Effects of community performance-based financing on community health workers’ service delivery in Kayanza health district, Burundi

International Journal of Community Medicine and Public Health Gervais M et al. Int J Community Med Public Health. 2023 Jan;10(1):21-28 http://www.ijcmph.com pISSN 2394-6032 | eISSN 2394-6040

DOI: https://dx.doi.org/10.18203/2394-6040.ijcmph20223521

Original Research Article

Effects of community performance-based financing on community health workers’ service delivery in Kayanza health district, Burundi

Manirakiza Gervais1*, Alfred Owino Odongo1, Apollo Maima2

1Department of Epidemiology and Biostatistics, School of Public Health, Mount Kenya University, Kenya 2School of Pharmacy, Maseno University, Kenya Received: 08 August 2022 Revised: 16 September 2022 Accepted: 30 November 2022 *Correspondence: Dr. Manirakiza Gervais, E-mail: mangev4@yahoo.fr Copyright: © the author(s), publisher and licensee Medip Academy. This is an open-access article distributed under the terms of the Creative Commons Attribution Non-Commercial License, which permits unrestricted non-commercial use, distribution, and reproduction in any medium, provided the original work is properly cited.

ABSTRACT Background: In many low- and middle-income countries, community services are often provided on a volunteer basis. To make their work more effective in the communities, some of them including Burundi adopted community performance-based financing as form of motivation of community health workers (CHWs). The study aimed to demonstrate the relationship between the community performance-based financing and health service delivery by the community health workers. Methods: It was a cross-sectional study using a comparative approach to assess the relationship with the performance-based financing scheme and the community services delivery. We used a stratified random sampling and inferential statistics used chi square and logistic regression. Results: The occupation of CHWs was significantly associated with counselling for early ANC (p=0.002), women attending early ANC and 4 ANC visits (p=0.000). Female were about 3 times likely to refer FP clients for FP methods uptake [p=0.043, CI= (1.030; 7.462), OR=2.773]. Farmers were 11 times as likely to perform this task as CHWs with other functions [p=0.000, CI= (3.890; 32.733), OR=11.284]. The CHWs who received 9 equipment and above [p=0.037, CI= (1.101)]. Conclusions: The community performance-based financing has a significant relationship with the community health workers service delivery and should be enhanced by trainings, formative supervision, and provision of equipment. Occupation was identified as to predict this association. But further studies be conducted to establish other factors that complement the community performance-based financing in community services delivery. Keywords: Community health workers, Community health worker’s Cooperative, Community performance-based financing

INTRODUCTION

In the line with the world health organization (WHO) which suggested that the selection of community health workers should consider the preservice education, and health system, governments in low-and middle-income countries (LMICs) have adopted health policies which aim to increase the utilization of health services and improve nature of wellbeing administrations, among them, the performance-based financing.1 In Benin, the

performance-based financing (PBF) was started in 2012 by a world bank project. Execution of performance-based financing was seen as promising as far as impact, possession, and supportability are concerned; and less assets burning-through. In any case, numerous nations are judged by the incorporation of the methodology in ordinary working of nearby wellbeing frameworks.2

The experience from Uganda showed that after the start of PBF, there was a positive trend in performance of the

International Journal of Community Medicine and Public Health | January 2023 | Vol 10 | Issue 1 Page 21

Gervais M et al. Int J Community Med Public Health. 2023 Jan;10(1):21-28

PBF-implementing health facilities relative to the controls. There was a positive trend in performance in the selected indicators in the PBF health centers (H/Cs). Nevertheless, a rigorous, regular, and independent data verification plan built within the implementation process has been highly recommended.3

Rwanda is one of the pioneers in implementing PBF beginning with a pilot phase in Cyangugu and Butare provinces. It was started by remunerating facility staff based on services delivered and by empowering them to identify creative ways to improve those services. Results from this initial phase showed improvement in terms of coverage, quality, and impact on health service delivery.4

In Burundi, after the government adopted the free care for childbirth and pregnant women in 2006, the workload for health workers increased significantly, but without commensurate increase in staff numbers. This affected the quality of, especially, maternal, family planning (FP) and reproductive health services. To solve this problem, a pilot phase of community performance-based financing was set up in southern province of Makamba.5,6 This study assessed the effect of performance-based financing in delivering community health services. Results from the study will hopefully inspire decision makers, health district leaders and health professionals to set proper strategies to work with CHWs in improving community health indicators. The findings herein should help public health leaders from the Ministry of Health to review policies and guidelines on CHWs’ work and incentives.

