The need to meaningfully engage individuals with lived experience (ie, patients, family members, caregivers, clinicians, researchers, and other advocates; henceforth referred to as knowledge users) in the conduct of health research—defined as research that includes clinical and basic medical sciences, such as care-based research, systems research, and preventative research—is being increasingly recognized by the scientific community. In particular, it is recognized that these individuals should be included at the onset of the research process, with the aim of developing research that meets the needs of individuals with lived experiences [1]. In fact, the lack of involvement of these individuals in such research priority setting has been identified as a key contributor to difficulties in effectively translating research findings into clinical practice and policy [2].

In parallel, the use of social media—defined as any web-based platform or mobile app through which users can engage with others—is gaining considerable traction within the research community, as researchers increasingly access Facebook, Twitter, and YouTube to support participant recruitment and other research activities [3]. The benefits of research-related social media use include enhanced connectivity between researchers and participants and the potential for rapid diffusion of scientific knowledge to target audiences [4]. The nature of web-based survey methods may also enhance anonymity for participants within the research process, potentially promoting the collection of more valid data [5]. Particularly, data collected via the web may be less vulnerable to contextual biases that can arise in focus group settings or when researchers administer surveys in-person [5].

In light of such potential benefits, a growing body of literature describing the use of social media to elicit and prioritize research uncertainties from knowledge users is emerging [6]. However, there remains sparse methodological guidance on how best to conduct social media efforts and their corresponding effectiveness in developing knowledge user–built research agendas [7].

Through this knowledge user–driven scoping review, we aim to identify studies that implemented and evaluated social media campaigns that promote participation in setting priorities for health research to address three overarching research questions:

An internal protocol was developed for this review. Our reporting process was conducted in accordance with the PRISMA-ScR (Preferred Reporting Items for Systematic Reviews and Meta-Analyses extension for scoping reviews) guidelines [8].

A comprehensive search strategy was developed in consultation with a tertiary hospital librarian (LR). We conducted tailored searches in PubMed, Embase, Cochrane Library, Scopus, and Web of Science. We searched all databases from their inception to September 14, 2021. Multimedia Appendix 1 shows the search strategy. Intradatabase and interdatabase duplicates were removed electronically. Using Covidence (Veritas Health Innovation), titles and abstracts were screened independently by 2 trained authors (KH and SS) according to our eligibility criteria. In cases of conflicting opinions on eligibility, studies were moved to full-text screening. Full-text articles were then screened independently by 2 authors (KH and SS). Any eligibility disagreements were resolved by consensus through discussion by at least three authors (AC, KH, SS, and LJ). The reference lists of relevant studies were also scanned to find other applicable papers.

We included studies (1) that discussed strategies to promote social media–based health research priority setting among key stakeholders and knowledge users and (2) measured the effectiveness of such strategies directly or indirectly. There were no restrictions on the language, country, and year of publication, nor the research content focus, as priority-setting research is cross-disciplinary. Although no explicit restrictions were placed on the language, the included studies were dominated by English language–based social media campaigns. We defined social media as any web-based platform or mobile app through which users can interact and engage with others. We defined knowledge users as patients (or potential patients), caregivers, clinicians, and other advocates (eg, health researchers). We excluded (1) studies where the purpose of the social media campaign did not include knowledge user engagement (eg, social campaigns used to disseminate smoking cessation information to knowledge users) [9]; (2) studies where the research prioritization campaign did not involve social media (dissemination techniques solely involved telephone calls, flyer distribution, etc); and (3) abstracts, dissertations, protocols, systematic reviews, scoping reviews, or case studies.

A standard electronic data collection form was created and piloted with our group, after which data extraction occurred independently (KH and SS). Discrepancies between the collected data were resolved through discussion with 3 authors (LJ, SS, and KH).

We used descriptive statistics to summarize quantitative study data and an inductive thematic analysis to synthesize qualitative data [10]. Our data collection form was uploaded to NVivo (version 12.6.0; QSR International) for analysis and was read through multiple times by 2 authors (KH and SS) who had previous experience with thematic analyses. One author (SS) then coded qualitative text within the table on a segment-by-segment basis. At frequent meetings, a second author (KH) reviewed the coding decisions using a constant comparative approach adapted from Thorne [11]. As a group, we (KH, SS, and LJ) then collapsed these codes into subthemes and themes based on the between-code relationships and in accordance with our research questions.

Figure 1 outlines our study identification process. Overall, 23 papers reporting on 22 unique studies were included in this review. The number of published studies increased steadily over time until 2020, which was the last complete publication year (Figure 2).

Included studies were conducted in 46 countries, most commonly in the United States (11/23, 48%), the United Kingdom (7/23, 30%), and Canada (5/23, 22%). Studies described participation by 13,640 individuals (median 332; range 31-4601), with sample size data missing from 4% (1/23) of the studies. Across studies, the median percentage of female participants was 77.28% (7404/9581). Sex data were missing from 52% (12/23) of the studies. Age data were variably reported and missing from 57% (13/23) of the studies; therefore, data were not collated. Sex data were missing from 39% (9/23) of the studies. Included studies used a variety of social media platforms to gather research priorities, including websites, emails, Facebook, Twitter, e-newsletters, web-based flyers, Survey Monkey, ExpertLens, blogs, YouTube, Choicebook, Instagram, WhatsApp, Snapchat, and web-based forums. The most common platforms used in the included studies were websites (12/23, 52%) and Facebook (9/23, 39%). The median length of a study’s social media campaign, when reported, was 3.5 months (range 1-24 months). Table 1 summarizes the characteristics of the included studies.

