Community participation in health services development, implementation, and evaluation: A systematic review of empowerment, health, community, and process outcomes

Background Community participation is widely believed to be beneficial to the development, implementation and evaluation of health services. However, many challenges to successful and sustainable community involvement remain. Importantly, there is little evidence on the effect of community participation in terms of outcomes at both the community and individual level. Our systematic review seeks to examine the evidence on outcomes of community participation in high and upper-middle income countries. Methods and findings This review was developed according to PRISMA guidelines. Eligible studies included those that involved the community, service users, consumers, households, patients, public and their representatives in the development, implementation, and evaluation of health services, policy or interventions. We searched the following databases from January 2000 to September 2016: Medline, Embase, Global Health, Scopus, and LILACs. We independently screened articles for inclusion, conducted data extraction, and assessed studies for risk of bias. No language restrictions were made. 27,232 records were identified, with 23,468 after removal of duplicates. Following titles and abstracts screening, 49 met the inclusion criteria for this review. A narrative synthesis of the findings was conducted. Outcomes were categorised as process outcomes, community outcomes, health outcomes, empowerment and stakeholder perspectives. Our review reports a breadth of evidence that community involvement has a positive impact on health, particularly when substantiated by strong organisational and community processes. This is in line with the notion that participatory approaches and positive outcomes including community empowerment and health improvements do not occur in a linear progression, but instead consists of complex processes influenced by an array of social and cultural factors. Conclusion This review adds to the evidence base supporting the effectiveness of community participation in yielding positive outcomes at the organizational, community and individual level. Trial registration Prospero record number: CRD42016048244.

definitions by George et al. (2015) [12], the concept of community and community participation is described in Box 1.

Data sources
We developed the search string in accordance with the underlying objective of the study and refined it with inputs from an information specialist. The following databases were searched from January 2000 to September 2016: Medline, Global Health, Embase, Scopus, and LILACs. The full search terms used for Medline are shown in Table 1.
Inclusion criteria. We included all studies that involved the community, service users, consumers, households, patients, public and their representatives in the planning, implementation, monitoring and evaluation of health services, policy, or interventions. These included studies that involved the community in disease prevention, promotion, or healthy living, and/ or health service delivery. Studies that involved patients in decision making of personal healthcare decisions only were excluded from our review. We also excluded studies where Community Based Participatory Research (CBPR) was used merely to suggest ideas rather than as part of implementation in a community program. For this review, we excluded editorials and theoretical studies but included reports which had a description of the community participation component. We did not impose any language restrictions but limited the search to published literature from high and upper-middle income countries as defined by the World Bank.
Search and retrieval of studies. Two reviewers (SS and AS) double screened titles and keywords for 20% of the total articles from the search in the databases (kappa coefficient = 0.82). The remaining 80% of the articles were distributed among SS and AS and screened only once due to the high initial Kappa coefficient. Following the title screenings, the abstracts included were double screened (kappa coefficient = 0.84). Any disagreement at this stage was discussed between SS and AS. In the absence of a consensus, opinion was sought from a third reviewer for resolution. Five reviewers (SS, AS, VH, FC, HLQ) conducted the fulltext screening. Articles in languages other than English (e.g. French, German, Spanish, and Portuguese) were screened by a reviewer who could read and understand the article. Disagreements were resolved by a third reviewer. Only papers that reported outcomes or effects of community participation were included in this review. The details of the studies screened and included at each stage are presented in a flowchart in Fig 1. Data synthesis. Two reviewers (VH and FC) conducted data extraction using standardized forms including categories on: (1) study characteristics including study design and setting, (2) type of community involvement described in the paper, and (3) outcomes reported. The two reviewers (VH and FC) met regularly to discuss and resolve any discrepancies or disagreements on the data extraction or interpretation of the studies. We conducted a narrative synthesis of the findings.

