Supervision of community health nurses in Ghana: a mixed-methods study on experiences and mentorship needs

Abstract

Adequate supervision is critical to maintain the performance of health workers who provide essential maternal and child health services in low-resource areas. Supportive supervision emphasizing problem-solving, skill development and mentorship has been shown to improve the motivation and effectiveness of health workers, especially at the community level, but it is not always routinely provided. Previous studies have assessed the uptake of supportive supervision among volunteer health workers and paid health centre staff, but less is known about the supervision experiences of paid community-based staff, such as community health nurses (CHNs) in Ghana. This mixed-methods study explores the frequency and content of CHN supervision in five districts in the Greater Accra and Volta regions of Ghana. We analysed quantitative data from 197 satisfaction surveys and qualitative data from 29 in-depth interviews (IDIs) and four focus group discussions (FGDs) with CHNs. While the majority of CHNs received supervision at least monthly, they reported that supervision was primarily focused on meeting clinical targets (48%) rather than on handling specific cases or patients (23%). Over a third (34%) of CHNs did not agree that supervisors help them with job-related challenges and nearly half (43%) were unsatisfied with their jobs. When asked about their mentorship needs, CHNs reported wanting feedback on how to improve their job performance (40%) and encouragement (30%). There were only slight variations in the frequency and content of supervision based on type of supervisor. During IDIs and FGDs, CHNs offered ideas for how to improve supervision, including more frequent field visits so that supervisors could see the on-the-ground realities of their work, greater respect and positive reinforcement. Overall, CHN motivation and job satisfaction may be strengthened by aligning supervision more closely with the principles of supportive supervision.

Community health, human resources for health, health workforce, nurses, maternal and child health, supervision, Ghana

KEY MESSAGES

• Community health nurses (CHNs) in Ghana are an important part of the health system, providing maternal and child health services outside of health facilities.

• Most CHNs reported meeting with supervisors at least monthly; however, their visits primarily focused on meeting clinical targets.

• CHN motivation and job satisfaction may be improved by aligning supervision visits more closely to the principles of supportive supervision, including a focus on skill development and encouragement

Introduction

In sub-Saharan Africa, where 6–8% of children die before the age of five, and the maternal mortality ratio is 2.5 times the global ratio at 533 deaths per 100 000 live births (UNICEF, 2019). community-based health programmes have substantially improved child survival, disease prevention and reproductive health outcomes (Lehmann and Sanders, 2007; Perry et al., 2014; Perry, 2020); however, they can be difficult to implement at scale. Typically, these programmes rely on volunteer or salaried community health workers (CHWs), who receive formal training to deliver essential health services to members of the communities where they live or have shared life experiences. As programmes grow, a recurrent challenge is human resource management: ensuring that community-based staff remain motivated, effective and satisfied with their roles (Roberton et al., 2015).

Research has shown that supervision is among the most important factors affecting the motivation and effectiveness of community-based health staff. While strong supervision can improve satisfaction and productivity, poor quality or infrequent supervision can lead to discontent and high turnover rates (Uys et al., 2004; Kane et al., 2010; Jaskiewicz and Tulenko, 2012; Strachan et al., 2012; Ludwick et al., 2018). Many studies in sub-Saharan African contexts have found that CHWs and nurses do not feel adequately supported by their supervisors and perceive supervision to be an act of criticism and control (Stekelenburg et al., 2003; Uys et al., 2004; Mathauer and Imhoff, 2006). Specifically in Ghana, a study of CHWs and nurses found that only 15% of those surveyed felt supported by their supervisors, potentially negating the benefit of supervision on productivity (Frimpong et al., 2011).

