Nature of review

Our scoping review forms part of an initial wider scoping review, conducted to identify and map the available evidence on economic evaluations of both vertical and integrated horizontal CHW programmes in LMICs published between 2015 and 2024. The protocol was uploaded to the Open Science Framework (OSF) on July 27, 2023 [13]. Due to the large number of studies identified and the heterogeneity between studies, the reporting of results has been divided into several publications, by health/disease area or type of CHW, for clarity and to facilitate comparisons between similar studies. This paper focuses exclusively on RMNCH.

A scoping review was chosen given the broad and varied nature of the field, with the goal of identifying updated evidence, mapping research methodologies, and highlighting knowledge gaps. This study was conducted according to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses Extension for Scoping Reviews (PRISMA-ScR) guidelines [14]. The PRISMA-ScR Checklist is available in the Appendix (S1 File).

Search strategy and study selection criteria

An initial search covering January 1, 2015, to July 11, 2023, was conducted in the following databases: Ovid MEDLINE(R) and Epub Ahead of Print, In-Process, In-Data-Review & Other Non-Indexed Citations and Daily (1946 to July 06, 2023); Ovid Embase Classic+Embase (1947–2023 July 07); Ovid APA PsycInfo (1806 to July Week 1 2023); Ovid Global Health (1910–2023 Week 26); Ovid AMED (Allied and Complementary Medicine) (1985 to June 2023); Cochrane Central Register of Controlled Trials (CENTRAL); Cumulative Index to Nursing and Allied Health Literature (CINAHL); Web of Science Core Collection; Scopus; and Latin American and Caribbean Health Sciences Literature (LILACS). To ensure this review was up-to-date, the search was re-run to capture relevant literature up to and including July 16, 2024.

Additionally, we searched the following sources to identify any relevant grey literature, using the search strategy outlined below and in supplementary files: Google Scholar, Bielefeld Academic Search Engine (BASE), DART-Europe E-theses Portal; e-theses online service (EThOS), Open Access Theses and Dissertations, and The OAIster database, plus websites of key organisations involved with CHWs (e.g., CHW Central, Community Health Impact Coalition, and Healthcare Information for All (HIFA.org)). Grey literature included (but was not necessarily limited to) theses or dissertations, preprints or unpublished research, and internal reports and the date span was January 1, 2015, to July 16, 2024.

The search strategy included all appropriate controlled vocabulary and keywords for ‘Community Health Workers’, ‘Economic Evaluations’ and ‘LMICs’, which are defined below. Reference lists of included studies were reviewed to identify any additional studies missed by database searches. Full database search strategies are available in the Supplementary Material (S1 Text).

Community Health Workers

For this review, we drew upon previous literature [5,15,16] to define CHWs as healthcare workers who:

1. (a). are primarily based in the community providing primary healthcare services;
2. (b). are part of the health system (i.e., government or non-governmental organization supported CHWs), performing tasks related to health-care delivery, and/or health education, promotion, or care coordination; and
3. (c). have received organised training and/or certification, but do not have a tertiary-level degree such as a nursing or midwifery degree.

Economic evaluations

Both full and partial economic evaluations were included. Full economic evaluations, as defined by Drummond et al., (2015) [17], compare the costs and outcomes of health interventions against alternatives such as the current standard of care or a no-intervention scenario. These may include Cost-Effectiveness Analysis (CEA), Cost-Utility Analysis (CUA), Cost-Benefit Analysis (CBA), Cost-Minimization Analysis (CMA), Cost-Consequence Analysis (CCA), Social Return on Investment (SROI), Multi-Criteria Decision Analysis (MCDA), Budget Impact Analysis (BIA), and Programme Budgeting and Marginal Analysis (PBMA).

Partial economic evaluations, on the other hand, consider costs and/or consequences without necessarily comparing alternatives or linking costs to benefits. They can include outcome descriptions, cost descriptions, cost-outcome descriptions, effectiveness evaluations, or cost analyses.

While full economic evaluations are preferred for guideline and policy development due to their comprehensive nature [11], partial economic evaluations are valuable for initial program development and in contexts where full evaluations are too costly, particularly in LMICs [18].

LMICs

The World Bank classification of economies was used to categorise LMIC countries as either ‘low’, ‘lower-middle’ or ‘upper-middle’ income based on the costing date for each respective study [19].

Reproductive, maternal and newborn health

In this review, we define reproductive health, maternal and newborn care as follows. Services focused on the health of women before and during pregnancy, childbirth, and the postpartum period (up to six weeks after delivery) could include, but are not limited to: Antenatal care (e.g., routine check-ups, nutrition counseling, screening for complications); Intrapartum care (e.g., skilled birth attendance, emergency obstetric care); Postnatal care (e.g., postpartum check-ups, family planning counseling, management of postpartum complications); Preconception care (e.g., nutritional support, folic acid supplementation, screening for infectious diseases) and Health education and counseling (e.g., birth preparedness, breastfeeding support, mental health support). Services focused on newborns (from birth through the first 28 days of life) could include but are not limited to: Essential newborn care (e.g., neonatal resuscitation, thermal care, early initiation of breastfeeding); Screening and management of neonatal conditions (e.g., neonatal sepsis, low birth weight); Immunization (e.g., polio, hepatitis B); Postnatal follow-up (e.g., growth monitoring, screening for developmental anomalies). Within this broad area of Reproductive health, maternal and newborn care, we distinguish two sub-areas: 1) Reproductive health, and 2) Maternal and newborn care.

