "higher" levels of engagement as a source of examples to describe patient as research partners. We believe our findings support this decision but suggest additional research be undertaken to further understand the similarities and/or differences between roles providing greater involvement, increased sharing of power and increased responsibilities and those that do not. 16 When assigning studies to one of the six levels of engagement, as described by the Patient and Researcher Engagement spectrum, 15 we assigned the level of engagement based on what was reported by the manuscript authors and acknowledge it is possible some studies were incorrectly assigned. We believe the duplicate review of each study and the addition of a third reviewer when consensus could not be reached minimized potential errors. Further, the patient partner-led grey literature search and data extraction may be perceived as design bias by some. However, we feel that the robust methodological processes we developed to conduct this review minimized any potential for bias, while supporting greater understanding and confidence among the patient partners.

| CON CLUS ION
While significant research exists that highlights how researchers are involving patients in health research, the engagement of patient as research partner is not well described or understood. Our findings suggest, with much research focused on the mechanisms of how and theoretical frameworks of why to engage patients, the linkage between purpose of engagement and study outcomes are evident. The data set also enabled a greater understanding of the role of patient as partner in health research. Creating opportunities for the involvement of patients as partners, in all aspects of research, and across research designs, helps researchers and patients in building a critical mass for change on an individual and an organizational level.

ACK N OWLED G EM ENTS
The authors would like to thank the additional patient partners who contributed to the collection and analysis of the data in this review: Melissa Hartwell, Char Vanderhorst and Sophia Khan. We would also like to thank the Health Technology Assessment Unit at the University of Calgary for their guidance and support during this Review. We also wish to thank Farwa Naqvi for assistance during the review.

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