The review’s eligibility criteria were developed using the Population, Concept and Context (PCC) framework.

Inclusion criteria:

Types of articles included:

Types of articles excluded:

A three-step process was used to select studies.^27,28 Three databases, such as PubMed, ScienceDirect and Google Scholar (first 10 pages), were searched. These databases were selected as they have been previously used in other African systematic reviews.^30,31 The search was conducted using the key terms ‘task-sharing’, ‘community health workers’, ‘challenges’, ‘non-communicable diseases’ and ‘Africa’. After the relevant articles were identified, the index terms were used to develop a search strategy for the databases. The database search was conducted on 26 March 2025. The Boolean operators used were ‘OR’ and ‘AND’. The full search strategies for all three databases are included in Online Appendix 1. The reference lists of the selected articles were also searched for relevant studies.

The references of the identified studies were uploaded to the Mendeley reference manager, and duplicates were removed. Titles and abstracts were then screened by two reviewers, N.L.M. and N.F. Screening was performed using Covidence, whereby potential studies and their full texts were screened against the eligibility criteria. During screening, reasons for exclusion were noted. Disagreements between the reviewers were resolved through discussion. The PRISMA chart was used to show the decision process.³²

Data were extracted from selected articles by two reviewers using a data extraction tool developed by the reviewers. The data extracted included the title of the study, year of publication, country where the study was conducted, objectives of the study, study design, sample size and outcomes in relation to the review. The data extraction tool is included as Online Appendix 2.

The selected studies were appraised using the Critical Appraisal Skills Programme (CASP) checklists. Although the Arksey and O’Malley Framework does not require the appraisal of selected studies for inclusion, the researchers deemed it necessary to appraise the studies. Two checklists were used, and these were for qualitative studies³³ and cross-sectional studies.³⁴ The reviewers, N.L.M. and N.F., appraised the studies independently, and any disagreements were discussed. The CASP checklists broadly assessed the validity of the results, results and the usefulness of the results. The overall decision would be negative (poor methodology) or positive (sound methodology).

A numeric analysis was first conducted of the selected studies. To answer the research questions, qualitative content analysis was used to analyse data independently by two researchers. Four steps outlined by Kleinheksel et al.³⁵ were used. The first step was identifying the unit of meaning aligned with the main study objective. The second step was labelling similar units with a code. The third step was grouping the codes into a category, and lastly (the fourth step), the categories were organised as themes.

A total of 189 articles were screened by title only, of which 137 did not meet the inclusion criteria, and 7 could not be retrieved. As such, 45 articles were screened by title and abstract. After screening, 14 articles were selected. Figure 1 shows the PRISMA flow chart of the decision process.

All the 11 studies assessed using the CASP qualitative checklist had an overall appraisal summary of positive methodological quality. The most common unreported item in the qualitative studies was the relationship between participants and researchers. The two studies that used a mixed methods design were assessed using the CASP cross-sectional and qualitative checklists. The overall appraisal decision for both studies was ‘positive’. One study was assessed using the CASP checklist for cross-sectional studies, and the overall decision was also positive. The outcomes of the quality appraisal are included as Online Appendix 3.

Fourteen (n = 14) studies were included in the review. Most (28.6%, n = 4) studies were from South Africa, three (21.4%) studies from Uganda, two (14.3%) studies from Nigeria and one (7.1%) study each from Kenya, Zimbabwe, Mozambique and Ethiopia. One (7.1%) study reported findings from South Africa, Mexico, Bangladesh and Guatemala. Selected studies were published between 2015 and 2024, with most (four each) being published in 2023 and 2022. Eleven (78.6%) studies used a qualitative approach, while two studies used mixed methods and one used a quantitative approach. Nine (64.3%) studies focused on mental health, while other studies reported on hypertension (7.1%), diabetes and hypertension (7.1%), cardiovascular diseases (7.1%), all NCDs (7.1%) and chronic obstructive pulmonary disease (COPD) (7.1%). Table 1 summarises the characteristics of the selected studies.

The studies reviewed described the contexts in which CHWs experienced challenges. In most studies, CHWs experienced challenges while working in rural areas.^39,42,44,48 In some studies, the challenges were experienced in peri-urban communities.^36,40 One study was conducted in both rural and urban settings.⁴⁶ Additionally, some studies named the study sites: one study³⁷ was conducted in four states in Northern and Southern Nigeria, and another study³⁸ was conducted in South-West Ugandan schools, while another study was conducted in Mbarara District, Uganda.⁴³ Munodawafa et al.⁴⁵ and Selohilwe et al.⁴⁷ conducted their studies in South Africa; Khayelitsha, Cape Town and Kenneth Kaunda District, respectively. One study described the study setting as a low socio-economic area.⁴¹

The studies reviewed described the socio-demographic characteristics of CHWs who experienced challenges in task-sharing for NCD mitigation. In most studies,^{36,37,38,41,42,43,46,48} the CHWs who participated were female. In one Mozambican study,³⁹ more male CHWs participated, while in another study,⁴⁹ there was an equal proportion of male and female participants.

Concerning the level of education of the CHWs, four studies^39,40,43,45 reported that CHWs had completed high school education. In some studies, CHWs had some high school education, such as Abrahams-Gessel et al.³⁶ who noted that the minimum level of education was Grade 8, while Wall et al.⁴⁹ note that 45.0% of the CHWs had some high school education. In one rural Ugandan study,⁴² most (70.0%) CHWs had primary school education, while on the other hand, most CHWs in a Nigerian study³⁷ had completed a basic national diploma.

The review found three themes of challenges experienced by CHWs. These were individual-level, organisation-related and community-related challenges.

At an individual level, the CHWs’ lack of knowledge and skills affected their ability to mitigate NCDs.

