Introduction

Community health workers (CHW) are an important part of the health system, particularly in low- and middle-income countries (LMICs); and play a crucial role in linking the community with the formal health system. Different terms are used to refer to CHWs depending on context [1, 2]. For example, in India, they are referred to as Accredited Social Health Activists (ASHAs) while in other contexts they are known as lay health workers, traditional birth attendants, lay counsellors. For this paper, we broadly refer to them as community health volunteers (CHVs), which also aligns with the terminology used in Kenya.

Literature from LMICs shows that when involved efficiently, CHVs can improve access to primary health care services including: maternal and neonatal health services [3, 4]; psychosocial support services such as counselling, to individuals experiencing mental health problems like anxiety and depression [5, 6]; and the provision of preventive health education on topics such as hygiene and sanitation [7], exclusive breastfeeding [8, 9] and family planning [10]. CHV programmes also promote the provision of services such as child immunization [11], and in responding to major health outbreaks as was observed during the Ebola outbreak in West and Central Africa [12]; and more recently with the COVID-19 pandemic in Kenya [13]. Thus, CHVs are important actors in ultimately improving health outcomes, particularly in low-resource settings [14–18].

A systematic review by Kok et al., showed that the success of CHV programmes depends on three key aspects: programme design factors, health system factors, and broad contextual factors surrounding the workings of such programmes as shown in Fig 1 [2, 19, 20]. This work was guided by the Kok et al., [2015] framework (Fig 1). According to Kok and colleagues, programme design factors are features of the programme or intervention design that have a direct influence on CHVs’ performance and can be directly modified and adjusted to motivate maximum CHV performance [20]. They include renumeration, recruitment, supervision and workload, community and health system links, and resources and logistics [2]. Contextual factors relate to the setting in which the CHVs are working. That is, the broader environment, community, and economy; and the health system policy and practice [21]. Health system factors on the other hand include issues related to human resources for health policies, service delivery, financing, and governance [20].

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 1. Kok et al. (2015) framework.

https://doi.org/10.1371/journal.pgph.0001815.g001

CHV performance can be measured at two levels, the individual (CHV) level and at the end-user (community) level [20]. Attributes that are measured at the individual level include the cognitive, affective, and behavioural changes in the CHV such as self-esteem, attitude, motivation, job satisfaction, competencies/knowledge levels and guideline adherence. Whereas at the end-user or community level, one can measure CHV-attributable outcomes in the community. Such outcomes include for instance; credibility or trust in the CHVs, increased use or awareness of a health service, or the acceptance and use of services provided at a health facility following referral by a CHV [20]. This work focused on assessment at the individual level as highlighted in bold in Fig 1.

Literature shows that context matters when it comes to the performance of CHVs. It is, therefore, important to understand factors influencing CHV performance in specific contexts such as urban informal settlements (UIS). For example, in addition to the factors highlighted above, a systematic review by Ogutu et al., [22] showed that specific to UIS in LMICs, some of the factors that negatively impacted on the performance of CHVs at the broader contextual level in UIS included insecurity and the demand for material or financial support by households.

In Kenya, it is estimated that more than half of the urban population lives in informal settlements, with 56% residing in the capital city Nairobi [23]. These populations are often characterized by a high burden of disease and limited access to health care [24–28]. To increase health care access in these and other settings, in 2006 the government rolled out the Kenya Community Health Strategy (CHS). This strategy, predominantly relies on CHVs to deliver a range of services, and act as a link between the formal healthcare system and the community [29–32]. As per the strategy, health service provision in Kenya is centered around four tiers–community, primary care, primary referral (county health services), and tertiary referral (National services) [32]. Tier one (community unit) is identified as the first level of health service provision. It focuses on creating appropriate demand for services, by referring members of the community for services. Tier two, three, and four (primary and referral services) focus on responding to this demand. The sub-county health management team (SCHMT), is responsible for the coordination of tier one services, with the focal person (sub-county community health strategy coordinator) tasked with the supervision, planning and monitoring of community health-related activities. Tier one services are implemented through community health units (CHUs). Each unit consists of approximately 1000 households and is assigned to 10 CHVs and one government-employed Community Health Assistant (CHA); formerly known as Community Health Extension Workers (CHEWs). CHAs are trained health workers linked to a specific primary health care facility and provide support and supervision to an assigned number of CHVs. Each CHU is governed by a Village or Community Health Committee (V/CHC). This committee oversees the coordination and management of the CHU, with support from the primary care facility management committee and the SCHMT as shown below in Figs 2 and 3 respectively.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 2. Kenya health service management structure.

