A Systematic Literature Review of Community-Based Participatory Health Research with Sexual and Gender Minority Communities

Purpose: The objective was to review sexual and gender minority (SGM) health research studies to gain an understanding of how the community-based participatory research (CBPR) framework has been operationalized. Methods: We used the Preferred Reporting Items for Systematic Reviews and Meta-Analyses guidelines to conduct a review of all SGM health research studies published in the past 10 years that cited a CBPR approach (PROSPERO Registration No. CRD42016036608). CINAHL, PubMed, and PsycINFO databases were systematically searched in October 2020. Dimensions of community involvement (e.g., shared decision-making; flexibility to community needs and priorities) and the strength of evidence for each dimension were rated using guidance from the Agency of Healthcare Research and Quality. Results: The 48 eligible articles identified reported a range of 0–11 (out of 13) community elements. Seven studies reported zero elements. Qualitative studies (n=28; 58.3%) had an average quality score of 2.32 (range: 1.43–2.5). The 15 (31.3%) cross-sectional studies had an average quality score of 2.08 (range: 1.64–2.27). Conclusion: Adhering to the CBPR framework is challenging. The benefits of striving toward its principles, however, can move us toward transformative and sustainable social change within SGM communities.


Introduction
Community-based participatory research (CBPR) is a framework built upon equitable collaboration among scientific researchers, community members, and other stakeholders to improve community health, reduce health disparities, and improve health equity. 1,2 This adaptive approach engages the community, recognizes and leverages the diverse strengths and contributions of all research partners, and is action oriented in that it seeks not only to understand problems but also to propose cocreated solutions. CBPR principles include the following: colearning between academic and community partners; capacity building and empowerment; mutually beneficial knowledge and findings; bidirectional leadership and decision-making; and long-term commitment. 3,4 CBPR shifts the traditional research paradigm which focuses on a specific set of research methods or techniques, largely developed in academic settings, by prioritizing the relationship between academic and community-based research partners and the creation of positive, transformative, and sustainable social change within communities. This represents a systematic effort to incorporate community participation, decisionmaking, and practices into the research practice. 5 Actively integrating community members during all phases of the research project helps to ensure that the methods used and data collected are culturally grounded and reflect the lived experiences of the population. 6,7 Community integration within research teams also has the potential to ensure that health research is acceptable and directly relevant to target communities, potentially improving the rigor of these research efforts overall.
CBPR has been used to examine a variety of health topics such as mental health, food insecurity, diabetes, homelessness, and HIV. It has also been used to assess community characteristics which are key to partnership sustainability, such as capacity, readiness, social capital, and empowerment. 8 It can be adapted for diverse community collaborations and may be particularly valuable for work with vulnerable and historically underserved communities. [9][10][11] Sexual and gender minority (SGM) is an inclusive term used to refer to a diverse array of people who are gay, lesbian, bisexual, transgender, queer (LGBTQ), gender non-binary or non-conforming, two-spirit, asexual, pansexual, intersex, and other sexual orientation and gender identities (SOGIs). National surveys estimate that *4-7% of the United States population (or 11 million people, roughly equivalent to the popula-tion of Ohio) are SGM but are likely underestimates given the failure of survey efforts such as the U.S. Census or American Community Survey to robustly collect this information. 12,13 Due to discrimination and social marginalization, SGM people are at risk for poor health behaviors and health outcomes. 14,15 Compared to their cisgender and heterosexual counterparts, SGM people experience higher rates of HIV infection and other sexually transmitted infections, smoking, drug and alcohol use, and mental health problems. 14 There is evidence that sexual minority women have higher odds of risk factors for hypertension, diabetes, and breast cancer. 14 Transgender adults may have higher risk factors for cardiovascular disease and myocardial infarction. 14 However, due to lack of data collection regarding sexual orientation and gender, the full extent of health disparities experienced by SGM people is not known. 14 Health disparities in SGM people are not caused by their gender or sexuality, but by the discrimination, minority stress, and sociopolitical barriers to optimal health that lead to exposures and behaviors known to contribute to disease and disability. 16,17 For example, SGM people face barriers to accessing basic health care across the life span, and many experience discrimination or refusal of service when seeking health care. 18 SGM people are also more likely than cisgender, heterosexual people to face employment discrimination and lack health insurance. 18 Even when care is accessed, health care providers are often underprepared to provide affirming health care to SGM people. 19,20 SGM communities are heterogeneous, and intersectionality must be considered. SGM individuals differ markedly by not only SOGIs but also in life experiences by age cohort, racial/ethnic group, socioeconomic strata, disability/ability, and immigration status. Integrating intersectionality into health disparity research emphasizes the need to consider this diversity in research, health care, and policy given its influence on an individual's health risks, screening behaviors, and treatment experiences. Intersectionality, a feminist sociological theory, considers the intersection of marginalized or minoritized identities 21 and how multiple oppressions coexist and interact on various and often simultaneous levels. Intersectional disparities among SGM have also been documented by older age, 22 lower socioeconomic status, 23 and immigrant status, 24 among others.
SGM communities have historical contexts for community-engaged research. ACT UP (AIDS Coalition to Unleash Power) was formed in response to social neglect, government negligence, and complacency of the medical establishment during the 1980s. 25 ACT UP has advocated for sustained investment in HIV/ acquired immune deficiency syndrome (AIDS) treatment and related coinfection research, equitable access to HIV/AIDS prevention and care, and tackling structural drivers (e.g., stigma, discrimination, and poverty) of the HIV/AIDS epidemic.
To make meaningful, impactful progress in SGM health equity, researchers must understand the influence of social determinants on SGM health, as well as the priorities and behaviors of this community, from their perspective. CBPR is well-positioned to more equitably include SGM communities in the pursuit of transformational research outcomes. 1,2,26 CBPR also has the potential to address noted gaps in SGM health research topic diversification. Although efforts have been made to establish a more comprehensive national SGM research agenda, HIV remains disproportionately prioritized, leaving gaps in other areas such as chronic disease and comorbidity, aging, methods and measurement, and social determinants of health (SDOH). 27, 28 It has been beneficial in the development of tailored assessment tools for specific populations 29,30 ; assessment of a wide range of chronic health conditions [31][32][33] ; and addressing the social determinants impacting the health of minoritized communities. [34][35][36][37] Although the incorporation of CBPR methodology into SGM health research likely has profound benefits, CBPR practitioners face a number of challenges which threaten to limit full implementation of CBPR principles. For example, uncurtailed community-academic partner power differentials, conflicting visions about the work, and limited structural support from funders and academic institutions may lead to minimal community engagement. 2,38,39 To effectively utilize CBPR to produce findings, knowledge, and outcomes with maximum benefit to SGM people, it is particularly important to characterize community involvement in the research process.
This systematic literature review contributes to the knowledge base by identifying gaps in implementation that limit SGM community members from becoming full partners, contributing their unique knowledge and experiences while also shielding their interests. We present the results of a review of peer-reviewed SGM health studies published over the past 10 years, which describe a CBPR or community-engaged approach, to gain a clear understanding of how CBPR has been operationalized in SGM communities across the United States.

