Timely Access to Maternal, Neonatal and Child Healthcare for rural communities in Rwanda: Job satisfaction of Community Health Workers delivering Community Based Maternal, Newborn and Child Healthcare


 Background

In Rwanda, although there has been progressing in health care delivery as expressed in the reduction in maternal and child mortality, rates are still high and geographically variable. For the improvement of equitable access to health services for maternal, newborn and child healthcare (MNCH), community-based maternal, newborn and child healthcare (CBMNCH) depends on the use of “community health workers” (CHWs). However, the CHW program faces challenges that disrupt the quality delivery of a full package of services. Yet little is known about the satisfaction of CHWs in delivering CBMNCH.
Methods

This quantitative cross-sectional study involved a survey of 500 sampled CHWs delivered CBMNCH in three selected rural districts of the southern province, Rwanda. Ordinal regression was used to examine the determinants of CHWs` job satisfaction.
Results

Multivariate analysis shows that the determinants of job satisfaction were motivation (OR = 8.59, p < 0.001), formal training in CBMNCH (OR = 2.24, p < 0.05), individual supervision (OR = 6.19, p < 0.001), and peer support (OR = 2.66, p < 0.01), knowledge about CBMNCH (OR = 0.51, p < 0.05), access to essential materials (OR = 0.32, p < 0.05), and incentives (OR = 0.53 (p < 0.01).
Conclusion

The findings indicated that the managers of CHW programs and other stakeholders need to improve the working conditions of CHWs to enhance their job satisfaction, to enable the effective provision of CBMNCH.


Abstract Background
In Rwanda, although there has been progressing in health care delivery as expressed in the reduction in maternal and child mortality, rates are still high and geographically variable. For the improvement of equitable access to health services for maternal, newborn and child healthcare (MNCH), community-based maternal, newborn and child healthcare (CBMNCH) depends on the use of "community health workers" (CHWs). However, the CHW program faces challenges that disrupt the quality delivery of a full package of services. Yet little is known about the satisfaction of CHWs in delivering CBMNCH.

Methods
This quantitative cross-sectional study involved a survey of 500 sampled CHWs delivered CBMNCH in three selected rural districts of the southern province, Rwanda. Ordinal regression was used to examine the determinants of CHWs` job satisfaction.

Conclusion
The ndings indicated that the managers of CHW programs and other stakeholders need to improve the working conditions of CHWs to enhance their job satisfaction, to enable the effective provision of CBMNCH.