The concept of supportive supervision is intended to mitigate these issues. The role of a supportive supervisor is to guide, mentor and coach workers, maintaining a focus on problem resolution, goal-setting and skill development, rather than fault-finding (Crigler et al., 2013). As outlined by the World Health Organization, the primary principles of supportive supervision are (1) frequent supervision visits, (2) a focus on problem-solving and skill development, rather than report checking or criticism and (3) ongoing support and encouragement (WHO, 2003). Strong two-way communication between supervisors and supervisees, including ‘real-time feedback’, forms the base of a supportive supervisory relationship (Bailey et al., 2016). In a systematic review of CHW supervision in four sub-Saharan African countries, Kok et al. (2018) found that supervision was perceived to be more supportive if it involved a problem-solving focus, joint responsibilities and skills-sharing, rather than being critical and hierarchical. Supportive supervision is hypothesized to foster a ‘virtuous cycle’, creating positive feedback that contributes to CHW recognition within communities, connection to the health system, confidence, effectiveness and satisfaction (Ludwick et al., 2018).

While previous studies have described the uptake and value of supportive supervision for volunteer health workers and salaried health centre nurses in sub-Saharan Africa, less is known about the experiences of formally employed, community-based workers, such as community health nurses (CHNs) in Ghana.

Country context

In Ghana, the first level of healthcare is delivered through the Community-based Health Planning and Services (CHPS) Program. CHPS was implemented as part of an experimental study in 1994, then rapidly expanded when initial results demonstrated that health services delivered by community-based nurses could produce dramatic improvements in maternal and child health (Nyonator et al., 2005). Within 3 years of the national rollout, Ghana saw a 50% reduction in the childhood mortality rate (Phillips et al., 2005). In addition, improvements were seen in breastfeeding and immunization rates, household involvement in diarrhoea treatment and prevention of waterborne diseases (Magawa, 2012).

Despite these triumphs, the maternal mortality ratio in Ghana remains high at 308 deaths per 100 000 live births and further progress is still needed on child and neonatal mortality reduction. Ghana also suffers from a shortage of healthcare workers, particularly in rural areas (WHO, 2019), and access to care remains an important equity issue. Less than 15% of physicians work in sub-district health centres, making community-based health staff critical for healthcare delivery at community levels (Alhassan et al., 2013).

CHNs in Ghana are compensated with salaries built into the national budget to provide maternal, newborn and child healthcare (Nyonator et al., 2005). Senior High School graduates may apply for a 2-year CHN training school certificate or a 3-year CHN training school diploma programme (Nyonator et al., 2005). After completing the certificate or diploma programmes, CHNs are posted to communities and after 3 years of service, may be promoted to senior CHN positions and/or apply for higher-degree programmes. Community health officers (CHOs) are a category of CHNs who, after completing CHN training, receive an additional 3–6 months of training focused on community engagement, outreach, community health planning and midwifery, and are posted to work at CHPS compounds or CHPS zones (clusters of communities being served by a CHO). For the purpose of this analysis, CHOs are referred to only as CHNs.

In 2007, the Ghana Ministry of Health developed a human resource development plan, which called for CHNs and CHOs to receive regular supportive supervision visits (Ghana Ministry of Health, 2007); however, the structure of supervision was found to be inconsistent across districts (Schwarz et al., 2019). CHNs may be supervised at various levels by those in charge at CHPS compounds (an ‘in-charge’ or i/c), sub-district public health nurses (PHNs), those in charge of Reproductive and Child Health Units (an ‘RCH in-charge’ or RCH i/c), sub-district heads or PHNs at the district level.

This study uses a mixed-methods approach to understand CHN supervision experiences and needs. The primary objectives of the study are to describe CHN supervision experiences, determine whether the frequency and content of supervision differ by type of supervisor and assess whether CHN supervision experiences align with the principles of supportive supervision.

Methods

The study was conducted as part of the Innovations for Maternal, Newborn, and Child Health Community Care Hub project, implemented by Concern Worldwide U.S. and JSI Research & Training Institute with ILC Africa as a local research partner. The conceptual framework for the project was based on the (Herzberg et al., 1959) theory of motivation, which posits that the presence or absence of ‘hygiene factors’ such as supervision, interpersonal relations, work conditions, salary and job security determine levels of worker dissatisfaction and motivation. The baseline study assessment involved two quantitative surveys administered to CHNs: a clinical knowledge assessment and a satisfaction survey, as well as in-depth interviews (IDIs) and focus group discussions (FGDs) with CHNs and supervisors. This manuscript includes results from the satisfaction survey, CHN IDIs and CHN FGDs. Results from other components are presented in (Sacks et al., 2015).