The objective of the study was to investigate the effects of community performance-based financing on CHW’s service delivery in Kayanza health District, Burundi.

METHODS

Study area and population

The cross-sectional study was undertaken in Kayanza and Matana health districts in Burundi. Kayanza health District is in Kayanza health province, north of Burundi, bordered by the Republic of Rwanda. Matana health district, the comparative study area, is in Bururi province, south of Burundi. The study aimed to demonstrate the relationship between the community performance-based financing and health service delivery by the community health workers. The population was constituted of 634 CHWs, 340 working in Kayanza health district and 294 working in Matana health district. The CHWs were from 15 health centers of Kayanza Health district and 23 health centers of Matana health district.

Study period

The study focused on the period of the implementation of community performance-based financing in Kayanza health district, beginning from January 2019 up to June

2020. This took into consideration a period of 18 months (one and half years).

Sampling, data collection and tools

Stratified random sampling method was used to select the participants (CHWs) from the 38 health centers in the health district of Kayanza (intervention area) and Matana (control area), and the commune entity constituted the stratum in the study. The size (234) for the study was constituted by sample sizes from each health district so that the sample size each (stratum) commune was proportionate to the population (CHWs) to the stratum. The research was a cross-sectional study. The study explored the level of services delivered by CHWs in terms of counselling, referral, and community-based distribution of commodities in maternal, child and reproductive health.

Quantitative data was collected using a pre-formulated questionnaire, and reports and registers were analysed to evaluate the level of achievement of services by CHWs. It was a structured questionnaire, guided on how to answer to avoid ambiguity. The researcher met respondent (CHWs) at health center where he also collected secondary data from registers and reports. The focus group discussions (FGDs) within which community health workers’ supervisors and health providers were involved were conducted to give their appreciation on the level of leadership within CHWs’ groups and the perception of CHWs on the in-kind incentives received by CHWs and their influence on the performance. Key informant interviews were conducted where FGD were not appropriate.

Variables

The services delivered by CHWs at community level were designed in dependent variables scored on level of the target reached or not. For maternal health, nine variables were created. The visit for pregnant women, the counseling for early first antenatal care, at risk pregnancies referred to health facility, and the early ANC attendance were evaluated. The study also evaluated the attendance for 4 standards of ANC. First, deliveries carried out at health facilities by health professionals, upon clients being accompanied by CHWs. Secondly, home deliveries referred immediately to health facilities by CHWs, then early post-partum visits as well as obstetrical fistula. For child health, services on the recovery of abandoned vaccination were evaluated. For reproductive health and FP, four variables were created such as the counseling on reproductive health services, referral for FP methods, commodities distributed in the village by CHWs and the recovery of abandoned use of contraceptive methods. Additionally, there was a variable of whether CHWs referred patients to health facilities for other health problems.

International Journal of Community Medicine and Public Health | January 2023 | Vol 10 | Issue 1 Page 22

Gervais M et al. Int J Community Med Public Health. 2023 Jan;10(1):21-28

The independent variables were grouped in two components. One in the socio-demographic characteristics with variables of age divided into 20 to 34 years, 35 to 49 years, and 50 years and above, gender (Male and female), and religious affiliations divided in to Catholic, Protestant, Muslim and Adventist. Then there was occupation divided into famer, farmer and other jobs, and other jobs; marital status divided into single, widow/widower, married and divorced; and the level of education divided into primary, secondary, college and university. The second is the in-kind motivation divided into trainings and supervisions to CHWs and the equipment distributed to CHWs (umbrella, raincoat, registers, bicycle, bag, t-shirt), and leadership of CHW’s groups.