Background
Poor access to health care services are resulting in high maternal, neonatal, and child mortality and morbidity, especially in middle and low-income countries like Rwanda 1 . Besides, inequitable access to health care services partly explains the disparity in a rural-urban dichotomy in morbidity and mortality in many countries 2,3 . Residents in urban areas tend to have improved health outcomes compared to their counterparts in rural communities 2,3 . In several parts of the developing world, there is a de ciency of skilled health professionals. For instance, sub-Saharan Africa remains below the WHO recommended a limit of 2.3 medical doctors, midwives, and nurses, per thousand individuals 4 . Medical doctors are particularly unevenly distributed among urban and rural communities, with de ciencies being considerably more intense in rural communities 4 . In response, CHWs are being depended upon in some countries around the world to address the issue of de ciency health providers by enhancing the provision of primary health care services and promoting health in rural settings or underserved communities 5,6 . The concept of universal health coverage as espoused by the Alma Atta Commission, and subsequently adopted by the WHO is well placed to reduce inequalities in health access 6 . Some countries are endeavouring to close the healthcare gap through universal health coverage, which guarantees reliable and timely access to improved health access regardless of geography and socioeconomic circumstances 7 . It has, therefore, been argued that enhancing equitable access to quality medical services, needs global and national investments that bring health care services closer to those in most need 8 .
In line with the need to improve universal health coverage to deprived communities, the role of CHWs in rural areas in developing countries cannot be overemphasized. A signi cant priority in health policy is that essential health care services should be accessible geographically: close to where individuals live and work 8 . CHWs are mostly residents in these rural communities. Therefore, the decentralization of healthcare services to remote and rural settings through CHWs has the advantage of geographical proximity and readily accessible within rural settings which also helps bridge socio-cultural and linguistic barriers to health care delivery 9,10 . In a meta-analysis of maternal and child health by Kassebaum et al. 11 , explained the positive impact of CHWs on reducing maternal and child mortality between 1990-2003 in varying contexts. However, although child mortality reduced about half since the 1990s, and maternal mortality dropped 1.3% every year since 1990 11 ; still over 17,000 children are reported to die annually from preventable causes 12 . Stillbirth rates have not signi cantly changed, and many women are losing their lives due to perinatal-related complications 12 .
In Rwanda, a country known as "the land of a thousand hills," in Eastern Africa on a highland plateau averaging 1,200 to 2,000 m in elevation, the concept of CHWs is not new. With a commitment to provide universal healthcare as part of its Vision 2020 Strategy, the Rwandan government has implemented a national CHWs program since 2007 as a bridge between local communities and the health care system 13 . Each village (around 100-250 family units) has one CHW (female), called an "ASM" (Animatrice de santé Maternelle), explicitly focused on follow up of women during pregnancy and after birth, and newborn. They provide Community-Based Maternal, Neonatal, and Child healthcare (CBMNCH) 14 .
Although there has been some progress in health care delivery as expressed in the reduction in maternal and child mortality, rates are still high and geographical variables 15, 16  Since its inception, community-based interventions have been generally seen to have fundamentally contributed towards current health achievements in Rwanda 17 . However, in the same way as other African nations, the CHWs program in Rwanda still faces huge di culties that upset the delivery of the quality of the comprehensive package of services. These di culties extend from the low limit capacity of CHWs to insu cient resources to sustain routine community health activities 17 . Based on this background, to strengthen its CHWs program, Rwanda Ministry of Health in partnership with Western Ontario University introduced in 2016 the "Training, Support, Access Model" (TSAM) project in the six districts with high MNCH needs amongst others include Rulindo, Gakenke, and Gicumbi in the northern province and Muhanga, Ruhango and Gisagara in the southern province. One of the key objectives of TSAM is to improve MNCH through community-based interventions using appropriately trained, mentored, and supported CHWs 18 . This project has implemented in the northern province and is now scaling up in the southern province. Unfortunately, there is little baseline knowledge about CHWs jobsatiaction ∈ theprovisionofCBMNCH, ∈ f or mationtŵod ̲ berequiredbef or eTSAM's ∫ ervention → ⊂ sequentlyexa min etheimpactmad job satisfaction prior to intervention in southern province.