Box 1. Definitions
Community: Communities are defined as constituted by those with a shared social identity; that is of members of the same set of social representations, which are the meanings, symbols, and aspirations through which people make sense of their world.
Community participation: Active group participation or participation of a person as representative of the group in activities where they not only provide ideas but are also involved in the intervention.
Risk of bias assessment. Two reviewers (VH and FC) assessed the studies for risk of bias. The Cochrane risk of bias tool was used to assess randomized control trials (RCTs) while observational studies were assessed using a proforma with 3 domains: selection bias, information bias, and confounding, then categorised as low, high, or unclear. Qualitative studies were evaluated for quality with an adapted checklist used in a previous series of mixed methods High income and upper-middle income countries "Argentina" OR "Albania" OR "Fiji" OR "Namibia" OR "Algeria" OR "Gabon" OR "Palau" OR "American Samoa" OR "Georgia" OR "Panama" OR "Angola"OR "Grenada" OR "Paraguay" OR "Azerbaijan" OR "Guyana" OR "Peru"OR "Belarus" OR "Iran" OR "Romania" OR "Belize" OR "Iraq" OR "Russian Federation" OR "Bosnia and Herzegovina" OR "Jamaica" OR "Serbia" OR "Botswana" OR "Jordan" OR "South Africa" OR "Brazil" OR "Kazakhstan" OR "St. Lucia" OR "Bulgaria" OR "Lebanon" OR "St. Vincent and the Grenadines" OR "China" OR "Libya" OR "Suriname" OR "Colombia' OR "Macedonia" OR "Thailand" OR 'Costa Rica" OR "Malaysia" OR "Turkey" OR "Cuba" OR "Maldives" OR "Turkmenistan" OR "Dominica" OR "Marshall Islands" OR "Tuvalu" OR "Dominican Republic" OR "Mauritius" OR "Venezuela" OR "Guinea" OR "Mexico" OR "Ecuador" OR "Montenegro"OR "Andorra" OR "Gibraltar" OR "Oman" OR "Antigua and Barbuda" OR "Greece" OR "Poland" OR "Aruba" OR "Greenland" OR "Portugal" OR "Australia" OR "Guam" OR "Puerto Rico" OR "Austria" OR "Hong Kong" OR "Qatar" OR "Bahamas" OR "Hungary" OR "San Marino" OR "Bahrain" OR "Iceland" OR "Saudi Arabia" OR "Barbados" OR "Ireland" OR "Seychelles" OR "Belgium" OR "Isle of Man" OR "Singapore" OR "Bermuda" OR "Israel" OR "Sint Maarten" OR "British Virgin Islands" OR "Italy" OR "Slovak Republic" OR "Brunei" OR "Japan" OR "Slovenia" OR "Canada" OR "Korea" OR "Spain" OR "Cayman Islands" OR "Kuwait" OR "St. Kitts" OR "Nevis Channel Islands" OR "Latvia" OR "St. Martin" OR "Chile" OR "Liechtenstein" OR "Sweden" OR "Croatia" OR "Lithuania" OR "Switzerland" OR "Curacao' OR "Luxembourg" OR "Taiwan" OR "Cyprus" OR "Macao" OR "Trinidad and Tobago" OR "Czech Republic" OR "Malta" OR "Turks and Caicos Islands" OR "Denmark" OR "Monaco" OR "United Arab Emirates" OR "Estonia" OR "Nauru" OR "United Kingdom" OR "Faroe Islands" OR "Netherlands" OR "United States" OR "Finland" OR "New Caledonia" OR "Uruguay" OR "France" OR "New Zealand" OR "Virgin Islands (U.S.)" OR "French Polynesia" OR "Northern Mariana Islands" OR "Germany" OR "Norway"OR "High income countr � " OR "upper-middle income countr � " OR "developed countr � " OR "developed nation � " OR "developed population � " systematic reviews [13,14] scored for ten core criteria. We classified studies with a score of eight to ten as having an overall low risk of bias, four to seven as having an overall medium risk of bias, and zero to three as having an overall high risk of bias. We did not conduct a risk of bias assessment on case studies; however, we have included these studies in our review as they give insight into the mechanisms of partnerships, inter-organisation collaboration, and stakeholder satisfaction.

27
,232 records were identified through database searching. 23,468 articles were screened by title followed by 1,740 abstracts screened for inclusion. The full text of 707 articles was obtained and assessed for eligibility. After screening for reported objectives, 49 articles met eligibility criteria for this review (Fig 1). Due to the heterogeneity in study design, intervention types, participants, and outcomes, we conducted a narrative synthesis of the findings instead of a meta-analysis.

Characteristics of included studies
Of the 49 studies that met inclusion criteria, 22 were quantitative, 14 were qualitative, and 13 were case studies. Of the 22 quantitative studies, 6 were RCTs, 8 were intervention studies, 7 were cohort studies, and 1 was a cross-sectional study. The studies could be categorised into five different disease categories based on the focus of the community participation initiative described. Of the 49 studies, 16 focused on community health in general, 13 involved initiatives that targeted healthy living, 9 focused on non-communicable diseases, 7 studies addressed infectious diseases, and 4 studies were related to environmental health. The description of each disease category and the number of relevant studies are presented in Table 2.