Adopted from Ministry of Health, [29, 32] and McCollum et al., [33].

https://doi.org/10.1371/journal.pgph.0001815.g002

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 3. Kenya CHS management and supervision structure.

Adopted from McCollum et al., [33] and Ministry of Health., [34].

https://doi.org/10.1371/journal.pgph.0001815.g003

Current Kenyan studies on CHVs have focused on nomadic, rural, and peri-urban areas with less attention given to UIS [35–42]. Drawing on the Kenya CHS as the ‘intervention’ or ‘programme’ under implementation; this study, therefore, aimed to explore factors that affect CHV performance in UIS within Nairobi, Kenya. This study also aimed to identify and explore ways in which CHVs can be better supported to enhance their wellbeing and strengthen community strategies.

Methods

Results

This section begins by presenting the general characteristics of the study participants, followed by the factors that influenced the CHVs’ performance as outlined in the Kok et al., [2015] framework.

General characteristics of study participants

The study involved 126 participants, including two sub-county officials, 14 community health assistants, 77 CHVs and 33 community health committee (CHC) representatives. The participants varied in age ranging from 22 to 63 years old. Of the 77 CHVs, 16 had primary school education, 34 secondary school education, and 27 had tertiary level education (24 diploma and three university level education). Length of service varied from one year to as many as 25 years in service. Table 1 below gives additional information on the study participants’ characteristics.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 1. Characteristics of study participants.

https://doi.org/10.1371/journal.pgph.0001815.t001

Discussion

This study explored factors that affect CHV performance in urban informal settlements within Nairobi Kenya and ways in which CHVs can be better supported to enhance their wellbeing and strengthen community strategies. The key programme design factors identified as influencing the performance of CHVs working in UIS included: CHV recruitment; training; the availability of supplies and resources; and the remuneration of CHVs. Health system factors that influenced the CHVs performance included: the nature of relationship between healthcare workers at local referral facilities and community members; availability of services and perceived corruption at the referral facilities; and CHV referral outside of the local health facility. Broader contextual factors that affected CHV performance at the community level included: demand for material or financial support; perceived corruption in community programmes and neighbourhood insecurity. These are discussed in turn below with some suggested recommendations for tackling the issues.

The voluntary nature of the CHV position influenced the kind of CHVs who were recruited and their availability to carry out their duties as assigned by their supervisors. This mirrors findings from other contexts. For example, Sarma et al., George et al., Aseyo et al., and Osindo et al. in their work conducted in Bangladesh, India, and Kenya respectively. In Sarma’s work although conducted in a rural setting, the availability of paid assignments made it difficult for the BRAC’s local office to find volunteer workers [48]. Whereas, in George, Aseyo and Osindo’s work, CHVs working in UIS found it difficult to devote time to voluntary work without any government allowances. This is because they faced additional pressure to provide for their families and therefore had to look for alternative sources of income. As such, some of them fitted their CHV-related tasks around normal daily income-generating activities [36, 49, 50].

In Bangladesh, Alam and others noted that a regular salary or income was critical to improving the CHVs’ level of activity [51]. Relatedly, to control for attrition, Bhutta et al., recommend regular, performance-based financing (PBF) incentives or hiring CHVs as full-time waged employees [52]. Nonetheless, given the significant resource and system challenges experienced in settings such as Kenya, sustainability of such financial incentives are likely to face challenges, including irregular or insufficient disbursement of payments [1]. Moreover, evidence from literature suggests that approaches like PBF incentives sometimes result in over reporting [53] or neglect of unpaid tasks by CHVs [2]. Therefore, for such incentives to work, there is need for a mix of both financial and non-financial incentives, robust accountability measures, and a consideration of context. It is, however, noteworthy that Nairobi County, Kenya is in the process of establishing a performance-based incentive (PBI) for CHVs. Under this initiative, CHVs will be issued with a monthly stipend of 3,500 Kenya shillings of which 500 shillings will go towards their medical cover. These payments will be in exchange for eight days of work in a month and subject to the CHVs successfully achieving an 80 percent performance target [54]. It remains to be seen how this approach will work. Further work is therefore needed to explore the effect of this PBI on the CHVs’ motivation, job satisfaction, attrition, and performance.

Although money was noted as a key motivator for CHVs working in UIS in this study, other non-financial incentives were also observed to play an important role. In particular, CHV appreciation, respect and recognition were seen to be powerful motivators in continuity of the volunteer CHV role. This finding concurs with a review by Jaskiewicz et al., [55] and studies by George et al., Laston et al., and Alam et al. [50, 51, 56, 57]. In their studies conducted in Ethiopia, Malawi, Zanzibar, Mozambique, Zambia, Nigeria, Burkina Faso, Iran, Pakistan, Nepal [55], India [50] and Bangladesh [51, 56, 57] these authors found that social prestige and positive community feedback to CHVs were important non-financial factors associated with CHV activity; and should therefore be considered by programmes working in UIS.