Studies^{38,39,40,41,43,45,47,48} included described inadequate knowledge, which made it difficult for CHWs to provide health education, and this lack of knowledge would culminate in poor skills in providing NCD care. In South Africa, Doresha and Mash broadly described the issue of a lack of skills in supporting the management of all NCDs.⁴¹ For example, the CHWs explained their inability to educate community members on insulin self-administration for diabetes care as they had not been trained.⁴¹ Other South African studies noted that despite training and having a manual, the CHWs often misinterpreted the therapeutic terms or modalities,⁴⁰ and it was challenging to provide psychosocial care, as mental health was associated with multimorbidity.⁴⁷ Likewise, Ethiopian participants explained that they had inadequate knowledge to answer hypertension-related questions from the community and lacked the skills to measure blood pressure.⁴⁸ Mozambican CHWs also had inadequate knowledge in three areas: (1) believed in false treatments of epilepsy, (2) had inadequate knowledge of its causes, and (3) believed in some cultural treatments.³⁹ Although CHWs did not have adequate knowledge on the three issues, Cumbe et al. note that, generally, CHWs had sufficient knowledge of epilepsy.³⁹ In Uganda, even after dementia training, CHWs shared how they were still not proficient.⁴³ Sometimes, the issue of poor skills was attributed to anxiety about performing a new task and would eventually resolve, with performance improving with continued service provision; however, in a few cases, CHWs’ performance may remain poor.⁴⁵ Poor skill levels at the onset of providing a counselling intervention would only be resolved in the third or fifth session.³⁸

Organisation-level challenges described in the reviewed articles emanated from the health facilities that supported the CHWs. These organisational challenges included referral problems, inadequate infrastructure and a lack of supervision.

Two Ugandan studies,^42,43 a South African study⁴¹ and a Nigerian study,³⁷ described problems with the referral process. In South Africa, some primary healthcare facilities disregarded referrals made by CHWs who had screened patients for hypertension and diabetes.⁴¹ Sometimes, community members in South Africa would be referred to health centres, and they would not afford the transport costs.⁴¹ Similarly, in rural Uganda, CHWs noted how community members screened for COPD would not reach the health centres.⁴² One Ugandan CHW noted, ‘We expect the patients to reach the … clinic and they don’t’.⁴² Moreover, in Uganda and South Africa, referrals were problematic as the service required was unavailable at the clinics.^41,43 In Nigeria, Ajisegri et al. also noted how CHWs would refer community members to district hospitals; however, there was no feedback about the referred patients, which compelled them to call community members for an update.³⁷

Another organisational challenge experienced by CHWs was the lack of working space and inadequate equipment.^37,42,47,48 In a South African study, it was revealed that CHWs would improvise for working space, and consequently, patients would not know where to find them for follow-up mental health services.⁴⁷ The lack of space was underscored when CHWs screened for COPD in the homes of community members in rural Uganda and found it difficult to maintain privacy with neighbours who were curious about the spirometry machine.⁴²

Regarding the lack of equipment, a community-based hypertension screening intervention in Ethiopia lacked sphygmomanometers and stethoscopes from the health posts, which affected service delivery.⁴⁸ Additionally, CHWs had to ‘improvise’ to measure blood pressure or blood glucose by sharing the sphygmomanometers between departments and asking patients to pay a fee for blood glucose testing.³⁷

Three studies^37,47,48 described the lack of supervision of CHWs. Teshome et al.⁴⁸ found that five participants explained that supervision is essential for programme monitoring. Similarly, in South Africa, irregular and inadequate supervision were identified as a barrier to scaling up mental health service delivery.⁴⁷ One South African CHW explained the issue of irregular supervision: ‘… they take a while to come back again’.⁴⁷ These findings are also quantified in the Nigerian study, which found that 52.2% of CHWs reported that inadequate supervision was a challenge.³⁷

Four community-level challenges were identified, namely a lack of trust in CHWs, concerns of neighbourhood crime, transport challenges and poverty among community members.

Although CHWs broadly operate from the communities they come from, in some studies,^42,49 there were challenges with community members trusting them. In rural Uganda, the CHWs reported that they were not trusted by the community, as they thought they were land grabbers.⁴² The Kenyan study reiterates this issue, highlighting how the community may perceive that their problems would not be kept in confidence when counselling is provided by CHWs.⁴⁹ This lack of trust in CHWs also made it difficult to include family members who were deemed essential for the counselling intervention.⁴⁹

In South Africa, two studies by Munodawafa et al.⁴⁵ and Abrahams-Gessel³⁶ described the issue of crime in the neighbourhoods from which CHWs operated. One South African CHW highlighted, ‘We saw skollies [thugs] when we were at the station and had to turn back’.⁴⁵ Similarly, in Cape Town, CHWs who often wear uniforms stood out and were often targets of theft.³⁶

In a rural Zimbabwean study, one of the barriers to ensuring community access to mental health services through the Friendship Bench initiative was the issue of transport.⁴⁴ Kidia et al. explain that CHWs would often travel long distances in extreme weather and at times on foot.⁴⁴ Similarly, in Kenya, CHWs also provided mental health services to elderly people and shared how they were unable to conduct home visits because of transport challenges.⁴⁶

Community health workers who provided mental health interventions were affected by poverty described by community members.^40,44,45,49 In the rural Zimbabwean study, some patients had to borrow money to access CHWs for mental health support.⁴⁴ Similarly, South African CHWs felt helpless when faced with patients with material needs, and they were unable to offer such material assistance.^40,45 Sometimes, there would be progress in mood improvement; however, this would be disrupted by an urgent financial challenge.⁴⁰ The Kenyan study further described how it was difficult to provide a mental health intervention when community members experience financial challenges and cannot afford food.⁴⁹