Outcome definitions and framework
Reported outcomes were classified as process outcomes, community outcomes, health outcomes, stakeholder perspectives, and empowerment (See Table 3). We define process outcomes as short-term outputs that reflect the effectiveness of collaborative processes and activities over time. Organizational processes are concerned with community-based group achievements, while community processes are linked to process-related changes in the targeted community. We define community outcomes as intermediate social effects that represent changes in community member's knowledge, attitudes, and behaviors. More extensively, it Overall, studies were located in North America (n = 25), Europe (n = 9), Asia (n = 5), South America (n = 6), Africa (n = 1), and Oceania (n = 3) (Fig 2). The community health category featured the most geographic diversity with studies from nine different nations represented. The United States was represented by studies in all categories.
includes outcomes that reflect impact on social capital, community development, socio-cultural, and environmental improvements. Health outcomes are those that reflect changes in community member's health status. We also describe those outcomes that deal with larger sociopolitical influences, as well as stakeholder perceptions. Studies also report on empowerment at the community or individual level, as an outcome. Studies that defined empowerment framed it as communities coming together to address a self-identified community problem and create positive change that is self-sustained, contextually appropriate, and fosters knowledge transfer between community members. These studies also point to complicated power relations and structural differences between community members and professionals or policy makers that underpin the challenges in defining and measuring community or individual empowerment (See Table 4).
Outcomes of community involvement initiatives may be viewed through a hierarchy, as some outcomes necessitate others (See Fig 3); for example in order to deliver a community involvement program that reports robust health outcomes, it is important to have functional and sustainable underlying organisational structures, as well as community awareness and involvement. Throughout this hierarchy, both organisation and community members may report perspectives on the process or outputs and may feel empowered at either a personal or community level.
The number of outcomes reported by disease category and study design can be found in Table 5. Twenty-nine studies reported process outcomes, of which twenty-three reported organisational processes and nine reported community processes; twenty-one studies reported community outcomes; sixteen reported perspectives of stakeholders on either processes or project outcomes; six reported on empowerment and twelve reported health outcomes. Process outcomes, especially organisational processes, were most often reported in studies involving community health (n = 12), while both infectious disease and environmental health category only had one study reporting these outcomes. Empowerment was the least reported across study categories; of 6 studies, 4 were in the community health category. Health outcomes were more often reported in healthy living (n = 4) and non-communicable disease initiatives (n = 5), while community health initiatives reported no health outcomes.

Definition of Empowerment Category Author/Date
"Individual levels of empowerment" described in terms of youth's ability to "reach out" and disseminate health information to the community. Focus on reaching out to and advocating for undocumented immigrants and helping them to gain confidence, knowledge and access services while "feeling empowered to motivate others to do the same."

Community Health
Ferrera et al 2015 [15] "When local people at all levels are drawn together with the purpose of employing local wisdom to solve a problem which they all face, the result is a sense of empowerment to make changes, which are intrinsically sensitive to local circumstances, widely accepted by the community, and because of this, more likely to be sustained"

Environmental Health
Sansiritaweesook et al 2015 [16] "Empowerment is related to the process of giving groups of communities autonomy and a progressive and selfsustained improvement of their lives."

Process outcomes
Study characteristics, along with the findings reported, and the risk of bias assessments for studies that report on process outcomes can be found in Table 6 (See S1 File for table legend for risk of bias). Nine studies presented process outcomes relating to contextually appropriate initiatives and mutually agreeable organizational processes to meet community's needs [15, 16, 25, 26, 28-30, 44, 45]. Four studies reported on how collaborative processes led to the creation of appropriate policies and community-led priority setting [19,22,34,43]. Two studies reported clearer role definition as a process outcome of community involvement in community health initiatives [3,46] while two studies reported how robust processes enabled the provision of more activities [20,47]. Yet, not all partnerships showed favorable results, due to conflicting stakeholder views, as well as underestimation of the time and resources required for collaboration [35].

Community outcomes
Study characteristics, along with the findings reported and the risk of bias assessments for studies that report on community outcomes can be found in Table 7 (See S1 File for table legend for risk of bias).
Eight studies provided evidence on community outcomes in the form of increased community knowledge and awareness [15,35,43,44,49,52,53,55]. Two studies involved interventions that focused on community health in general [15,44], 1 on community mental health [43], 3 on infectious diseases [35,52,55], 1 on environmental health [53], and 1 on a healthy living intervention involving a physical activity trial [49]. Five studies reported on community outcomes relating to improved self-efficacy and confidence [22,27,46,52,54]. Two studies that reported on such outcomes had contextually tailored interventions on HIV and AIDS [52,54]. Both studies reported positive impact on its target population including increased confidence and personal development among peer educators and sex workers, decreased HIV stigma, reduced proportion of men reporting that they had engaged in unprotected sex, and increased positive attitudes in condom use.

Stakeholder perspectives
Study characteristics, along with the findings reported and the risk of bias assessments for studies that report on stakeholder perspectives can be found in Table 8 (See S1 File for table legend for risk of bias).

Process
Outcome 1) Villagers collaborated to conduct a situation analysis, design, and trial a prototype intervention, scale up to a full system design and trial that was followed by system improvement and dissemination. 2) 80% of networks were cooperative in submitting timely reports and using them for action. 3) Accuracy of information in reports increased from 65% to 90%.