As with the study conducted by Kok and colleagues [2], this study found that inconsistent CHV training; inadequate supportive supervision; and lack of, or insufficient, CHV resources and supplies negatively impacted CHV motivation and job satisfaction. This in turn adversely affected CHVs’ wellbeing and performance. This finding concurs with that of several other authors that have conducted work in LMICs including Aseyo et al., George et al., Goudet et al., Kok and Muula, Karuga et al., Laston et al., Martinez et al., Osindo et al., Odhiambo et al., Sarma et al.; and systematic reviews by Jaskiewicz et al. and Kok et al. [2, 36, 48–50, 55, 57–62]. For example, Aseyo et al., George et al., and Sarma et al., found that inconsistent training left CHVs with knowledge gaps related to the technical aspects of their roles, thus compromising their credibility. This was especially evident when CHVs were not able to respond to (health and other) questions asked by community members [36, 48, 50].

As in this study, work conducted in other resource-poor settings also show that inadequate supportive supervision of CHVs and insufficient resources and supplies; have also been linked to loss of credibility and subsequent demotivation of CHVs resulting in poor performance [2, 36, 48, 49, 55, 60, 61, 63]. Previous work in Kenya by Karuga et al., concurred with the findings of this study on the reasons for inadequate supportive supervision [60]. Karuga et al. found that the main reasons for insufficient supportive supervision were heavy workload on the part of supervisors, inadequate training by trainers who also act as supervisors, and inadequate resource inputs. According to Karuga et al., CHEWs supervised many CHVs, which affected their ability to provide adequate regular supervision. Therefore, to boost CHV supervision in UIS, CHV programmes need to adopt an approach that takes into consideration the trainers’ workload, the voluntary nature of the CHV position, and the limited resources in terms of operational support.

In this study, sub-optimal health systems and related challenges also contributed to loss of credibility for CHVs. This finding resonates with observations by George et al., Sarma et al. and a systematic review by Jaskiewicz et al. [48, 50, 55]. According to Jaskiewicz and others for instance, community trust or respect was dependent on many factors within the community, with the health facility as an institution having a higher degree of influence [55]. Both George et al., and Sarma and colleagues noted that community trust or respect was an important factor to the success of CHV programmes; and when compromised, it rendered CHVs ineffective. As in this study, George, Sarma and colleagues, and the systematic review by Jaskiewicz and others found that sub-optimal healthcare worker-community members interactions, drug stockouts, service unavailability, and perceived corruption in referral facilities; led to loss of community trust in the CHVs. This in turn resulted in a preference by community members for private clinics over government facilities; and subsequently demotivated CHVs since they were evaluated based on referrals made to public health (link) facilities [64, 65].

Another health system challenge reported to demotivate and negatively affect CHV work was poor interactions between the CHVs themselves, and referral facility healthcare workers. In this study, this was particularly prevalent when CHVs referred or accompanied community members outside of their usual locality due to lack of clarity on the CHV roles. According to Kok et al., attitude of healthcare workers towards CHVs had a major effect on how they felt and performed [2]. Negative attitudes from formal healthcare workers towards CHVs resulted in communication challenges, and adversely affected the community-health facility referral process [66]. This suggests a need for awareness raising amongst relevant stakeholders to appreciate the role played by all cadres within the health system, clarity of roles and remit for CHVs, and potentially also communication training for formal healthcare workers to improve interactions with both community members and CHVs. Some of the training that the formal healthcare workers can undergo is the communication skills and emotional competence training [67]. With this training, the formal healthcare workers will gain knowledge on the importance of communication skills and emotional competence in interacting with colleagues.

In this study, the demand for financial or material support, perceived corruption in community programmes, and insecurity hampered CHVs’ service delivery to those living in UIS. These mirror findings from earlier studies conducted in Nairobi and other Kenyan counties by Aseyo et al., Osindo et al., and Odhiambo and colleagues [36, 49, 62]. In these studies, the community expected tangible support from the CHVs. As such, out of sympathy, some of the CHVs used their own financial resources to provide this support to community members, which also led to community acceptance of them. Those CHVs who were unable to meet such demands lost favour especially with needy households, resulting in CHV dropout. This emphasizes the need for clarity of roles including community sensitization of the same [2]. This also suggests a need to integrate CHV activities with government social support programmes, that provide assistance to the very needy and shield CHVs from unrealistic and unfeasible community demands. Such social support programs would, however, require proper monitoring and accountability checks to ensure that only genuinely needy community members benefit.

