The purpose of this study was to assess the influence of community health volunteer’s practices in advancing social accountability. Prior studies have shown that CHVs serve as a link between the community and the health system [15, 23, 29] which provides them with an opportunity to influence health system responsiveness and priorities . For example a study in Nepal showed that Female Community Health Volunteers collected concerns from the community and reported them during regular meeting with the health providers . The next sections reflect on the CHVs practices (information sharing, documentation of complaints/compliments and providing feedback) and their influence on social accountability.
According to these study findings, CHVs informed the community about their health rights and entitlements, but not adequately. For instance, majority of the CHVs did not inform the community the role they play in improving service delivery. This practice was limited by CHVs lack of awareness on social accountability, similar to  findings. As a result of CHVs' lack of social accountability capacity, community members may be unaware of their health rights and entitlements. Limiting how they express their concerns, as most informal settlement household members rely on intermediary structures such as CHVs to empower them in health-related matters . Findings from this study are consistent with those of  which stated that if CHVs feel empowered, they will be able to empower the communities they serve. In addition, their performance would improve, allowing them to realize their potential as social change agents . Other studies from India, Nepal, and the Democratic Republic of the Congo, have reported the impact of CHVs/FCHVs on providing the community with information on health services and educating them about the importance of reporting concerns [25, 32, 18, 33]. These studies found that community empowerment on health rights could lead to increased use of services and the ability to seek help [25, 32, 18, 33] and the reverse was it could result to poor adherence to treatment .
In this study CHVs reported to actively encourage the community members to speak up and they would listen to their issues during household visits. This study did not exhaustively investigate the factors that influence or limit health clients from speaking up even when they have issues to complain about  but social economic status [10, 28], knowledge and power asymmetries  have been documented to influence how they raise health concerns. By not proactively seeking the opinion, expectations or complaints about health services, could result to the voice of the vulnerable and marginalized in the society being left out in health system decision making. Lodeistein  reported intermediaries like Health Facility Committees not to proactively seek health users’ opinions or complaints about health services while Accredited Social Health Activists in India  concentrated in achieving service delivery targets like immunization coverage compared to their advocacy role. Successful social accountability mechanisms should deliberately encourage health clients to speak up and encourage responsiveness, if we are to achieve person-centred services.
Documentation of Complaints and Compliments
The CHVs practice of recording complaints and compliments was very low in both areas of study. Discussions revealed that verbal reporting was commonly used to express concerns from the community. Lack of complaint and compliment indicators on the Ministry of Health community reporting tools was one of the reasons for no documentation. When health concerns are not documented, they may have an impact on accountability because health authorities will lack a reference point for information. Documenting complaints and compliments can help in tracking health concerns and how they are handled. Evidence of this study on documenting complaints corroborate with those from Democratic Republic of Congo , and Nepal . Their studies established that by not having proper systems for recording and analysing complaints resulted to community concerns not being addressed. A study in India  established that CHVs' ('Matinins') practice of submitting written complaints to officials for action improved how health authorities handled issues that could not be resolved through mediation.
Studies  have proposed the need for instruments that allow for more systematic data collection and documentation after identifying gaps with health facility committees collecting complaints and getting lost due to lack of documentation. A good documentation system that enables systematic methods of collecting, analyzing and responding to complaints/compliments is required for effective use of data. Complaints must be documented in order to understand their frequency and nature [12, 19], increasing the community's voice in health-care management significantly. Documentation will also demonstrate which mechanisms are more suitable and effective .
Feedback to Health System and Community
These study findings showed inconsistency in CHVs practice of reporting complaints/compliments to the CHA or facility in- charge for action. The reasons could have been lack of awareness on formal channels of complaint handling and fear by the CHVs. Informing CHVs on formal channel of complaint/compliment handling mechanisms in the community health system, would help in ensuring there is a standardized and systematic way of handling community concerns of the health system. Lack of formal systems at local health centre’s or a representative of the population to present complaints or concerns to health providers were reported to be lacking in Democratic Republic of Congo by . In addition, their study reported community groups lacked the capacity and expertise to express their concerns or exert pressure on public officials or health care providers.
Findings from this study showed not all CHVs provided feedback to clients after they had raised a complaint. The decision to provide clients with feedback was dependent on if the CHVs were informed on action taken after the complaint was raised. The findings demonstrated lack of formal complaint handling mechanisms with clearly defined feedback loops in community and facility systems. Feedback is an important component of social accountability because it could increase customer satisfaction  and enhance enforceability and answerability . Citizen feedback must be communicated to relevant actors or decision-makers who can act on the information and/or who may incur costs as a result of the information . Well-functioning feedback loops necessitate a response capacity (which can be improved through social accountability) to ensure that reported issues are noted and action is taken to resolve problems as they arise, preferably with regular communication from relevant higher authorities to frontline service providers about resolution plans and timelines .
Feedback to the health system and clients can promote transparency, performance, fairness and respect especially by the health providers . Positive feedback provided directly by the clients to the health providers or indirectly through intermediaries can elicit feelings of happiness, achievement and accomplishment. However, negative feedback, on the other hand, elicits feelings of incompetence and demotivation . Whichever the outcome, the benefits outweigh the risks; therefore social accountability mechanisms should strengthen the feedback component. Social accountability approaches can improve service providers' responsiveness to service users' needs and their understanding of the challenges they face, foster better government-citizen relationships, and provide valuable feedback on the status of basic service delivery in a given country .
First limitation was the questionnaire was administered to a small sample size compared to the total number of CHVs in Nairobi County which could affect generalization. However this bias was minimized by use of mixed methods, which aided in the in-depth understanding of the CHVs practices in social accountability and thus the findings, can be used in similar settings. The second limitation was social desirability bias. This bias was reduced by triangulating the data sources and validating the findings with study participants at the end.