Process
Outcome 1) BPC schools demonstrated better outcomes with more activities and lessons than BP schools. 2) In year 2 there was a higher mean number of physical activity and healthy eating programs being implemented in BPC schools (mean = 3.71 programs) compared to BP schools (mean = 2.73 programs).

Neto et al 2003 [21]
Brazil Intervention Study 1,524 households in intervention area; 1,564 households in control area Infectious Disease A preliminary diagnosis presented to the community to launch a discussion aimed at defining future actions, implementation of the actions in the study area with community participation.

Process
Outcome Changes in the study area included: vector control workers began demonstrating preventive measures without removing potential breeding places or using larvicide; use of educational aids specific to the local reality; activities related to the residents' priorities; and activities such as music, theater skits, scavenger hunts, and games to demonstrate the vector cycle.

Reeve et al 2015 [25]
Australia Cohort N/A Non Communicable Diseases A health service partnership between an Aboriginal community-controlled health service, a hospital, and a community health service that implemented an integration of health promotion, health assessments, and chronic disease management.
Process Outcome Short-term outcomes-1) Increase in occasions of service (from 21,218 to 33,753) particularly in PHC in remote areas (from 863 to 11,338). 2) Increased uptake of health assessment (from 13% of eligible population to 61%, then to 73% of those identifies with DM placed on a care plan). Medium-term outcomes-1) Over a 6 year period, improvements in quality-of-care indicators, i.e. glycated hemoglobin checks and proportion of people with DM receiving anti hypertensives. 2) Increase in proportion of patients identified with chronic disease or risk factors. 3) Increased PHC episodes and follow-up.

Nelson et al 2006 [26]
Canada Cohort 79 Consumer Survivor Initiative members Non Communicable Diseases Consumer Survivor Initiatives-organizations that are operated by and for people with a history of mental illness.

Process
Outcome Members participated most often in internal activities (e.g. social-recreational, committees) and least often in external activities (e.g. advocacy, planning, education) with an average of 3 activities per month.

Litt et al 2013 [27]
United States Cross-sectional 59 collaborative groups representing 22 states Healthy Living Collaboratives formed to improve the built environment and policies for active living.
Process Outcome Groups made progress in identifying areas for environmental improvements and in many instances received funding to support these changes: 1) Groups' environmental improvement scores ranged from 1.5 to 5.0, with an average of 3.5 (SD: 0.9). This average indicated that groups typically had funding to support their initiatives and had started but had not completed the planned improvements.

Trettin et al 2000 [29]
United States Qualitative 6 to 14 participants of 3 focus groups (total n = 60) Community Health Volunteer-based community health advisory program developed to increase residents' access to health services, stimulate their interest in health, disease prevention, and awareness of health-related environmental issues, and empower residents to be more involved in community health.  Process Outcome 1) Coalition facilitated an average of at least 3 times as many community changes (i.e., program, policy, and practice changes) per month following the intervention. 2) After intervention, there was increased implementation of 3 key prioritized coalition processes: Documenting progress/using feedback (75% increase in stakeholders involved in designing the documentation system); making outcomes matter (50 to 100% increase in activities relating to incentives, accountability, and use of longer term outcomes with accountability); and sustaining the work (42% to 75% increase in identification of sustainability decision makers, determining what to sustain and duration of sustained effort). 2) A 1-year probe following the study showed that majority of the community changes were sustained.

Bursztyn et al 2008 [38]
Brazil Case Study Not mentioned Community Health A project was developed and implemented in primary health centers to improve young men's adherence to a teenage health care program using participatory planning techniques, and rapid assessment procedures.
Process Outcome 1) Self-assessment workshops were held with the local teams. Despite good awareness among the health professionals, the project's results varied between health centers. Over-centralization and lack of flexibility appear to be related to lower capacity to incorporate new practices. 2) Health centers where specific strategies were observed showed more successful results.

[39]
Mexico Case Study Not mentioned Community Health Use of participative strategies and the creation of support networks for poor pregnant women.
Process Outcome Coordination and community participation were relevant in relation to major resources allocation and availability, particularly housing and transportation.

Case Study
Not mentioned Healthy Living 'Active Living by Design' partnerships were established to change environments and policies, and support complementary programs and promotions to increase physical activity.
Process Outcome The connections among diverse community partners created a foundation that enhanced lead agency efforts to form, implement, and maintain policy changes and physical projects, as well as promotional and programmatic approaches, to support active living.

Rapport et al 2008 [41]
United Kingdom Case Study Focus groups with project steering group Healthy Living Action research project-organized to respond to a context of funding and service delivery, helmed by a Project Steering Group made up of community members, study organizers, statutory board members.
Process Outcome 1) Community members involved acquired new skills and "strengthened individual competencies," heightened knowledge amongst the community and Project Steering Group of community members' needs and desires," influenced working practices, altered perspectives and raised awareness of issues surrounding trust and communication within partnerships.
2) The data generated by the community interviews was perceived as more robust evidence that could be "taken seriously and gave credibility to the communities' comments and requests."