Table 2 shows the particular social media strategies used to enhance knowledge user engagement in research priority-setting exercises grouped by platform. Of studies using email as their primary social media platform [12,16,17,19,23,24,27,29,30,33] study teams emailed messages with embedded research prioritization survey links (including to researchers’ existing mailing lists) and integrated tell a friend tool in emails to prompt recipients to invite colleagues to participate. Facebook-specific methods to engage stakeholders included embedding survey links within Facebook posts, using the platform’s boosting feature (ie, paid advertisements), and hiring a Facebook advertising specialist. Informational Facebook pages were also used and involved private and public question-and-answer pages and a resource center with links to relevant documents [5,17,14-18].

Twitter-specific methods to engage participation included the use of hashtags within tweets and question-and-answer threads for prospective participants [5,14,15,17-19,25,31]. In addition, Salmi et al [27], hosted live chats on Twitter, in which host Twitter accounts tweet about predefined topics with questions during a set period, to which Twitter users respond via tweets and engage in discussions with each other. A web-based forum strategy led to the creation of a space where families and researchers could share ideas on the priority-setting research project [31]. Informational videos were created and hosted on YouTube for people potentially interested in contributing research priorities and were later posted on other platforms [28,29]. For studies involving blogs, researchers posted stories and internal updates related to the project to enhance interest in participation [28,29]. Studies also distributed e-newsletters to existing networks, sending them monthly to promote participation [16-18,28,30]. In addition, several studies used posts on Reddit and websites and web-based connection with the research team through video-calling platforms (eg, Skype, WhatsApp, or FaceTime or video chat on Facebook Messenger) to promote participation in priority-setting research [22,28,32].

Table 3 summarizes techniques to disseminate actual web-based research priority-setting surveys through social media. Snowball recruitment, in which current participants’ friends and family were approached for participation, was used [14,15,29,30]. Study teams also provided partner organizations with toolkits, templates, and promotional materials [5,12,13,15,17,23,29]. Then, organizations could use these materials to support the broadcasting of participation opportunities through social media. Individuals embedded in research prioritization exercises, such as steering group members, were additionally asked to promote the participation opportunity to their networks via social media [12-16,19,23,30], including by providing such individuals with preworded statements to tweet [19].

Across all the 23 included studies, 21 (91%) claimed to be successful in conducting health research priority-setting exercises via social media–based methods.

The direct effect of social media campaigns in securing stakeholder participation in research priority-setting was assessed as the (1) number of survey responses [12,14,15,20,33], (2) number of survey responses within a set period [14,15,20,33], (3) proportion of surveys fully completed [21], and (4) number of visits to external survey administration sites [14,15].

Indirect metrics for campaign effectiveness were (1) audience reach (ie, extent to which the survey sample was characteristic of the target population [13-15], number of countries and local communities represented in the sample [12,21], and number of national associations and external organizations contacted [12]); (2) campaign interaction (ie, number of clicks and impressions on posts [14,15,18,23,25,27], frequency of post views [26], volume of comments left by target stakeholders [26], number of searches for campaign pages or downloads of resources [17,18], number of bespoke hashtag clicks or uses [25,27], and Google Analytics [18]); (3) participant satisfaction [17,28,31]; and (4) platform-specific methods (ie, number of website views or likes [12,14,15,17-19,21,29], number of registered participants in an email chain or total number of delivered emails [12,13,16,19,21,29,33], new followers and likes on Facebook pages [14,15,17,18,26], and Twitter followers gained [14,15,25]).

Recognizing the importance of engaging key stakeholders in developing research agendas, we sought to use the extant literature to understand how social media might support research priority-setting, how effectiveness of the method might be measured, and the method’s benefits and drawbacks. We show that multiple social media strategies, which differ depending on the social media platform, have been used to promote participation in research priority setting—with strong success rates in generating research agendas. Metrics to quantify the reach of these strategies included the number of impressions on posts (eg, likes and other reactions) and the volume of comments left by stakeholders. In addition to the benefits, limitations of the use of social media in research priority-setting were also identified. Results from this review can guide methods for research priority-setting by patients, family caregivers, health care professionals, and other advocates and support the engagement of these stakeholders in developing future research agendas.

Social media–based strategies that incorporated platform-specific amplification (eg, Facebook boosts) and components that encouraged active engagement by participants (eg, question-and-answer forums and shared resources) enabled researchers to reach a broad audience of possible participants. This finding agrees with the literature showing that Facebook [34] health promotion posts receiving a paid boost reached significantly more users. Hashtags were also used in the included studies to increase visibility of tweets, which aligns with previous research showing hashtag use as effective in influencing social media conversations related to mental health [35] and in cases where the desired participant pool is small [36].