Materials and Methods
This systematic literature review was conducted in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses 40 and has been registered with PROSPERO (Registration No. CRD42016036608), an international database of prospectively registered systematic reviews. 41 Articles were included in the review if they described original SGM health research studies conducted in the United States with mention of CBPR or community-engaged methods and were published between January 2010 and October 2020. ''Health research'' was categorized broadly, allowing for ''health adjacent'' topics such as SDOH. Editorials, systematic review articles, meta-analyses, case studies, and methodological articles were excluded.
We searched article titles and abstracts in the CINAHL, PubMed, and PsycINFO databases using keyword combinations specific to CBPR and SGM health research (Supplementary Appendix SA1). The date of the last search was October 12, 2020. Article screening, extraction, and assessment of CBPR criteria were conducted using Covidence Ò , a web-based software for management of systematic literature reviews. 42 Each of the articles included in the final review was independently assessed by two members of our research team with the reviewers meeting to discuss and reach consensus on discrepant items. We also met as a research team to ensure that data extraction elements and quality criteria were understood and applied consistently during the review process. Data extraction and assessment of CBPR criteria were conducted using the data elements recommended by Viswanathan et al. in their report on assessing CBPR evidence for the Agency of Healthcare Research and Quality (AHRQ) (Supplementary Appendix SA2). 43 Assessment of SGM community involvement comprised eight CBPR elements: (1) shared decision-making; (2) community participation barrier removed; (3) socioeconomic determinants of health addressed; (4) flexibility to community needs and priorities; (5) capacity building; (6) findings disseminated to participants; (7) findings applied to health-related intervention or policy change; and (8) intervention sustainability. 43 A quality assessment was also conducted to classify each article as good, fair, or poor quality (Supplementary Appendix SA3). 43 Quality of each element was rated on a scale of 1-3. One indicated that the article provided insufficient information or that element was poorly captured. A score of three represented the highest quality. A final score (range: 1-3) was calculated by averaging the scores of all relevant quality elements. Scores were assigned based on what was described in the article and, thus, may under represent elements included but not described in the study design.
The general SGM community was not engaged in the design or conduct of this systematic review. Insofar as the intended audience was SGM health researchers, our author panel comprised members of the Equitas Health Institute Midwest Health Research Consortium, who are all SGM health researchers.