Not mentioned
Infectious Diseases Ecohealth approach used as a strategy to ensure active participation by the community, diverse sectors, and government. The approach allowed holistic problem analysis, priority setting, and administration of solutions.
Process Outcome 1) The strategy had been sustained two years after concluding the process. 2) 93.5% had attended trainings under the project and 89% knew that the inhabitants of the neighborhood had organized themselves into groups promoted by the project. 3) 93.5% considered that the community improved its ability to identify problems that affected its ecosystem and proposed solutions. Community participation in health services: A systematic review on outcomes

Qualitative 30 members of community interviewed Infectious Diseases
A community-based peer education program led by sex workers as an initiative in grassroots participation in sexual health promotion.

Community Outcomes
Increased confidence and personal development among peer educators and increased confidence among some sex workers.

Chervin et al 2005 [35]
United States Qualitative 364 in-person interviews with project staff, evaluators, and community and agency members

Centers for Disease Control and Prevention's Community Coalition
Partnership Program (CCPP)-building a community's capacity to prevent teen pregnancy through strengthening of partnerships, mobilization of community resources, and changes in the number and quality of community programs.
Community Outcome 1. Increased community awareness of the problem of teen pregnancy and willingness to discuss the issue; 2. Improved knowledge and skills relating to addressing teen pregnancy.

Mexico Case Study Not mentioned Community Health
Use of participative strategies and the creation of support networks for poor pregnant women.

Community Outcome
Governmental actors' involvement and leadership favored linking and coordination. Authorities, relatives, volunteers and users supported the referrals for obstetric emergencies, the identification of pregnant women in isolated areas, and their referral to health services. Around one-third of the users indicated geographical, economic, and cultural access barriers to health services in the four states, particularly those living in rural areas. Even though most of the informants received timely attention with a favorable evaluation of the treatment received in the units, testimonies were collected from users reporting feeling abused by transporters and suppliers.

Setti et al 2010 [53] Brazil Case Study 24 participants Environmental Health
The Neighborhood Ecological Program that involved the participation and empowerment of citizens in health promotion and sustainable development

Community Outcome
The program is reported to promote empowerment and community strengthening, dissemination of information and knowledge, development of critical thinking, and the creation of support networks.

N/A (Continued )
In five studies, participants reported positive experiences or satisfaction with the community participatory initiative [15,58,59,61,62], three of which involved community-academic partnerships [58,61,62]. Six studies reported on stakeholder perspectives that reflected positive where and how to access information, developing directories of local service user groups/resources, and understanding the value of advocacy. 5) Positive changes in attitudes towards partnership with service users. 6) Positive changes in behavior at individual level, e.g. students more conscious of sharing decision-making and using a needs-led approach following awareness of the imbalance of power between service users and professionals. 7) Positive changes in behavior at organizational level, e.g. the setting up of service user groups, ensuring user views are fed into planning decisions, supporting service users on staff recruitment panels, writing leaflets for users/carers about services offered, and collating info on resources for users.

Wilson et al 2014 [54]
United States Case Study 71 participants Infectious Diseases CBPR used to develop the Barbershop Talk With Brothers (BTWB) program-a community-based HIV prevention program that seeks to improve individual skills and motivation to decrease sexual risk, and that builds men's interest in and capacity for improving their community's health.
Community Outcome 1) Proportion of men who reported not having engaged in unprotected sex in past 3 months increased from baseline to follow-up administration of survey (25% to 41%, p = 0.007). 2) Proportion of men who reported having unprotected sex with two or more women in the past 3 months declined (46% to 17%, p = 0.0001). 3) Proportion of men reporting favorable attitudes towards condoms and confidence in their self-efficacy to use condoms consistently increased (p<0.05). 4) HIV stigma decreased, but difference did not reach statistical significance (Mean = 24.7; SD = 8.4 to Mean = 22.8; SD = 8.8; p = 0.11).

Diaz et al 2009 [42] Cuba Case Study Not mentioned Infectious Diseases
Ecohealth approach used as a strategy to ensure active participation by the community, diverse sectors, and government. The approach allowed holistic problem analysis, priority setting, and administration of solutions.

Community Outcome
At the outset, 85% of the outbreaks of the dengue vector were in tanks located in the patios of the houses. Two years later only 29% were located in the patios. Currently, no outbreaks have been identified in the deposits located in the houses. It was found that 16% of the 4,878 courtyards in the territory were unhealthy. Two years after the end of the study, these constituted less than 1%; The number of unprotected tanks decreased from 62% to 8% (n = 4,678).