Analysis

Data was cleaned and entered SPSS version 21 for analysis. Descriptive statistics were computed and the inferential statistics using chi-squared test to evaluate the co-relationships between independent and dependent variables. The logistic regression using the STATA was conducted to identify the influence of independent variables on the dependent variables.

Ethics approval and consent to participate.

The study involved human participants. It was approved by the Mount Kenya university ethics committee ref

MKU/ERC/1834, Number 907. Before taking part in the study, participants gave informed consent in the participation. Participant were required to sign consent form before being in the research. During data collection and analysis, an identification code replaced personal and identifying information such as names, and the final report to protect identity and to avoid disclosing personal and identifying information. Through signing the informed consent, they were assured that information given was solely for the purpose of the study.

RESULTS

Socio-demographic characteristics of participants

The Table 1 summarizes the results from sociodemographic characteristics analysis. The majority of CHWS were aged between 36- 50 years (Matana: 53.2% and Kayanza: 65.6%). Most participants were female (51.4% in Kayanza and 59.2% in Matana) and most of them were married (91.7% in Matana and 98.4% in Kayanza). The CHWS with primary school were 40.4% in Kayanza and 80.8% in Matana had, 44% in Kayanza and 17.6% in Matana had college level, while those with secondary school were 15.6% in Kayanza and 0.8% in Matana. Those whose occupation was farmer were 67.9% in Matana and 48.8% in Kayanza while 51.2% were combining farming and business. Most of participants were Catholics (61.5% in Matana and 75.2% in Kayanza).

Table 1: Socio-demographic characteristics of participants.

International Journal of Community Medicine and Public Health | January 2023 | Vol 10 | Issue 1 Page 23

Gervais M et al. Int J Community Med Public Health. 2023 Jan;10(1):21-28

Association of different variables and achievement

Analyses were conducted with logistic regression and results from Table 2 shows that gender, function, equipment, and non-financial motivation were not significant predictors of service delivery by CHWs in terms of visiting all pregnant women at home. On the other side, Table 1 shows that female CHWs were about 3 times as likely to refer FP clients for FP methods uptake [p=0.043, CI= (1.030; 7.462), OR=2.773], and CHWs whose occupation was famer were 11 times as likely to perform this task as CHWs with other occupations [p=0.000, CI= (3.890; 32.733), OR=11.284]. In terms of home, deliveries referred immediately at the health facility, Table 1 shows that CHWs with farming as occupation were as likely to perform as others [p=0.003, CI= (0.005; 0.344), OR=0.042] while CHWs who perceive equipment and non-financial motivation provided to them to be high are about 5 times as likely to perform the task of home deliveries referred immediately at health facility [p=0.037, CI= (1.101; 22.429), OR=4.969].

Association of occupation and achievement of CHWs

Table 3 shows a significant relationship between the occupation of CHWs and visit of all pregnant women 3 times in the village (p=0.001), women counseled for first antenatal care (p=0.002), women who attended the early ANC before 3 months (p=0.000), women who completed 4 standards ANC (p=0.000), FP clients referred for FP

methods uptake (p=0.000), home deliveries referred immediately to a health facility (p=0.001), FP commodities distributed to women and men in the village by CHWs (p=0.000), recovery of abandoned vaccination (p=0.003), recovery of abandoned use of contraceptive methods (p=0.007).

Comparison of performance of community health workers’ services delivery at Kayanza and Matana health districts

Table 4 shows that the counseling on reproductive health services, counseling women on early antenatal and early attendance of ANC (before 3 months) were highly associated with implementation of CPBF (p=0.000) in Kayanza health district than in Matana health district. The similar results were noted on the indicators on health facility delivery, FP clients referred for FP methods uptake, and home deliveries referred immediately to a health facility, women presented in post-partum consultation within 15 days after delivery; where FP commodities were distributed to women and men in the village by CHWs; and to pregnant women who received prevention against tetanus (p=0.000).