Reflexivity statement
The authors of this article include gender diverse (cisgender, transgender, and/or non-binary) and sexuality diverse (queer, gay, and straight) researchers who are nurses, a physician, and public health experts in SGM health.

Results
Overview Forty-eight studies  were identified for inclusion in this review, as outlined in Figure 1. A search of the da-tabases revealed 454 nonduplicated records, 353 of which were irrelevant based on a review of the titles and abstracts. Of the remaining 101 relevant records, 20 described studies done outside of the United States, 11 were methodological articles which did not share the results of original research, 5 did not use CBPR/action research approaches, 5 described program evaluation only, 4 were dissertations, 4 were related to topics other than health (e.g., education), 2 were book chapters, 1 was a duplicate article not identified on initial screening, and 1 was unrelated to health. As seen in Table 1, there has been a steady increase in the number of articles published on this topic.

Study design
The included studies are described in Table 2. Most of the studies were qualitative (28; 58.3%) or crosssectional (15; 31.3%). Two reported the results of randomized controlled trials (RCTs), and another described a secondary review of survey data obtained during one of the aforementioned RCTs. All three of the articles sharing results from RCTs were done by the same CBPR partnership in North Carolina.

Study quality
Among the 28 qualitative studies included in this review, the average quality rating ranged from 1.43 to 2.5, with an average score of 2.32. Among the 15 cross-sectional studies included in this review, the average quality rating ranged from 1.64 to 2.27, with an average score of 2.08.

Study characteristics
Participants. Study sample size ranged from 7 to 763 participants ( Table 2). Three studies (6.3%) exclusively included older adults, although the lower age limit for ''older adult'' ranged from 40 to 60 years old. 47,67,90 Sixteen studies (33.3%) exclusively included the perspectives of people of color (POC); one of these recruited only Black men who have sex with men living with HIV, 55 and nine focused on the health of Latinx SGM individuals. In an additional eight studies (14.6%), White non-Hispanic participants were the minority.
There was a wide variety of SGM identities represented within the 48 included studies. As seen in Table 2, 32 studies (66.7%) included bisexual participants, although 9 of these had 5 or less bisexual participants. Ten studies (20.8%) included lesbian participants, but no study focused solely on experiences of these women. More than half of studies in this review (n = 33, 68.8%) included transgender/gender nonconforming (TGNC) individuals. Nine of these included exclusively TGNC participants and 10 included 5 or less TGNC participants.

Elements of community involvement reported
Within the reviewed CBPR studies, a wide range of community involvement was reported. The distinction between academic and community partners was not always precise because some descriptions of community partners included academic partners with SGM identities. The number of community elements incorporated ranged from 0 to 11, with seven studies not specifically describing any of the AHRQdetermined community involvement elements in their article. The study with the most community elements was a qualitative study examining health services among older adults. 66 Examples of elements were: selecting the research question; developing the research proposal; collecting the data and other implementation activities; interpreting, disseminating, and applying the findings.
The most commonly reported community involvement elements were recruitment (56.3%) and study design (52.1%) (Table 3). Furthermore, more than half (54.2%) reported community partners assisting with interpretation of the findings. The research question of 21 projects (43.8%) was selected in partnership with community members, but only 3 (6.3%) study proposals were developed with community partner input. A small majority (4.2%) described community partners assisting with retaining the participants. No study reported shared financial responsibility with community members.