Methods
The study aimed to provide insight into the satisfaction of CHWs in the provision of CBMNCH for timely access to maternal, neonatal, and child healthcare in rural communities, Rwanda. The quantitative cross-sectional study was conducted in three selected districts in the southern province including Gisagara, Ruhango, and Muhanga district (working area of TSAM in the south). The study population comprised CHWs delivering CBMNCH within study areas. With the collaboration between the TSAM project and MoH Rwanda, through Rwanda Biomedical Centre (RBC), the sampling frame was obtained from the RapidSMS database. At the time of the survey, the database included 1388 CHWs who were actively providing CBMNCH in the three study districts. With a con dence level of 95% and α error probability, 0.05 applied to the population of 1388 CHWs, a sample size of 301 CHWs was calculated as the minimum sample threshold for unbiased ndings for this study. Even though the estimated minimum sample size was 301, we oversampled by 200 to give the sample size more power, therefore, a sample of 500 CHWs was used for this study. Because the population size was not equal in each district, at the district level, the sample size was calculated based on district proportionate allocation sampling technique, "probability proportional to size" 19 .
Systematic random sampling was used to select the study sample from the total population of CHWs 19 . For data collection, a comprehensive questionnaire for CHWs was developed. This questionnaire was designed from previous CHW studies in Rwanda and elsewhere and it is, therefore, not a standardized instrument, but it was presented to the team of TSAM experts to discover whether the content is relevant in comparison with the context to ensure its content validity. Then, the research was presented to two Ethics Committees (Western University and the University of Rwanda) for approval. Participation in the research was voluntary. Thus, potential participants had to sign a consent form. The survey was conducted from June 2019 to September 2019.
Job satisfaction is a dependent variable. The Cronbach's alpha was used to generate satisfaction scales by aggregating questions which were asked CHWs to rate their satisfaction on the job as maternal health service providers in rural communities. Scores to these questions were estimated based Likert scale, (5) strongly agree, (4) agree, (3) neutral, (2) disagree, and (1) strongly disagree. Regarding data analysis, we employed univariate, bivariate, and multivariate analyses to understand the factors associated with job satisfaction among CHWs. While our dependent variable (job satisfaction) is ordinally corded (1 = Low; 2 = Middle; 3 = High), therefore, we em0ployed the ordinal logistic regression, which is suitable for an ordered dependent variable. Models were built sequentially. We accounted for structural variables in Model 1 and individual-level variables in Model 2. Findings were reported in odds ratios (ORs) where ORs larger than 1 indicate higher odds of being satis ed on the job, while ORs smaller than 1 indicates lower odds of being satis ed. Overall, CHWs reported they do not have protective materials. As per their mandate, CHWs were sometimes required to accompany clients to a Health Center.

Univariate
The results show that the average travel time to HC was 94 minutes. Also, the mean number of households that CHW is responsible for is 172 households, with CHWs reporting serving an average of 27 clients per month. About 80% of CHWs had over four years of working experience. Only 15.2% of CHWs received in-kind payment for their services. 31.8% of CHWs were members of a local level for pro t cooperative. In terms of socioeconomic status, the results show that 6.4% of the sample are in "Ubudehe" category 1, whereas 40% and 56% are in "Ubudehe" category 2 and "Ubudehe" category 3, respectively. Also, 22% of CHWs were 35years or less, while 39% and 39% were aged between 36-49 years and 50 years and above, respectively. All CHWs received basic education with about 23% of them having more than primary education, while the majority (77%) had primary education level. In terms of primary occupation, 95% of CHWs are smallholder farmers. 87.4% of CHWs were married, and about half of them belonged to households with over seven members.
Bivariate Table 2 shows ndings from the bivariate analysis. Broadly, some structural and individual-level factors were signi cantly associated with increased job satisfaction among CHWs. At the structural level our ndings show that CHWs highly performed on the job (OR=11.91, p<0.001) or highly motivated (OR=15.85, p<0.001), were more likely to have high job satisfaction compared to CHWs poorly performed or with low motivated respectively. Individual supervision (OR=6.53, p<0.001) was associated with higher odds of job satisfaction. CHWs who received formal training on CBMNCH (OR=2.28, p<0.001) and peer support (OR=1.82, p<0.01) had higher odds of reporting high job satisfaction. Lacking assessment tools (OR=0.38, p<0.001) was associated with lower odds of job satisfaction.
Increased years of experience (OR=2.76 p<0.001) were signi cantly associated with higher odds of job satisfaction. Lacking in-kind payment (OR=0.32, p<0.001) was associated with lower odds of job satisfaction. Among the individual-level variables included in the analysis, education was the only signi cant predictor of job satisfaction: those with more than primary education (OR=1.69, p<0.01) had higher odds of job satisfaction compared to their counterparts with primary education.
Multivariate Table 3 shows ndings from multivariate analysis. Findings were largely consistent with bivariate results. At the structural level, we found that CHWs highly motivated (OR=8.59, p<0.001) had signi cantly higher odds of job satisfaction compared to those with low motivation. CHWs highly performed on the job (OR=7.08, p<0.001) had signi cantly higher odds of reporting high job satisfaction compared to CHWs lowly performed. However, CHWS with a high level of knowledge about CBMNCH (OR=0.51, p<0.05) had lower odds of satisfaction compared to their counterparts. CHWs who were supervised once every month (OR=6.19, p<0.001) had higher odds of being satis ed compared to those who never received supervision. CHWs with formal training (OR=2.24, p<0.05) had higher odds of being satis ed compared to their counterparts. CHWs who received peer support (OR=2.66, p<0.01) had higher odds of being satis ed compared to CHWs without peer support. CHWs who lacked assessment tools had lower odds (OR=0.32 (p<0.05) of being satis ed compared to CHWs with all required tools. CHWs who do not receive in-kind payment for their work performance were more likely to be unsatis ed compared to CHWs who received inkind payment (OR=0.53, p<0.01).