Overall, the results from the chi-square analysis revealed that there was a significant association between the performance level of CHWs and the CPBF implementation scheme in Kayanza Health district (p=0.000) for all variables), where it has been implemented.

Table 2: Logistic regression on various variables.

International Journal of Community Medicine and Public Health | January 2023 | Vol 10 | Issue 1 Page 24

Gervais M et al. Int J Community Med Public Health. 2023 Jan;10(1):21-28

Table 3: Association of occupation and achievement of CHWs.

Perception and opinions about leadership in CHWs’ groups

Key informants interviewed gave various factors that could increase performance in service delivery by CHWs apart from the community performance-based financing. These included regular training, materials, and support from local administration and health officials. They stated thus: “quarterly refresher training provided by our health facility providers and materials provided with the district team used to keep us knowledgeable on how to provide family planning services” (KII with CHWs leader in Kayanza). They further observed that good leadership within the cooperative was important in keeping CHWs motivated to perform desired tasks, adding that: “this is essential in implementing the income-generating activities for our cooperatives, enabling us to fulfill our domestic needs” (Female CHW in Matana during KII).

Focus group discussions revealed that other CHWs’ incentives came from the psychosocial satisfaction that

they got from the satisfaction of community members with their performance, a fact that made the population to pay for their services. “The consideration of our work by the community shows how our contribution was important in improving community health, and the population pays courtesy in our activities” (FGD with supervisors of CHWs).

Finally, the workload of each CHW was a hindrance to their performance. Given the planning and monthly activities of CHWs’ groups, viewed against what they could potentially perform within the community and the need to work for their daily subsistence, some activities were abandoned. To overcome the ensuing challenges and for better performance by CHWs, there was need for good leadership that could coordinate all planned activities and services considering that the CHWs also needed to cater for their families “suppose our group’s leadership is not strong enough to coordinate our activities well. We could change it because we also must fulfill the family’s duties apart from being CHW” (KIs on leadership in CHWs’ groups.

International Journal of Community Medicine and Public Health | January 2023 | Vol 10 | Issue 1 Page 25

Gervais M et al. Int J Community Med Public Health. 2023 Jan;10(1):21-28

Table 4: Comparison of effects of CPBF for CHWs in Kayanza and Matana health district.

International Journal of Community Medicine and Public Health | January 2023 | Vol 10 | Issue 1 Page 26

Gervais M et al. Int J Community Med Public Health. 2023 Jan;10(1):21-28

DISCUSSION

The study aimed at demonstrating the relationship between the community performance-based financing and health service delivery by the community health workers

Results on age revealed that most CHWs working in Matana (53.2%) and Kayanza (65.6%) districts were between 36 and 50 years old. No CHW was above 65 years since, in MOH Burundi policy, CHWs retires at the age of 65 years. The study conducted by Denys in Rwanda on the assessment of CHWs incentives on maternal and newborn health services performance in Rwinkwavu health district explained that the majority of CHWs were aged between 36 and 50 years.7

The study revealed that most respondents were female (51.4% in Kayanza and 59.2% in Matana). The study of Denys in Rwanda found the same trend of gender dominance where 100% were female.6 However, it contrasts with a study in Kenya on the effects of selected socio-demographic characteristics of CHWs on the performance of home visits during pregnancy, it reported that 58% were male. The study showed that most of CHWs were married.8 While the majority of CHWs in our study had primary school level of education studies in Rwanda and Kenya showed that most CHWs had a secondary level of education and confirmed that a high level of education is associated with good record keeping.7,8

Farming as an occupation associated with community services. This means that CHWs live within the community and this in accordance with the Burundi policy. The policy in place emphasizes that CHWs should be chosen by community members and the fact that the majority of Burundian depend on agriculture and business.6