Discussion
Despite wider recognition of the value of communityacademic partnerships in SGM health research over the past decade, this work remains sparse and adherence to the CBPR framework is variable given inherent challenges and an understanding that a single set of principles may not be appropriate for all communities and all contexts. 4 On the continuum of communityengaged research, CBPR moves beyond community involvement and collaboration and signifies the highest level of engagement. It is hallmarked by shared leadership and a strong and sustained partnership that ideally integrates the community into each phase of the project. 92 Of the 48 studies identified in this review, none met this standard. No study included all 13 community involvement elements identified in the 2004 AHRQ report, and 7 studies included none at all. 46,49,50,53,[60][61][62] Of note, only those elements described in the articles were assessed, and it may be likely that limitations such as article word limits may account for incomplete descriptions of methods and community involvement. Among the four studies with the greatest number of elements, the quality varied, with no study achieving an optimal assessment score. 60,69,81,83 Considering all studies, community involvement was disproportionately concentrated in the study design, recruitment, instrument development, and interpretation of results. In most cases, the community did not select the research topic and/or the research question, an activity essential to creating a shared vision and ensuring the project is community driven. Excluding the community at this critical point can limit identification of priorities and outcomes most important to them and instead promote the researcher's agenda and expertise or the funder's priorities. 2 Diminishing bidirectional collaboration and power sharing also damages the community-researcher relationship and can create a climate that the community does not perceive as open and just. This impacts not only the active study but also may propagate sentiments of distrust that negatively impact future research efforts.
There was limited evidence that studies prioritized removing barriers to community participation. The most commonly reported strategies included the provision of monetary incentives and transportation assistance (i.e., bus pass). Although these may address some challenges, other social and structural barriers may also impede participation of community members in the research process. Academic partners have the opportunity to build community capacity by providing education and resources for navigating research processes (Table 4).
Defining ''community'' was critical to understanding how authentically each study's findings may have represented the SGM experience. We therefore limited our definition of community to include SGM individuals representing the community to be served by the research efforts. In several studies reviewed, researchers and/or community advisory board members included stakeholder groups such as SGM service organizations and clinical providers treating SGM patients; this alone was not sufficient to warrant community involvement.
Difficulty in recruiting a sufficient number of SGM community members is not unusual in health research, particularly when working in a small geographically defined community or with some subpopulations (e.g., TGNC). In this case, supplementing the community member group with relevant community leaders and stakeholders can be helpful. Similarly, some studies focused on exploration of topics for which the health care and/or service provider perspective was relevant. But because it could not be assumed that all stakeholders who participated were well qualified to speak with authority about the SGM experience, if no SGM community members with lived experience with the health topic of interest were involved in a particular study element, credit was not given for that element. SGM POC, bisexual, and transgender people remain inadequately represented on the general SGM health landscape. Within CBPR studies, however, a greater proportion of these populations were represented. Twentyfive studies reported mostly or entirely POC samples, with significant representation of Latinx and Black/African American participants. Over half of studies included bisexual participants, although only one focused exclusively on bisexual health despite evidence that most sexual minority adults, particularly younger cohorts, identify as bisexual. 93 Twenty-five percent examined transgender mental health and health care use and access experiences. Few studies, however, included gender diverse participants such as genderqueer, gender non-binary, and gender non-conforming individuals. 54,60,66,83 Intersectional perspectives of some groups such as immigrant Latinas, juvenile justice-involved girls, and autistic and neurodiverse youth were also included.
Despite these strengths in representation, inconsistencies in SOGI conceptualization, measurement, and operationalization were observed across studies. Variations were likely influenced by study timing (i.e., language reflective of conventions of the time), geographic region, and knowledge and preferences of research teams (it is unclear if SGM partner input was integrated as no studies explicitly reported this). For example, some studies recruited under the broad ''LGBTQ'' umbrella without disaggregating sexuality and gender subgroups. 51,52,62,67,80 Sex and gender are often mistakenly conflated with the assumption that they do not differ from each other.
Of note, only two studies explicitly reported representation of indigenous third gender/non-binary roles (i.e., two-spirit, fa'afafine, m ah u). 62,81 Our operationalization of SOGI was not entirely inclusive, and the search structure used missed relevant publications as we failed to include the appropriate MeSH terms and keywords-a major limitation. In addition, it is likely that people who endorsed these identities comprised a small proportion of total participants in included studies and were reported under ''other'' and ''additional'' SOGI categories.