CHNs were recruited for the study during Care Community Hub mobile-application user training sessions, which took place over a period of 7 weeks between May and August 2014, and was mandated for all current CHNs in five districts in Ghana. These districts were predominantly rural: Ningo Prampram, Ada East and Ada West in the Greater Accra region and South Dayi and South Tongu in the Volta region. The implementation team worked with trained, local qualitative data collectors. A 2-day training session was held in May 2014 during which the tools were pilot-tested by data collectors and revised to better fit the local context. The final data collection tools, including consent forms, surveys and structured interview guides, are available in (Alva et al., 2015), along with baseline results from the original study.

At the Community Care Hub user trainings in each district, the quantitative surveys were administered to all CHNs in attendance and completed by over 95%. CHNs were informed about the purpose of the study and invited to voluntarily participate. Those who agreed provided written informed consent in their language of choice. The survey was paper-based and self-administered in private rooms within the health centres where the trainings were taking place. Data collection was monitored and quality checked by senior members of the research team. Unique identifiers were not collected, and paper surveys were destroyed after the data were transferred into Excel for analysis, in order to ensure confidentiality.

The CHN satisfaction survey included 39 questions on topics including job satisfaction, motivation, relationships and communication with peers and supervisors, career goals and job-related challenges. As a multiple-choice question format was used to allow for quick completion among a broad range of respondents, the surveys alone were not sufficient to gain a full understanding of CHN satisfaction. For instance, to assess the type of feedback CHNs received from their supervisors, CHNs were asked ‘Most of the time the feedback I receive from my supervisor is about…’ with five answer options and ‘none of the above’. In order to understand the contextual factors and elicit further elaboration on responses, a mixed-methods approach was used to strengthen the understanding of CHN supervision experiences.

IDIs and FGDs were used to gain a nuanced understanding of CHNs reasons for becoming a CHN, work experiences, job satisfaction, relationships with peers and supervisors, and professional aspirations. The study design included IDIs with eight CHNs, at least one of whom identified as male and half of whom were under age 30, and FGDs with five to eight CHNs, in each district. Participants were randomly selected from the list of CHNs registered for each district’s Care Community Hub training session based on the above criteria and asked to voluntarily participate. An effort was made to include CHNs with a range of years of experience and levels of seniority who were posted at both CHPS compounds and CHPS zones. Interviews were conducted in English by Ghanaian members of the research team. All research staff were trained in methods to encourage participation, avoid biased interpretation and manage group dynamics.

IDIs and FGDs were typically held in centrally located, rented office spaces in each district, providing privacy and independence from health centres or community demands. They ranged from 40 min to 2 h, and participants received compensation of 20 cedis (about $3.50 USD) to cover travel costs. All respondents were asked to sign a consent form before the interviews began agreeing to audio-recorded interviews and note taking. Respondents were assured that their supervisors and employers at Ghana Health Service would not know if they had chosen to participate and that they could discontinue at any point or refuse any question. Names and other identifying information were redacted from the written transcripts and the audio files were stored on a password-protected computer.

Data analysis

Quantitative data were entered into Excel then transferred to Stata for analysis by the research team. Descriptive statistics were generated on frequency of supervision, report of feeling supported by supervisors, type of feedback received and job satisfaction. As most CHNs had multiple supervisors at different levels (i/c, sub-district PHN/RCH i/c, sub-district head and PHN), the survey allowed CHNs to provide responses regarding their relationships with each supervisor type. Therefore, for each topic, results are presented by type of supervisor and as a single percentage across supervisor types. Results are not analysed by district, as initial analyses of these data indicated that district-level differences in CHN experiences were statistically insignificant.

Data from the qualitative portions of the study were compiled in the form of de-identified interview and focus group transcripts and interviewer notes. All qualitative data were coded in NVivo (Version 10) by the research team using thematic content analysis (Creswell, 1994). Pre-defined codes were developed based on previous formative research and new codes were added based on interviewer suggestions and emergent themes in the transcripts. The research staff met to check interrater reliability using randomly selected transcripts and agreement on appropriate codes applications and definitions was reached through consensus.