Results showed that community services delivery in Kayanza health district was significantly associated with community-based financing. The results were different from the study conducted in Rwanda where he found that there was no relationship between the community performance-based financing and the performance of CHWs services delivery.7 He explained that there was low provision of other incentives like training, income generated activities and materials in his study. The provision of equipment to CHWs was a significant prediction of performing the task of home deliveries and immediately referral at the health facility. The study in western Kenya identified that providing incentives like a bicycle for transportation, uniform, training materials, training skills on home base lifesaving and provision of first aids kits would increase the engagement in community maternal and new born services delivery where monetary allowances were not available.9 In Morogoro Region, Tanzania a qualitative study on the sources of community health worker motivation found that apart from the financial motivation, material supplies

support (for example, notebooks and pens) from community members motives them for working as CHWs.10

Drawing from qualitative findings, community health workers’ services delivery was dependent upon other incentives provided. In Rwanda and Kenya, results on qualitative studies revealed that provision of refresher training and in-kind incentives from community members contributed to better performance on community care of pneumonia.7,11 The CHWs might be motivated to deliver community health services if community programs were adapted to local settings, including the provision of training and materials, and supporting the system of governance in the cooperatives. The study recommends that further study could evaluate more factors that complement the community performance-based financing in community services delivery.

The training and supervision provided to CHWs were found as a factor that increases motivation. In India, they found that the CHWs considered motivation through getting regular supportive supervision and streamlining of the means of strengthening their responsibilities and they added that identifying and training more experienced volunteers for CHW’s supervision in resource-constraint settings was very challenging.12 Those items were complements of motivating our participants in terms of achieving community services. These finding were consistent within previous research where was found that CHWs considered understanding their voluntary role when supported by the facilities.13 According to this literature and results from this study, it is evident that training, supervision visits, and material provision to CHWs were important for the motivation of CHW’s works.

The focus groups and the KIIs were used as additional data collection tools apart the question and those are sometimes subjected to potential limitations. The main limitation was the unwillingness of the respondents to give true information of the questions of the questionnaires especially on the amount each CHW can get with PBF system.

CONCLUSION

The community performance-based financing had a significant association with the community health worker’s service delivery. In Kayanza, farming as an occupation was highly predicting this effect of CPBF on CHWs service delivery, and this was independent of the age, education, religious affiliation, and marital status of community health workers.

However, this effect was also predicted by other incentives provided to CHWs like training, supportive supervision, provision of materials that were important and helpful in improving the community services delivery.

International Journal of Community Medicine and Public Health | January 2023 | Vol 10 | Issue 1 Page 27

Gervais M et al. Int J Community Med Public Health. 2023 Jan;10(1):21-28

ACKNOWLEDGEMENTS

We are grateful to community health works of the Matana and Kayanza health, and to the staff of the two district who participated in the study and the collection of the data. Grateful are also directed to Dr Owino Alfred and Professor Maima for their support during study development.

Funding: No funding sources Conflict of interest: None declared Ethical approval: The study was approved by the Mount Kenya university ethics committee ref MKU/ERC/1834, Number 907

REFERENCES

1. Igna B, Robert S, Ellen VP, Olivier B, Longin G, Frank VL, et al. Introduction of performance-based financing in burundi was associated with improvements in care and quality. Health Affairs 2014;33(12):2179. 2. Elisabeth P. Performance-based financing to strengthen the health system in Benin: challenging the mainstream approach. Int J Health Policy Manag. 2018;7(1):35-47. 3. Godfrey BZ, Philip G, Robert ADO, John FM, Rogers A. The effects of performance-based financing on the health centres of Jinja Diocese, Uganda. Int J Public Health Res. 2015;3(4):162-7. 4. Eichler R, Levine R. Performance incentives for global health: potential and pitfalls. CGD Books; 2009. 5. Adrien R, Olivier B. Community PBF Project in Makamba Province- Appraisal Report, Bujumbura. March 2013. 6. MOH Burundi. Manual of procedures for the implementation of performance-based financing at the community level. Bujumbura. 2017. 7. Denys N. Assessment of community health workers’ incentives on maternal and newborn health services performance, in Rwinkwavu district hospital,