King et al 2011 [55]
American Samoa Community-academic partnership. Members included a non-profit agency, university representatives, and participants from health, education, government, and lay leadership sectors.
Stakeholder Perspectives 1) Participants expressed satisfaction with the formation and maintenance of the committees and noted that the committees were still actively meeting in the community 2 years after they were formed. 2) Satisfaction with committee participation in community events. 3) Satisfaction with raising awareness about the committee in the community. 4) Participants spoke of individual benefits of becoming personally more aware of nutrition and physical activities.

Qualitative 61 individuals interviewed
Healthy Living Community Engagement Model-Well London program, community specific interventions for healthy eating, physical activity, and mental wellbeing delivered in socioeconomically deprived neighborhoods.
Stakeholder Perspectives 1) Positive benefits reported by those who participated in project activities. 2) Extent of benefits experienced was influenced by physical and social factors of each neighborhood. 3) Highest level of change in perception occurred in neighborhoods where there was social cohesion, personal and collective agency, and involvement and support of external organizations.

Qualitative 35 key informants interviewed
Healthy Living 'Lay food and health workers' and professionals involved in delivering local food and health initiatives in less-affluent neighborhoods.
Stakeholder Perspectives 1) Salient benefits identified were increased service coverage, ability to reach the "hard to reach", as well as personal development and enhanced social support. community-level outcomes [57][58][59][61][62][63]. Two of these studies reported greater awareness of the targeted health issue or services among the community, both of which involved community-academic partnerships [59,62]. Three studies reported perceptions relating to the processes of involving the community, although results were mixed [44,57,58]. Two of the studies reported stakeholder satisfaction with service coverage, staff development, enhanced networks, and creation of new alliances [44,58]. However, another qualitative study that investigated perspectives of a United States Qualitative 6 to 14 participants of 3 focus groups (total n = 60)

Community Health
Volunteer-based community health advisory program developed to increase residents' access to health services, stimulate their interest in health, disease prevention, and awareness of health-related environmental issues, and empower residents to be more involved in community health.
Empowerment Sense of empowerment fostered among participants when they were given greater control over the direction of the program.

Qualitative 23 youths interviewed
Community Health CBPR used to form youth advisory board and youth involved in decision making and programming, as well as in a feedback and improvement role.
Empowerment Improved sense of agency amongst students. Community participation facilitated an understanding of how students may have a positive impact on their community. "Individual levels of empowerment" described in terms of youth's ability to "reach out" and disseminate health information to their family members and the immigrant community. Reaching out to and advocating for undocumented immigrants helped them to gain confidence and knowledge on accessing services. They felt empowered to motivate others to do the same.

Qualitative 35 key informants interviewed
Healthy Living 'Lay food and health workers' and professionals involved in delivering local food and health initiatives in less-affluent neighborhoods.
Empowerment Empowerment was perceived as both an individual benefit and a benefit to the community resulting from the program. Community participation in health services: A systematic review on outcomes  Health Outcome 1) In the year after system implementation the nonfatality drowning rate in target areas fell to zero, the non-fatality rate in control areas increased. 2) Fatality rate in target areas dropped to 4.5 per 100,000 but remained the same in control areas. Incidence rate ratio of injuries in the comparison areas was 23.32 times higher than in the target areas (95% CI: 3.081-176.599, p = 0.002).

[20]
United States Community participation in health services: A systematic review on outcomes Community participation in health services: A systematic review on outcomes health impact assessment among native participants reported otherwise, highlighting the need to account for a community's history of colonization and forced assimilation in the community engagement process [57]. At a more fundamental level, community participation has been perceived to have facilitated community ownership and development as reported in two studies [57,62].

Empowerment
Study characteristics, along with the findings reported and the risk of bias assessments for studies that report on empowerment can be found in Table 9 (See S1 File for table legend for risk of bias). Three studies described how participation in a community initiative fostered engagement [28,42,53]. Two studies described how greater agency, i.e. the capacity of individuals to act on their own accord, interacted with empowerment [15,29]. One study involved a volunteerbased community health advisory program that sought to increase access to health services which reported a sense of empowerment among participants after they were given greater control over program direction [29]. The other study, involving a youth advisory board formed through CBPR, reported an improved sense of agency amongst students [15]. One study described specifically how gaining skills through participation led to empowerment. The study involved a community-academic collaboration that led to resident empowerment through skills based training that was included in the CBPR research process [28]. In another study on active participation strategies for environmental solutions, community groups were reportedly mobilized to make changes in their own community, resulting in the strengthening and empowerment of the community [42].

