Rwanda’s community health workers at the front line: a mixed-method study on perceived needs and challenges for community-based healthcare delivery during COVID-19 pandemic

Objective During the COVID-19 pandemic, community health workers (CHWs) served as front-line workers in the COVID-19 response while maintaining community health services. We aimed to understand challenges faced by Rwanda’s CHWs during a nationwide COVID-19 lockdown that occurred between March and May 2020 by assessing the availability of trainings, supplies and supervision while exploring perceived needs and challenges. Design and setting This study was a mixed-method study conducted in three Rwandan districts: Burera, Kirehe and Kayonza. Main outcome and measure Using data collected via telephone, we assessed the availability of trainings, supplies and supervision during the first national lockdown, while exploring perceived needs and challenges of CHWs who were engaged in COVID-19 response, in addition to their existing duties of delivering health services in the community. Results Among the 292 quantitative survey participants, CHWs were responsible for a median of 55 households (IQR: 42-79) and visited a median of 30 households (IQR: 11-52) in the month prior to the survey (July 2020). In the previous 12 months, only 164 (56.2%) CHWs reported being trained on any health topic. Gaps in supply availability, particularly for commodities, existed at the start of the lockdown and worsened over the course of the lockdown. Supervision during the lockdown was low, with nearly 10% of CHWs never receiving supervision and only 24% receiving at least three supervision visits during the 3-month lockdown. In qualitative interviews, CHWs additionally described increases in workload, lack of personal protective equipment and COVID-specific training, fear of COVID-19, and difficult working conditions. Conclusion Many challenges faced by CHWs during the lockdown predated COVID-19 and persisted or were exacerbated during the pandemic. To promote the resilience of Rwanda’s CHW system, we recommend increased access to PPE; investment in training, supervision and supply chain management; and financial compensation for CHWs. We applied a mixed-method cross-sectional design using a convergent parallel approach.19 It was nested in a larger study conducted in three rural Rwandan districts of Burera, Kirehe and Southern Kayonza, to understand the effect of COVID-19 on community health indicators and CHWs’ work. These three districts are supported by Partners In Health/Inshuti Mu Buzima (PIH/IMB), a Boston-based international non-governmental organisation that supports health system strengthening. Of these three districts, Kirehe is on the border of Rwanda and Tanzania and experienced substantial truck traffic from Tanzania, even during the national lockdown. This transit route resulted in larger number of imported COVID-19 cases in Kirehe district, relative to the other two districts. Each Rwandan village, which is composed of 100–250 households, has three CHWs including one female Agent de Sante Maternelle (ASM) and a male–female pair known as Binômes. ASMs provide maternal and newborn healthcare through home visits and follow-up while Binômes provide a wider range of services, including diagnosis and treatment of childhood illness and antimalaria services for people of all ages, malnutrition screening and provision of contraceptives. Nonetheless, the roles of CHWs are not mutually exclusive based on CHW cadre. Except Binomes who are responsible for treatment and provision of medications, any CHW can perform health promotion activities, referrals to health facility, home visits and check-ins. In PIH/IMB supported regions, a fourth CHW is being added in most villages to support health promotion. Although performance-based financing is provided in some instances, CHWs currently work as volunteers and do not receive regular financial compensation.20 Consequently, CHWs engage in other activities to support their family’s livelihood. CHWs are mainly supported by the Rwanda Ministry of Health through their local health centres. CHWs receive a comprehensive kit of supplies, including a prepaid cell phone which can be used to contact collaborators in local government and health sectors. CHWs are directly supervised at a cell level, which is an aggregate of approximately seven villages. A CHW cell coordinator keeps direct contact with the nearest health centre and supervises all CHWs from villages in his/her administrative cell through monthly meetings or one-on-one visits in the community. PIH/IMB provides additional support to the CHW in its catchment districts by providing them with trainings to improve their service delivery and reporting as well as financial support by funding income-generating cooperatives for CHWs. We used a stratified simple random sampling methodology to select participants out of 5767 CHWs from the three PIH/IMB-supported districts. We stratified the populations of CHWs into nine groups cross-classified by district and cadres (ASM, female Binome, male Binome) and sampled 5% of individuals from each of the nine groups at random. We excluded CHWs who had served their current village for less than 1 year at the start of the COVID-19 pandemic. Our target sample size for the quantitative survey was 292, which would allow us to report 95% confidence intervals with a precision of at least±6% for the overall CHW population, district-specific estimates with a precision of at least±14%, and cadre-specific estimates with a precision of at least±12%. To achieve the minimum sample size of 292, selected CHWs who were unreachable on the phone were replaced by the next CHWs on the sampling frame, which expanded the overall sample size to 349. Informants of the qualitative survey were purposely selected among CHWs who participated in the quantitative survey and the sample size for the qualitative portion was determined by thematic saturation. Participants had no particular involvement in the implementation of the study, other than answering to the surveys. Our quantitative data collection tool was developed by adapting questions from validated questionnaires previously used among healthcare workers in similar settings.21 22 We conducted a phone-based data collection between 30 August 2020 and 3 October 2020. CHWs were recruited via phone calls to set up an appointment for data collection, received a text message containing a shorten generic consent information and asked to consent to the study by replying yes or No to voluntary participation to the text message. Data were collected by short-term research staff who do not typically have interactions with the CHWs and who do not play a supervisory role to the CHWs. In effort to reduce non-response bias, we made three phone-call attempts on three different days; and CHWs who were not reachable directly, we contacted a peer CHW serving in that village to help us reach the CHW sampled for the study. CHWs who were not reachable at three different attempts on three different days were considered non-response. Quantitative data were collected via a 40 minutes-long phone interview and stored into a password-protected REDCap database, a data management application first developed by Vanderbilt University and hosted on a Rwanda-based server.23 For qualitative data, we developed an interview guide comprising nondirective open-ended questions framed around challenges and needs of CHWs during COVID-19 pandemic. Probes to elicit deep responses were predetermined and included in the interview guide.24 Interviews were conducted in Kinyarwanda, audiorecorded, transcribed verbatim and translated to English. Quantitative and qualitative data were collected concurrently. All variables were based on participants’ self-reported information. We reported demographic characteristics, including age, level of education, marital status and main occupation, for quantitative and qualitative study participants. Among demographic variables, we also collected Ubudehe categories for the participants. Ubudehe is a four-rank Rwandan home-grown categorisation of socioeconomic status, where Ubudehe 1 represents the poorest and Ubudehe 4 represents the wealthy.25 We assessed whether CHWs had received any training in the past 12 months prior to COVID-19 (March 2019–March 2020) as well as whether they had received training on specific health-related topics and presented responses by CHW cadre. We assessed whether CHWs had a full kit of supplies normally provided to CHWs. Among Binomes, we assessed supply availability both at the beginning and throughout the lockdown for the following items: injectable contraceptives, oral contraceptives, condoms, reproductive cycle beads, malaria diagnostic test, malaria drugs, mosquito nets, antibiotics, deworming tablets, vitamins, registries, referral forms, timer, measuring tapes, weight-scale and other supplies. Among ASMs, who have a narrower scope of work and narrower range of required supplies, we assessed supply availability of mosquito nets, measuring tapes, weight scale, timer, registries referral forms and other supplies. We also assessed the frequency of supervision visits during a 70-day lockdown and means of supervision. Based on ideal practice of supervising CHWs at least one time per month, we categorised supervision into: no supervision, once, two, three times and more than three times.