Safety of CHVs especially at night was also noted to be of concern in this study. Indeed, this particular finding seems to be of greater significance in urban informal settlements (including in Kenya) compared to rural and peri-urban contexts. For example, reviews by Glenton et al., and Ogutu et al., [22, 68] showed that safety among CHVs working in UIS was of great concern especially due to high poverty levels in these settings, and other social challenges including criminal activities in the local areas. In Brazil for example, CHVs lived and worked in dangerous neighbourhoods with high criminal activities such as prostitution, drug use and violence. Because of this, they felt insecure working in such neighbourhoods for fear of being victims of aggression [69]. In previous Kenyan studies it was noted that insecurity hampered the CHVs’ access to certain neighbourhoods or working late in the evenings consequently blocking CHVs from reaching their full potential [49, 62]. It is therefore pertinent to boost safety and security measures in settings such as where this study took place. Some of the initiatives to achieve this include providing robust and corrupt-free national and local community policing including responsive emergency hotlines as needed. One example of an initiative that holds promise in such settings is the nyumba kumi initiative [70]. The Nyumba Kumi initiative is a community-oriented policing policy that was introduced in the year 2013 by the government. This policing concept was borrowed from Tanzania and is closely linked to the Africa philosophy of Ubuntu or Undungu—meaning brother’s keeper. Though this initiative anchored at the household level, the policy focuses on getting residents to know each other and establish communication structures among themselves as well as with the police. Among the key responsibilities of Nyumba Kumi include sharing information, and identifying problems affecting residents. These problems could be issues concerning insecurity, the environment, culture, and the like, and jointly solving problems that might result in social disorder. Other initiatives include provision of adequate ambulatory services to avoid CHVs having to accompany community members to facilities at night while on foot or motorbikes; tackling social challenges such as high unemployment and drug and alcohol abuse that contribute to crime; ensuring proper and adequate amenities such as local street lighting and good road networks are provided by the relevant national and sub-national authorities; and maintenance of relevant infrastructure. These efforts would require intersectoral and intergovernmental cooperation with both the national and county governments playing a role supported by relevant private and NGO stakeholders.

Limitations of the study

One key limitation of this study was that it was conducted during a transition phase, when a new CHS (2020–2025) was being rolled out. At the time of conceptualizing the study, this transition had not been anticipated. This meant that the research team had to take into consideration changes in both the new and outdated policy. To mitigate against this, the research team took this change into consideration when designing the data collection tools and factored in interview questions that referred to both the new and outdated policies during discussions.

Additionally, this study was conducted during the COVID-19 pandemic, and subsequently required adherence to COVID-19 safety protocols and restrictions. These always included social distancing and wearing facemasks which was challenging given the study contexts i.e. urban informal settlements. These COVID-19 safety protocols potentially also affected the quality of audio-recordings, especially where the background environments were noisy (given that respondents had to be masked and sit far apart from each other). To mitigate against this challenge, the research team did two interviews for each category of participants. The research team also asked the participants to raise their voices while speaking, and where necessary—especially for soft-spoken participants–to lower their masks only while speaking for audibility. For the soft-spoken participants the interview moderator also repeated their responses to enable capturing by the audio-recorder and to facilitate discussion and responses from others present. The research team also did a thorough quality check of the transcripts as explained in the data analysis section above.

Conclusion

Like in other settings such as the rural and peri-urban settings, community health volunteers working in UIS in Kenya face a myriad of challenges that impact on their wellbeing and work performance. These cadre of health workers can be better supported by being financially remunerated in addition to other non-financial incentives. Other programme design requirements that are vital for CHVs’ wellbeing include timely holistic training, adequate supportive supervision and sufficient resources and supplies to undertake their job roles. Additionally, there is need for clarity of roles and scope of work, training of healthcare workers on appropriate relations with both the community and CHVs, ensuring availability of services, and putting strategies in place to safeguard against corrupt practices in public health facilities and community programs. In addition, specifically for the urban poor context, there is an urgent need for improved and adequate security measures at the community level, to ensure safety of the CHVs as they undertake their roles.

Acknowledgments

We acknowledge immense support from the Ministry of Health and in particular the Division of Community Health Services. Special thanks to the Nairobi County CHS coordinator and both Mathare/Ruaraka and Lang’ata/Kibra sub-county CHS coordinators for facilitating our community entry and mobilization of both sub-county and community respondents. We also acknowledge the continual support of the Nairobi County Department of Health, and all the study participants for giving their time during the study period.