Discussion
Health workers' job satisfaction plays a key role in the quality of health service delivery. Given the speci c responsibilities of CHWs in the context of Rwanda where they ll an important gap in health service delivery, understanding the predictors of their job satisfaction is key in proffering policy solutions for improving the quality of their work. Overall, CHWs satisfaction was found to be signi cantly associated with mainly structural level factors including level of motivation, role performance, individual supportive supervision, formal training, access to required materials, peer support and in-kind bene ts from the community, among others. This nding emphasizes the important role of overarching structural factors in guaranteeing CHWs job satisfaction in the context of Rwanda. However, overall, the results of this study indicate that the proportion of CHWs who are highly satis ed with their work is more than those who may be dissatis ed. These ndings were found to be largely consistent with the studies of Ding et al. 24 and Mpemberi et al. 25 also emphasize the role of structural factors in guaranteeing the job satisfaction of health workers.
Given the evidence of a link between job satisfaction and performance, researchers often use workers' performance as a proxy to measure their job satisfaction as workers who are high performers also tend to report higher job satisfaction 26,27 . Consistent with this observation and a recent study by Khatri, Mishra, and Khanal 28 , the middle and high performing CHWs workers in this study had a higher likelihood of reporting job satisfaction compared to CHWs who reported low performance. Earlier, studies 29,30 have highlighted a bidirectional relationship between job satisfaction and performance where 11.8% of health workers' performance was explained by their job satisfaction. In this regard, in evaluating CHWs job satisfaction in Rwanda, stakeholders can achieve this through performance appraisals, and where CHWs performance is seen to be below a set target, appropriate measures can be put in place to address any outstanding issues that may be adversely affecting their job satisfaction.
Furthermore, the nding that motivation was a signi cant predictor of CHWs satisfaction may not be too surprising. Speci cally, community health workers with middle to high motivation were more likely to belong to a high category of job satisfaction compared to those with poor motivation. This nding is consistent with the current literature`s suggests that motivation, whether intrinsic or extrinsic, plays a key role in health workers' job satisfaction 25,34−37 .
Besides, Lambrou et al. 31 also found that intrinsic factors measured by internal thought processes and perceptions about motivation, and extrinsic factors measured by monetary rewards and recognition for work done, greatly in uenced health workers' job satisfaction. However, CHWs` motivation and its link to their job satisfaction could also be due to the overwhelming community recognition of their work 16,33 . Similar links were revealed by a study conducted by Liverpool School of Tropical Medicine Centre for Maternal and Newborn Health, in collaboration with UNICEF and Rwanda Biomedical Center (RBC), where most CHWs exhibit intrinsic motivation as recruitment into the program was mostly voluntary with no nancial compensation. In this regard, only individuals with a desire to help address persistent maternal and child health care challenges in the community opted to be trained and commissioned as o cers despite insu cient remunerations 32 . It is therefore imperative for the Rwandan government to effectively harness this high level of motivation to ensure the delivery of high-quality health services in rural communities by CHWs.
The ndings further revealed that CHWs with high knowledge about their primary mandate, speci cally, maternal, newborn and child health care were less likely to report job satisfaction. Although this nding may seem counter-intuitive, it is possible that CHWs with high knowledge of their mandates and conscious of the important function they serve in the healthcare delivery chain are dissatis ed with some existing ine ciencies impeding their ability to effectively discharge or execute their mandates in reducing maternal and child mortality. Consistent with this observation, Mathauer and Imhoff 34 have revealed that, where health workers are unable to perform their duties because of bureaucracies and other delays in accessing the necessary tools to perform their duties, they become frustrated and often report job dissatisfaction.
Both quantitative and qualitative literature have discussed the importance of supportive supervision on job satisfaction among health care workers in several contexts [38][39][40] . According to previous studies, the supervision of community health workers in developing countries is critical to ensure that they perform well, deliver quality services and be motivated [41][42][43] . Consistent with these scholarships, ndings from this study revealed that lack of supervisory support negatively affects CHWs w or ksatiaction. Thus, CHWswithat ≤ * o ≠ ⊃ p or tive ⊇ rvisionamonthwerem or elikely → be ∈ ahighcateg or yofw or ksatiactioncom decisions and feelings as they perceive their work is valued and appreciated and through this, enhance their greater work satisfaction 44 . In addition to supervision from superiors, Hill et al. 38 have also emphasized the importance of community and peer supervision for CHW which was also found to be associated with improved work performance. Earlier studies in Rwanda identi ed su cient supervision as a major barrier affecting effective service delivery by CHWS 16 . Thus, this nding suggests that stakeholders of the CHWs program might need to pay particular attention to this critical to improving supportive supervision.