Health outcomes
Study characteristics, along with the findings reported and the risk of bias assessments for studies that report on health outcomes can be found in Table 10 (See S1 File for table legend for risk of bias).
The health impact of community participation interventions was the most evident among studies involving non-communicable diseases. All five studies reported positive health outcomes including decreased hospital admissions [25,65], reduced clinical symptoms [22], improved behavioral risk factors such as exercise [46,49,64,66], improved quality of life [43], and decreased mortality over time [16]. Two studies on infectious diseases reported positive health outcomes in terms of greater community compliance to the prevention and treatment of lymphatic filariasis which was the targeted disease of the community participation program [55], and a lower rate of increased vector density of a dengue control intervention [17]. Two out of 4 studies relating to healthy living reported positive results relating to improvements in obesity rates [20,46], while the other 2 studies targeting physical activity did not find these interventions effective in promoting health outcomes [49,64]. Only one study on environmental health reported on health outcomes where the implementation of the local drowning surveillance system resulted in reductions in non-fatal drowning rates, drowning fatality rates and incidence rate ratios of injuries [16].

Discussion
This review explores reported outcomes of community involvement and participation and presents a conceptual model to frame these outcomes, beginning with a foundation of process outcomes and community outcomes as necessary to achieving robust health outcomes, while recognizing the influence of stakeholder perspectives and empowerment.
Our review highlights the importance of both process and outcomes evaluations when assessing community involvement interventions. Process outcomes, especially those that reflect on organizational processes, are the results of intra-and inter-organizational negotiating and learning, that over time results in "trust" and "authentic" relationships which ultimately drive partnerships forward [66]. Few studies report on the community processes that result from these initiatives, such as increased outreach, volunteerism or other "conversion" of community members into active members. From an organizational perspective, many studies reported on the learning phases wherein organizational relationships are established and built. Partnerships in this phase mostly report process outcomes as they learn ways of working both together and with the community [43]. This learning curve is important in developing contextually appropriate interventions and those studies that invest in this stage report success in program development and implementation [25].
Failing to account for contextual learning can result in failure to work together to achieve goals, and this is especially important in vulnerable populations and those communities with a history of colonization and forced assimilation [55]. This speaks to the international Aboriginal self-determination movement which calls for program development for indigenous people by indigenous people that integrates underlying theoretical and cultural frameworks into applied public health [17]. Past research has shown how community participation interventions have been viewed as an initiative to improve health outcomes rather than a process to implement and support health program to sustain these outcomes [20,46]. However, our findings highlight that examining community participation as a "process" is equally as important, and furthers the understanding that outcomes could be influenced by shifts in social, economic, and political contexts over time.
Overall, community-level outcomes were the most common measure reported across the studies. Findings from our review demonstrate that successful community outcomes were most evident among interventions that included outreach activities such as: health camps, community fairs, and partnerships with schools and religious groups [49,64]; targeted interventions that delivered tailored and specific health knowledge [16]; and interventions that encouraged relationship building with the wider community [28,41,44]. CBPR was also beneficial in developing trust between community and academic partners through the creation of a level-playing environment where members could decide on health priorities collectively [28,29,67]. In another review that examined the effectiveness of community engagement in health intervention planning and delivery, community participation initiatives were reportedly linked to positive gains in social capital, social cohesion, and in capacity building among the community [16,22]. Furthermore, a systematic review addressing what indigineous Australian clients valued about primary health identified how community participation influences access, acceptability, availability, responsiveness and quality of services, with the potential of increasing utilisation and ultimately improving health outcomes [68]. Another study also identified how increased community participation could also address the social determinants of health outcomes through increased local or Indigenous employment services [69]. In our review however, very few studies reported on such community outcomes, which are inherently more difficult to define and measure given its subjectivity.
In terms of population level outcomes, our findings indicate that there is a problematic reliance on empowerment as an outcome measure of community participation interventions. Some studies report on community empowerment and empowering of participants as a community level improvement resulting from participation in a community project or initiative [67]. Empowerment is perceived as beneficial and a positive outcome of community participation, often constructed through qualitative exploration of participants and residents' perceptions, but without a robust definition and measurement of impact, caution is required in attributing the outcomes reported to actual community empowerment. Furthermore, care must be taken not to reduce empowerment to a component of a bureaucratic process while conflating these debatable definitions and measures of empowerment to represent tangible power and influence [70]. Empowerment as an outcome requires sustained community engagement, which is dependent on program sustainability. While there may be many barriers to sustainability, the greatest challenges can be political [71].
Findings from our review indicate that the ultimate aim for most community involvement programs is to improve health and wellbeing of a particular community; however, indicators were difficult to obtain and measure. Changes in health status usually require long-term monitoring and may not be measurable over a single program cycle. In our review, health outcomes are most commonly reported for community involvement interventions addressing non-communicable diseases and healthy living, and findings presented are generally mixed. For instance, some healthy living interventions reported no significant effect of physical activity interventions on health outcomes [15,17,24,46,55,57] while others reported the contrary [22,65]. Nonetheless, interventions that are contextually targeted which have specific goals at the outset that are monitored over time seem to have greater success in achieving positive health outcomes [16,44,54]. As highlighted in other reviews, identifying that a positive outcome or change is specifically attributable to community participation is a complex task [44]. Community participation initiatives usually do not happen as a direct and linear intervention to improve health, but rather consists of complex processes and interactions [7]. Our review reports promising evidence that community engagement has a positive impact on health, especially when supported by a strong organizational and community foundation.
Despite the variability in interventions, there are some positive community participation examples that provide convincing evidence of benefits as demonstrated by the six RCTs identified in this review, two of which were of high quality given its overall low risk of bias [17][18][19][48][49][50]]. Boivin's study elucidates that community involvement is central to setting priorities in driving healthcare improvement at the population level [19] while Caprara's study presents social participation as an effective tool in facilitating environmental management for improved dengue vector control [17]. It should be noted however, that all studies described were context specific, hence the external validity of these studies are inevitably limited. Ultimately, there is 'no one size fits all' approach to community participation that will ensure intended positive outcomes and community participation that is tailored to context is fundamental in ensuring the provision of equitable health care and optimization of interventions to improve health [64].