Results

Respondent characteristics

Data were collected from 197 satisfaction surveys, 29 IDIs, and 4 FDGs with four to seven participants each for a total of 23 individual FGD participants across the five districts. The CHNs included were similar across districts with respect to sex, age and marital status, with 91% (180/196) of the sample female, 79% (155/196) under age 30 and 41%, (80/197) married or living with a family member. Nearly all of the CHNs interviewed (93%) were posted in a different district than their home district. Approximately half of the CHNs (48%) were posted to health centres. The other half (51%) were posted to CHPS compounds or CHPS zones, giving them the title of CHO.

Supervision structure

Nearly all CHNs reported having multiple supervisors at different levels within the CHPS system; however, the structure of CHN supervision varied between and within districts. Across districts, 90% (178/197) of CHNs reported having an i/c supervisor and those who did not were typically i/c supervisors themselves. In some districts, CHNs served as sub-district heads and therefore did not have supervisors within the categories included in this study. The supervision structure in South Dayi at the time of the assessment is provided as one illustrative example for reference (Figure 1).

Figure 1

South Dayi district CHPS supervision structure. Blue indicates a sub-district head, red a health centre in-charge and green other clinical or centre staff. An asterisk indicates that the staff member serves as both a sub-district head and a centre in-charge.

Frequency of supervision

The majority of supervisors met with CHNs at least monthly [55.4% across supervisor types; range 48.7% (92/189) to 60.8% (115/189) by supervisor type] (Figure 2). However, nearly a quarter of supervisors (22.7%) met with CHNs less than four times per year. A number of CHNs (14.2%) also reported meeting with supervisors weekly. CHNs supervised by i/c supervisors reported the widest range of frequencies: while ∼30% (56/193) met with i/c supervisors weekly or more often, 25% (49/193) met fewer than four times per year.

Figure 2

Frequency of CHN supervision by supervisor type.

Type of feedback received from supervisors

CHNs were asked to select the primary form of feedback they receive from each their supervisors. Across the four supervisor types, CHNs reported receiving input on meeting clinical targets (47.5%) most often, followed by advice on handling specific cases or patients (23.4%). Time management, attitudes towards work, punctuality and other topics were also reported as the primary feedback received during supervision. Grouped together, these ‘other topics’ were reported by 30% of CHNs [range 22.6% (43/190) to 38.0% (73/192) by supervisor type]. CHNs with i/c supervisors reported other topics as the primary feedback they received most often [38.0% (73/192)] (Figure 3) (Supplementary Table S1).

Figure 3

Type of feedback received by CHNs during supervisory meetings.

Support for job-related challenges

Across supervisor types, the majority of CHNs agreed that their supervisors generally help them with challenges related to their work (66.0%) (Table 1). In particular, CHNs supervised by PHNs and RCH i/c supervisors were the most likely to agree that their supervisors help them with job-related challenges [70.1% (136/194)]. However, nearly a third (34.0%) of CHNs felt neutral or disagreed that supervisors help them with job-related challenges. The largest percentage of CHNs who disagreed were supervised by a sub-district head [19.2% (37/193)].

Table 1

Extent to which CHNs feel that their supervisor helps with job-related challenges (n = 197)

	In-charge		Sub-district PHN/RCH in-charge		Sub-district head		PHN		All supervisor types
	No.	%	No.	%	No.	%	No.	%	%
Agree	128	65.6	136	70.1	114	59.1	133	69.3	66.0
Neutral	28	14.4	30	15.5	39	20.2	35	18.2	17.1
Disagree	30	15.4	22	11.3	37	19.2	24	12.5	14.6
Not applicable	9	4.6	6	3.1	3	1.6	0	0	2.3
Total	195	100	194	100	193	100	192	100	100

Job satisfaction

Nearly half of the CHNs (43% across supervisor types) responded that they were either ‘unsatisfied’ or ‘very unsatisfied’ with their jobs. Neither supervisor type nor frequency of supervision was related to job satisfaction. The percentage of CHNs who reported being satisfied with their jobs was similar across supervisor types [range 56.7% (101/178) to 57.3% (102/178)], and a similar percentage of CHNs who were satisfied and unsatisfied with their jobs received supervision monthly or more frequently (56.1% and 53.8%, respectively).