In addition to supportive supervision, training of CHWs before the commencement of their duties and other in-service training is seen as particularly useful in enhancing their knowledge and skills for service provision. Training as an indispensable tool in the work of community health workers is useful in the transfer of useful skills and information for the effective delivery of health services to hard-to-reach populations 44,45 . CHWs, therefore, consider training as an essential component in achieving their mandate as health workers. Among others, CHWs explained that more frequent training improved their e ciency, con dence, and knowledge base as most of them are not originally trained as a health professional. They feel empowered and respected within their respective communities when they receive training from superiors who are active health service professionals with many years of work experience 16 . To this regard, it may not be surprising that in this study CHWs who received formal training in CBMNCH were more likely to be in a high category of job satisfaction compared to those who have never received formal training in CBMNCH. Therefore, CWHs who received limited or no training were less likely to be satis ed with their job given its adverse in uence on the effective delivery of their mandates as community volunteers. These ndings further suggest that stakeholders in the CHWs program should focus on making the training of CHWs more frequent to give them opportunities to improve their knowledge and skills. This will likely lead to better performance of their assigned tasks which could also improve their feeling of accomplishment (satisfaction) from their work.
Furthermore, in the health delivery literature, peer support is a crucial factor in the retention of health workers as they share knowledge and discuss how to surmount challenges in the performance of their daily duties. In this study, peer support, as one of the facets of job satisfaction, was also examined and it was found to be a predictor of job satisfaction for CWHs. CHWs who had good peer support were more likely to be in the high category of job satisfaction compared to their counterparts. These results were consistent with other studies such as Jayasuriya et al. 46 , who found relationships with colleagues and other forms of peer support to be a strong predictor of job satisfaction. Similarly, other scores of scholars 47,48 from a qualitative enquiry approach have shown inter-personal relationships as an important ingredient in health workers' motivation. This nding is very useful for policy consideration among CHWs stakeholders as they can target strengthening peer support activities as this is currently not implemented in the study context.
Access to working materials and other essentials is necessary for meeting targets and effective discharge of responsibilities for health workers particularly in the context of developing countries. Therefore, although nancial rewards are important for motivating, retaining and ensuring health worker satisfaction, the presence of adequate resources in the form of supplies and essentials is very useful in improving the morale and work satisfaction of health workers signi cantly 49 . In this context, it may not be too surprising for this study revealed that CHWs who had limited access to assessment tools which is important in the discharge of their duties were less likely to report being satis ed with their jobs compared to those that had regular access to these materials. Other studies 50,51 have reported the same ndings. These suggest that it is necessary to su ciently equip CHWs to perform their work and that in turn could improve their work satisfaction.
While nancial rewards and motivation are linked to health worker job satisfaction and retention in many contexts 33,42,43 , in Rwanda, the community health worker program is mainly voluntary and based on little or no remuneration. However, CHWs are encouraged to form cooperatives where they initiate income Loading [MathJax]/jax/output/CommonHTML/jax.js generation activities and pro ts from these initiates may be used by CHWs as nancial compensation for their work. Thus, cooperatives serve as the main source of nancial remuneration for CHWs. The ndings reveal that CHWs who perceived their cooperatives to be pro table, implying they may be gaining some nancial rewards from their cooperatives were more likely to report job satisfaction compared to those who did not belong to a pro table cooperative. In this regard, it can be argued that although the CHW program was established voluntarily, nancial remuneration may still be playing a key role in CHWs job satisfaction as reported by earlier studies 25,43,52 . Furthermore, CHWs who received payment in kind for their services were more likely to have job satisfaction compared to those that were not receiving any payments from community members. Given the voluntary base of services rendered by CHWs to their community, payments in kind make CHWs feel appreciated for their work, explaining why they are more likely to have better job satisfaction 16,34 . Based on these ndings, it may be critical for stakeholders of the CHW program in Rwanda to rethink how they can provide a suitable nancial incentive to CHWs to engender strong feelings of governmental support and in turn, better job satisfaction.
In general, work satisfaction revolves around feelings and attitudes that an individual has with regards to their work that motivates them to ful l an anticipated target or achievement 52 . Given the multiplicity of factors that are associated with health workers' job satisfaction, this study argues that in the context of Rwanda, CHWs can be satis ed with some aspects of their job and the same time remain dissatis ed with other aspects that fail to meet their expectations. Therefore, this study suggests a holistic approach in considering all the possible factors associated with work satisfaction of CHWs.
Whilst this study provides greater insights into the CHWs' job satisfaction in the provision of CBMNCH services for timely access to MNCH in rural communities Rwanda, there are some limitations. First, the study was conducted in high need districts, making it di cult to extrapolate the nding to the rest of the country. Second, given that respondents were asked to recall most of the responses, some of the responses may be subject to recall bias which may in uence the reliability of the collected data. However, the likelihood of recall bias was reduced by reducing the recall period to a maximum of 12 months.
Lastly, given the cross-sectional nature of this data, ndings are only restricted to statistical associations and therefore, causal-effect could not be inferred.