Strengths and limitations
This systematic review on outcomes of community participation in high and upper middle income countries is the first of its kind to be conducted. A strength of this review was the use of a wide range of databases and the inclusion of papers in multiple languages to ensure broad representation. However, majority of the studies identified were conducted in the United States which could be a result of publication bias. It is highly likely that not many real world community participatory initiatives are evaluated robustly according to epidemiological standards, and it is possible that studies with null findings are less likely to be published. Additionally, given the broad scope of our inclusion criteria, the search produced a large amount of literature on community participation for eligibility assessment and synthesis. Nevertheless, prioritizing studies that had the best quality evidence in outcomes reported allowed for the data extraction and synthesis process, and the risk of bias assessment, to be done comprehensively and with rigour. Implications for research. Our review shows that while community participation and involvement is well documented from a case study and qualitative perspective, there is a need for more robust program evaluations and studies that measure and report long-term outcomes. Studies were largely descriptive or only had a evaluative component as part of a case study. While descriptive reports provide insight into program successes and operationalisation they would benefit from more robust methodology and reporting to determine stronger causal linkages between intervention components and desired outcomes.
Our review included six RCT studies that serve as positive examples for evaluating community participation programs. However, it must be noted that while RCTs are considered the gold standard in research methodology; difficulties in applying experimental designs at the population level is evident and well documented [7]. A particular challenge will be to account for the multi-faceted health and social dimensions of community participation in drawing definitive linkages and pathways that explain how community participation leads to a desired community or health outcome [6].
Importantly, no studies reported on outcomes relating to costs. Further evaluations are needed to examine the cost-effectiveness of real-world interventions and draw comparisons between the varying approaches of community participation and involvement. Such research is imperative to support evidence-based policy-making by identifying community participation programs that can achieve the greatest health return on investment.

Implications for policy
Evidence garnered from this systematic review presents some of the successes of community participation in yielding positive outcomes at the organizational, community, and individual level in high and middle-income countries. It is a worthwhile endeavour for policymakers to devote resources in enabling community engagement, creating platforms for involvement, and in facilitating successful collaborations or partnerships within the health sector and beyond. Nonetheless, addressing issues of power relations, developing trust with the community, and understanding the political, social, and economic contexts in which initiatives are supported, is imperative in any form of community engagement effort.
Based on the findings of this review, we have developed a new outcomes framework for community participation which policy-makers can utilise to prioritise program outcomes and justify resource allocation in program design and implementation. Consideration of the interplay of social and cultural factors is essential when exploring perspectives of community members on outputs of such initiatives, while empowerment and power relations are key elements that should be taken into account with more robust measurements. As policy-makers consider new and effective ways of planning, implementing, monitoring, and evaluating community involvement programs, the evidence here can contribute in providing some clarity to the process and supporting the development of evidence based policies.

Conclusion
Community participation is a fundamental element of an equitable and rights-based approach to health that is proven effective in optimizing health interventions for positive public health impact. This review adds to this evidence base supporting the utility of community participation in yielding positive outcomes at the organizational, community, and individual level across a wide range of health domains. Our findings present process and community outcomes as necessary to achieving robust health outcomes. This supports the notion that participatory approaches and health improvements do not happen as a linear progression, but rather consists of complex processes influenced by an array of contextual factors. Overall, it is evident that community involvement is key in priority setting to drive healthcare improvement and that interventions utilizing community involvement can benefit from a contextualizing learning phase whereby organizational relationships and trust can develop. Our review highlights the need for more robust program evaluations of community participation initiatives that measure long-term outcomes and cost-effectiveness, in more settings globally.
Supporting information S1