Aggregating the diversity under the LGBTQ + umbrella risks mischaracterizing experiences of power, sexuality, and relations. SOGIs are core aspects that shape opportunities and experiences of discrimination that influence health; therefore, accurate conceptualization and measurement is crucial. Continued efforts have been made to improve measurement of sex, gender, and sexual orientation, most recently, the groundbreaking 2022 National Academies of Science, Engineering, and Medicine report. 94 This report is the most comprehensive to date of measurementrelated research for these constructs. As utilization of these recommendations permeate the research community, data collection can be enhanced, demonstrating respect to and making visible the SGM participant, partner, or patient. Dissemination will also enable data harmonization between data sources, further building the bodies of work that will inform future health interventions and practice.
SGM CBPR health research covers a range of topics. Although our definition of ''health'' research was expansive enough to include topics such as mental health and SDOH, it is possible that other relevant healthrelated topics were excluded from consideration, limiting the results. In addition, we did not search gray literature, which might have resulted in the identification of additional studies that would have been eligible for inclusion. In studies identified there was a concentration in HIV/AIDS, reflective of patterns in SGM health research overall. More research is needed in a large number of diverse areas, including demographic research, intervention research, social influences, and health inequities. 28 Use of CBPR can play a key role in addressing these gaps, particularly exploration of approaches to addressing the SDOH. 2 The majority of studies reviewed were exploratory or descriptive with either a qualitative or cross-sectional survey design. This decreased overall community involvement scores as credit could not be given for activities such as intervention implementation or participant retention. In addition, capacity building, partnership sustainability, and use of findings for policy change were commonly lacking, independent of study design. These are significant and interrelated gaps. Moving the needle from exploratory to interventional research can provide a foundation of growth, sustainability, and innovation within a community that over time can enable transformational change to occur. To facilitate this shift, commitment from fundersparticularly those of federal mechanisms-is integral.
CBPR studies are often unfunded, as was the case for 30% of studies in this review, or funded by smaller Ricks, et al.; Health Equity 2022, 6.1 http://online.liebertpub.com/doi/10.1089/heq.2022.0039 mechanisms, offering little-to-no incentive or sustainable support to engage in robust CBPR work. Adequate and flexible funding opportunities are needed to support the extended time needed upfront to build relationships with community partners, as well as the additional time-months or years-often involved as the community participation process evolves. Increased success has been demonstrated in programs that have had access to ''braided funding'' from multiple streams, a model that affords more flexibility in terms of concrete ''deliverables'' and provides programs the authority to fund efforts that are based on emerging community needs. 95 Similar flexibility should be explored and expanded in other government and foundation research grant programs to move beyond the rigid model of preprogram budgets.
Other structural and educational barriers to CBPR work have been documented (Table 4) and may have limited CBPR projects included in our review. For example, the typical timetable for promotion and tenure at academic institutions may not be amenable to the pursuit of CBPR as this methodology is more timeconsuming than traditional research approaches. Researchers must contend with the time required to build sustainable partnerships and recruit and train community researchers, shifts in community priorities and leadership, and other unanticipated hurdles. 96,97 In addition, although public health juggernauts such as the Institute of Medicine encourage academic researchers to foster community-academic partnerships that share the strengths of each and call for CBPR to be included among traditional curricula, 98,99 most graduate programs do not require training in community-based research methodology. This may limit acceptability of CBPR as a viable, academically acceptable option. It certainly restricts the ability of untrained academic practitioners to effectively conduct CBPR research. 100 Conclusion Implementing the CBPR framework with true fidelity is challenging. Across studies we found wide variation in the extent to which communities were involved in research activities, reflecting the diversity of CBPR partnerships, settings where SGM health research is conducted, and the inherent difficulty in adhering to the model. Although achieving a perfect score on the CBPR principles is difficult, if not impossible, the framework represents targets to strive for in the pursuit for more equitable and collaborative research con-ducted in SGM communities. Prioritizing this work can have a transformational impact on reducing the fundamental inequities that threaten SGM health.

Author Disclosure Statement
No competing financial interests exist.

Funding Information
No funding was received for this article.

Supplementary Material
Supplementary Appendix SA1 Supplementary Appendix SA2 Supplementary Appendix SA3