Conclusion
The ndings of this study provide in-depth insight into the current job satisfaction of CHWs in the provision of CBMNCH in rural communities, Rwanda. The ndings indicated that there a high need for the managers of CHW programs and other stakeholders to improve the working conditions of CHWs to enhance their job satisfaction, to enable effective provision of CBMNCH. This study was conducted before the intervention of the TSAM project in the study areas. Therefore, it would be useful to do a post-intervention analysis to do a comparative study examining the impact of TSAM intervention on job satisfaction of CHWs regarding the provision of CBMNCH in Rwanda. Ethical approval for the study was obtained from both the Research Ethics Board at the University of Western Ontario, Canada, and the College of Medicine and Health Sciences Institutional Review Board (IRB), at the University of Rwanda. The authorization to conduct the study locally was granted through the districts` administrative authorities. Before the study, informed written consent was obtained from CHWs with the option to withdraw at any time during the study.

Consent for publication Not applicable
Availability of data and materials The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.

Competing interests
The authors declare that they have no competing interests Funding The Training, Support, and Access Model (TSAM) project provided the fund regarding only logistic issues to facilitated eldwork activities (data collection).
There was no funding body in the design of the study, analysis, interpretation of data and writing the manuscript.

Authors' contributions
All authors designed the full protocol of the study. JBB collected, analyzed, and interpreted data and wrote the manuscript. IL read and approved the nal manuscript   Standard errors in parentheses * p < 0.05, ** p < 0.01, *** p < 0.001