A larger percentage of CHNs who were satisfied with their jobs received feedback on meeting clinical targets compared with CHNs who were unsatisfied (52.6% vs 40.6% across supervisor types) (Table 2). In addition, CHNs who were unsatisfied with their jobs reported other topics as the primary type of feedback they received more often than CHNs who were satisfied (28.9% vs 19.1%). Among the unsatisfied, 30% reported that the primary feedback given during supervision was on other topics. Details on job satisfaction by type of supervisor and type of feedback recieved are available in Supplementary Table S2.

Table 2

CHN job satisfaction by type of feedback received during supervision visits (n = 197)

	In-charge		Sub-district PHN/RCH in-charge		Sub-district head		PHN		All supervisor types
	No.	%	No.	%	No.	%	No.	%	%
Satisfied
Meeting clinical targets	46	46.0	58	56.9	51	51.0	57	56.4	52.6
How I handle specific cases/patients	19	19.0	54	52.9	25	25.0	27	26.7	22.8
Other	28	28.0	22	21.6	21	21.0	6	5.9	19.1
Not applicable	7	7.0	1	1.0	3	3.0	11	10.9	5.5
Total	100	100	102	100	100	100	101	100	100
Not satisfied
Meeting clinical targets	31	39.2	31	40.8	31	40.8	32	41.6	40.6
How I handle specific cases/patients	16	20.3	18	23.7	17	22.4	20	26.0	23.1
Other	27	34.2	23	30.3	27	35.5	12	15.6	28.9
Not applicable	5	6.3	4	5.3	1	1.3	13	16.9	7.5
Total	77	100	76	100	77	100	77	100	100

Supervision needs

Across supervisor types, the most important type of feedback that CHNs reported wanting from supervisors was advice on things to improve on (39.9%), followed by encouragement when they do things well (30.0%). The percentage of CHNs who reported these forms of feedback was consistent across all supervisor types, however, wanting greater respect from supervisors was more common among CHNs supervised by i/c supervisors than CHNs with higher-level supervisors [16.1% (31/192) for i/c vs 12.7% for others].

Realities in the field

CHNs who participated in IDIs and FGDs wanted more routine supervision visits within the communities where they worked.

At least once in a while [supervisors] should come to the community and find out what is going on and how the work is going. They do visit the CHPS, but they don’t come to the community itself to find out how we relate with them (Female, IDI, age 20–30, 1.5 years as a CHN).

My supervisors, they normally don’t come…our information arrives late sometimes, and at times they deliver a message to our sub-district that the reason [for the delay] is that my place is far. Meanwhile, you are the one who put that clinic there …. the supervision should come to the ground field (Female, IDI, age 20–30, 3 years as a CHN).

A few CHNs noted that supervisors did not offer enough training, so they had to rely on colleagues for instruction and learn on the job.

When you enter the field at first, is it your colleague nurse who will train you. Your supervisor won’t train you, because what you learn in school is different from what you do in the field, so when you come, they will have to teach you certain things…if they don’t teach you, you won’t know (Female, IDI, age 20–30, 1 year as a CHN).

Communication and respect

CHNs noted that their supervisors were generally easy to talk to and that they were not apprehensive to go to their supervisors with certain types of problems, although they sometimes felt criticized, and recognized that support could be limited.

The relationship is very good. She doesn’t make decisions on her own; we all come to the table and decide on one (Female, IDI, age 30–40, 5 years as a CHN).

Always when [our supervisors] come around, they tell us that they are not here to criticize. They are here to see what we are doing and what we are not doing right and make corrections. So then even if we do something that is not right, they don’t insult us. They don’t abuse us much. They only tell us the right thing to do…so it’s cordial (Female, IDI, age 40–50, 14 years as a CHN).

Some CHNs said that they did not feel respected by their supervisors or by Ghana’s Health Services system because of their lower-status role. They felt that their work in the community was important but was not always recognized.