Our finding that snowball sampling is used to disseminate priority-setting surveys and expand participant pools aligns with other research showing that options to like, tag, or share posts expand a social media campaign’s reach [37]. This method may be particularly advantageous in cases where the campaign target audience is a specific and relatively small group (eg, people with lived experiences of less common diseases) and campaign participants may have contacts within their social network who they can engage in the process. Our results also suggest that there are priority-setting advantages in asking relevant external organizations and internal research and clinical team members to circulate survey links and use their personal or organization-affiliated social media accounts to expand reach.

We identified several metrics used by researchers to evaluate the effectiveness of social media campaigns, including the number of post impressions, frequency of viewed posts, volume of comments left by stakeholders, and number of times a bespoke hashtag was clicked or used. The heterogeneity in metrics likely reflects the exponentially growing number of social media platforms. However, the collection and interpretation of these social media impact metrics support ongoing consideration of the campaign’s effectiveness and subsequent content adjustments to maximize campaign reach and engagement [35].

Commonly identified benefits of priority-setting via social media include the speed at which participation opportunities can be disseminated and the capacity to build a sense of community among participants—possibly enhancing engagement. Research has also indicated that social media may be particularly useful in targeting information at some rarely reached groups such as individuals with depression [38]. In addition, moderators might humanize the campaign, build possible participant’s trust, and enhance campaign engagement by these individuals [39].

In contrast, limitations of social media–based methods for priority-setting research include the uncertainty of who is being captured through the posts [40]. Our study found that researchers commonly cite fears that social media–based methods may unexpectedly include or exclude the research priority perspectives of certain groups. In these cases, there are limited ways to assure that the recruited team of participants is the valid group of people that will render reliable results. This is problematic from ethical and methodological (ie, sampling bias) points of view and its mitigation requires careful planning. Moreover, when survey links are disseminated via social media, the true number of individuals that are reached cannot be calculated. This is because not all users will engage (ie, like, comment, and share) with the post [13,20]. In addition, although the platform analytics (ie, number of follows, comments, and likes on posts) are often used as an indication of survey engagement, these data may not be representative of the sample that opens the survey link or completes the survey.

Recommendations were also made to establish a consistent tone and branding, with a focus on using attractive graphic designs within priority-setting research campaigns. This consistency may increase the recognizability of the research project and authenticity to the effort, resulting in increased participation in priority-setting research efforts [41].

The definition of social media varies substantially in the literature and some definitions used did not meet our inclusion criteria. Our conclusions regarding the recruitment for priority-setting research projects may differ from those arising if a different definition was used. Varying definitions of social media may also have rendered our decision-making process during the screening phase susceptible to error. However, we screened in duplicate with good consistency and used third-party arbitration of discrepancies. Finally, amid the COVID-19 pandemic, the number of studies adapting to web-based research methodologies, especially using social media, may have increased after the search strategy was performed. Considering such rapid growth, it is important to note that this review is a snapshot at a particular point in time that does not account for novel methods that may have emerged after our search.

Social media appears to be an effective means to recruit and involve participants in the research process. Thus, researchers should consider using web-based social networking as a method to recruit knowledge users, collect data, and translate knowledge into practice. The study team’s efforts to build knowledge user trust in prioritization efforts, including by humanizing the campaign through moderating chats and engaging with participants, may improve engagement. On the basis of our findings, efforts can be supported by optimizing the visual representation of data through illustrative posts containing text and graphics. Moreover, to enhance participation by a wide group of knowledge users, researchers should focus on developing accessible and inclusive web-based materials. In addition, investing in platform-specific boosts (eg, Facebook boosts) and paid advertisements may be an effective tactic to enhance participant recruitment and enrollment.

Given the relatively recent emergence of digital platforms, social media–based methods are understudied compared with traditional recruitment means. We have identified some possible limitations of the method, such as potential limited access to individuals of lower socioeconomic status or older age. However, few studies have determined the extent to which these limitations impact prioritization efforts and, in the case of older adults, contrary evidence exists indicating good engagement with social media and technologies [42]. Should the identified limitations of social media–based priority-setting be significant, research into ways to mitigate these shortcomings is needed. Further research is needed to understand how to enhance the capacity of social media recruitment to capture representative samples. More research is also needed to understand which social media strategies and dissemination techniques are likely to be successful for research prioritization efforts, with the understanding that these strategies and techniques are likely to change over time as new social media platforms and features become available. Finally, given the highly public nature of information exchange on social media, considerations of the data privacy and security implications of social media–based research prioritization efforts are needed.

In this review, we synthesized the rapidly emerging data assessing the effectiveness of social media strategies to engage knowledge users in research priority-setting efforts across several social media platforms. The benefits of social media–based recruitment included the speed at which participation opportunities can be disseminated and the sense of community built among participants. As it is likely that social media–based research methods, including for research priority-setting, will be increasingly used by the scientific community, lessons and recommendations from this review can support scientists to more fully engage those who are most impacted by health research in setting associated research agendas.