The aspect I do not like is with… Ghana Health Service, because community health nurses are not given the needed recognition we deserve (Female, IDI, age 20–30, 1 year as a CHN).

If [supervisors] see the kind of work we are doing and then they call you an auxiliary [assistant] nurse, your spirit will go down. It is like what you are doing, they are not appreciating it (Female, FGD, age 20–30, 2 years as a CHN).

You are trying to kill yourself for the job. You try everything. You are working hard. There is no way the top people will recognize your work. There is no single day they will praise you, always fault finding… When it is raining, I will be in the rain. In the village, I know nobody…nobody cares whether you are sick or not…[supervisors] don’t regard us, but they need our figures for the ongoing office work (Female, IDI, 30–40, 2 years as a CHN).

Advice and encouragement

In IDIs and FGDs, many CHNs reported wanting positive reinforcement and respect from their supervisors and commented on the tone used by supervisors.

Sometimes when they come, they only come to point at mistakes…and then bombard you with insults…and don’t correct you (Female, IDI, age 20–30, 6 years as a CHN).

They can support us by using appreciative words. It’s would motivate us to work by giving us a chance to express ourselves without them becoming defensive…They should show us respect (Female, IDI, age 20–30, 6 years as a CHN).

Discussion

Overall, our results indicate that while most CHNs maintain good relationships with their supervisors and receive frequent supervision, the consistency and content of supervision visits could be strengthened to improve CHN job satisfaction. While a small number of CHNs reported experiences in line with the principles of supportive supervision (including frequent visits, a focus on problem-solving and encouragement), many did not.

Despite national guidelines recommending monthly supervision for CHPS staff (Ghana Health Service, 2016), CHNs reported a range of supervision frequencies. These results align with a national survey of CHWs in Ghana conducted in 2017, which found that although the majority received supervision monthly or more often, there was substantial variation in supervision frequency (Schwarz et al., 2019). It would be logical for supervision frequency to differ according to supervisor level, with more routine supervision occurring among lower-level i/c supervisors, as they might have more time to spend on CHN oversight and duties that more closely resemble CHN’s day-to-day activities. However, this was not the case, perhaps due to individual variation in how much time supervisors and CHNs spent at the same posting sites. During interviews, the location of supervision visits, rather than their frequency, emerged as a key theme, with CHNs wishing that their supervisors would come to the communities where they work more often to observe and provide feedback on patient care and better understand the realities ‘in the field’. In future studies, it would be valuable to collect quantitative data on the frequency of observation-supervision, as research suggests that this form of supervision has a large impact on clinical skill and healthcare quality (Bailey et al., 2016). It would also be helpful to further understand which aspects of CHN performance might be most important to community members in order to tailor supervision most relevant to their field-level work.

Our results suggest that while some supervisors emphasize problem-solving and skill development, many primarily focus on report-checking and criticism. Across types of supervisors, roughly a quarter of CHNs reported that the primary focus of supervision was on handling specific cases or patients, which may include feedback in line with the principles of supportive supervision; however, half of the CHNs reported that the main type of feedback they received pertained to meeting clinical targets and a quarter mentioned other topics. Likewise, in interviews and FGDs, some CHNs noted that they work with supervisors to make decisions and receive advice on how to improve their job performance, while others reported that they lacked sufficient training and that supervisors routinely criticized their work without providing mechanisms for improvement.

The pattern of limited uptake of supportive supervision has been identified within numerous country contexts. In a mixed-methods assessment of CHW supervision in Tanzania, 90% of CHWs reported that supervisors primarily checked registers during meetings, while only 40% reported receiving feedback on work and only 13% reported receiving sufficient training (Roberton et al., 2015). Likewise, in the MEASURE Evaluation study in Haiti, data validation was the most common supervision activity mentioned, noted by all community-level participants, while only two participants mentioned training or mentoring (Marshall and Fehringer, 2013).

CHNs in our study also lacked consistent support for job-related challenges and encouragement and reported wanting more positive reinforcement and respect from supervisors. In a previous qualitative study of nurses in Ghana, Kwansah et al. (2012) found that nurses posted to rural locations described a sense of being lost or forgotten in the system and felt that there were strong differences in mentorship across regions, with those in rural postings receiving the poorest quality mentorship. CHNs in our study, who were recruited from predominantly rural districts, similarly expressed lacking respect and recognition within Ghana’s health system, and a lack of understanding by supervisors of the unique challenges of working in rural communities.

Finally, our results indicate that aligning CHN supervision content more closely to the principles of supportive supervision could result in greater CHN job satisfaction. Across supervisor types, nearly half of the CHNs were unsatisfied with their jobs, but job satisfaction was higher among those who received more feedback on how to improve their work, encouragement and respect from supervisors.

This study has some limitations. The CHNs were recruited from those attending Community Care Hub training sessions in five pre-selected districts in Ghana rather than a full roster of employees; however, as this was a mandated training by the Ministry of Health, this should be a reasonable proxy for CHNs in each district. We did not assess the supervision experiences of CHNs in other regions of Ghana, which may have yielded different results, particularly if districts with a larger percentage of the population residing in urban areas had been included. Job satisfaction was self-reported, so desirability bias may have been present. Furthermore, where survey options were presented as multiple choice, respondents were asked for the best choice; analysis of the most frequent quantitative survey responses did not preclude other responses from being important, hence our use of qualitative data for further exploration. We also recognize that the definition of a ‘good’ relationship with a professional supervisor may vary by respondent and, given the dynamic nature, may change depending on when asked. We are limited by the language in the surveys, but have relied on triangulation and findings from IDIs to provide further nuance about the important aspects of this interpersonal relationship. Since we did not assess job performance in this study, we cannot know if supervision focused on constructive criticism was warranted due to poor performance, or was unnecessary as reported. In addition, due to transportation and communication constraints in rural districts, we only conducted four FGDs, yet similar themes emerged during IDIs, contributing to trustworthiness and the reliability of the results. Finally, although it was beyond the scope of this study, it would be important in the future to understand the drivers of supervisor behaviour and the facilitators and barriers faced by supervisors in providing effective mentorship.

This study also provides a number of valuable contributions. Firstly, data were drawn from a fairly large sample of CHNs across more than one region of Ghana. Secondly, the use of mixed methods allowed for a nuanced understanding of CHNs supervision experiences and motivators for stronger job performance. Finally, understanding the supervision experiences of CHNs is critical, as they are part of a growing cadre of health workers at the intersection of the formal health system and the community. Many CHW programmes in sub-Saharan Africa emerged in an ad hoc fashion to address pressing health crises, such as HIV and tuberculosis, and have relied heavily on CHW’s volunteer labour (Swartz and Colvin, 2015). As these programmes expand, many governments are working to formalize CHWs’ roles within the health system by creating salaried community-based positions and rigorous training programmes (Zachariah, 2009). Ghana’s CHPS system has proven to be a strong, sustainable strategy for scaling up community-based health programmes, and other countries are likely to look to the CHN model for inspiration (Kweku et al., 2020). Therefore, understanding CHNs’ supervision experiences and needs can offer important lessons for new and expanding national CHW programmes.

Conclusion

CHNs play a critical role in extending primary healthcare services to communities throughout Ghana. It is therefore essential that they are adequately supervised and feel supported to withstand and resolve job-related challenges, and advance in their careers. Better aligning supervision activities with the principles of supportive supervision may improve the motivation and effectiveness of health workers, especially at the community level.

Supplementary data

Supplementary data are available at Health Policy and Planning online.

Conflict of interest statement. None declared.

Funding

This work was supported by the Gates Foundation: Innovations for Maternal, Newborn, and Child Health.

Acknowledgements

Thanks to Anne LaFond, Leanne Dougherty, Amanda Makulec, Nicole Davis, Savitha Subramanian, Jennifer Pierre, Anitha Moorthy, Akuba Dolphyne, Jean-Christophe Fotso, Patricia Porekuu and Jahera Otieno for assistance during project implementation. We also thank Linda Vesel, Frédérique Vallières and Peter Rockers for input on the research instruments.

Ethical approval. This study was approved by the Institutional Review Board at John Snow Inc. and the Ethical Review Committee at Ghana Health Service (GHS-ERC: 07/09/13).