Child health

Child health extends beyond the neonatal period up to childhood - commonly up to five years of age. Child health services could include but are not limited to: Integrated Community Case Management (iCCM) (e.g., diagnosis and treatment of common childhood diseases such as pneumonia, diarrhea, and malaria); Nutritional support and supplementation (e.g., vitamin A, iron supplementation, therapeutic feeding for malnutrition); Routine immunizations (e.g., measles, diphtheria-tetanus-pertussis [DTP], Haemophilus influenzae type b [Hib]); Growth monitoring and health promotion (e.g., regular weight and height measurements, health and hygiene education); and Early childhood development interventions (e.g., psychosocial stimulation, early learning support). In this article, we focus on four sub-areas of Child health: 1) preventing child mortality and promoting survival, 2) child development, 3) nutrition, and 4) infectious diseases.

Inclusion and exclusion criteria

Studies were included if they:

• Primarily evaluated CHW programs, excluding those focused exclusively on other healthcare professionals such as doctors, nurses, or midwives.
• Evaluated CHW programs focused on RMNCH.
• Provided details of an economic evaluation, including either full or partial evaluations.
• Were published between August 2015 and July 2024, as the previous review on this topic covered literature up to July 2015.
• Evaluated interventions or programs located in LMICs per the World Bank classification in the year the study was costed.

Studies were excluded if they:

• Were letters to the editor, commentaries, protocols, opinion pieces, policy briefings, or conference abstracts. Although systematic reviews were excluded, their reference lists were searched for potentially eligible studies.
• Assessed the economic impact of digital add-ons to CHW programs (e.g., mobile phone interventions), as the focus of our review was on the economic evaluation of CHW-led interventions themselves, not digital add-ons.

No restrictions were placed on the time frame of the analysis or language of publication. Although the search was conducted in English, full texts were reviewed in any language. Studies were not excluded based on quality due to the high diversity in study types and the interest in exploring the breadth of available evidence. Full eligibility criteria are detailed in the PICO framework in the Supplementary Material (S2 Text).

Study screening process

Following a search of the databases and grey literature by a qualified information search specialist, citations were exported to the Covidence platform [20]. Duplicate results were removed using an automated ‘de-duplicate’ feature within Covidence. A team of 18 researchers took part in the screening process. Each title and abstract were reviewed independently by two reviewers. Conflicts were resolved by a third reviewer. Two independent reviewers conducted a full-text review to determine final inclusion or exclusion. Conflicts at this stage were resolved by a third reviewer.

Data extraction

Data were extracted into a custom Google Sheets document and rigorous quality control was conducted on the extractions. A third reviewer (a health economist) was available to resolve any disagreements. The data extraction form was tested for user-friendliness and completeness by all extractors independently and discussed during a joint video conference call. The spreadsheet captured the article meta-data, information about the study site and CHWs involved in the study, methodological and reporting data, as well as outcomes and cost data.

Outcomes were categorised into five categories: (i) Service Provision (e.g., visits, number of medications distributed, number of household visits); (ii) Population Coverage (e.g., households covered); (iii) Mortality and Morbidity outcomes (e.g., reduction in mortality, lives saved); (iv) Cost Savings and Cost Recovery outcomes (e.g., amount of money saved); and (v) Societal Outcomes (e.g., economic growth).

We documented whether costs were reported in the following categories: (i) Cost per CHW; (ii) Cost per Consultation; (iii) Cost per Service; (iv) Cost per Capita; or (v) Cost per Beneficiary. We also extracted other cost reporting that is specific to RMNCH such as the Cost per Mother-child dyad seen. We documented whether cost per outcome was reported in the study (i.e., Cost per Disability-Adjusted Life Year (DALY) Averted). All costs were converted to and reported in 2024 US$ to facilitate comparison. For costs reported in US$, we first converted costs to local currency units (LCUs) of the same year using that year’s exchange rate (World Bank, ‘US$ per LCU, period average’). With the costs reported in US$ now in LCUs, and for costs originally reported in LCUs by the resource, we inflated costs to 2024 LCUs using LCU inflation rates reported by the International Monetary Fund (‘inflation, average consumer prices’). With all costs in 2024 LCUs, we converted costs to 2024 US$ using the ‘LCU per US$, period average’ official exchange rate for 2024. For cost-effectiveness, we report the incremental cost-effectiveness ratios (ICERs) converted to 2024 US$. An ICER reports the difference in total costs (incremental cost) of the CHW program and the comparator divided by the difference in the chosen measure of health outcome or effect (incremental effect) to provide a ratio of ‘extra cost per extra unit of health effect’.

We also report on whether the study authors drew conclusions on the cost-effectiveness and affordability of the CHW program. To determine cost-effectiveness, we looked for comparisons against: (i) thresholds (willingness to pay or gross domestic product (GDP)/capita); or (ii) an alternative service or delivery modality, such as facility-based care. For affordability, we noted whether authors reported how the intervention affects the overall healthcare budget (budget impact analysis), including whether the intervention is affordable within the current budget constraints. That said, we report on cost-effectiveness and affordability based on the authors’ determination or conclusions from the original study, regardless of whether a threshold was used.

We used Microsoft Excel to organise and analyse extracted data.

Patient and public involvement

Patients and the public were not consulted as part of this scoping review.

Ethics approval

A self-assessment was conducted via the University of Washington Human Subjects Institutional Review Board (IRB) which determined that this study was not human subjects research and did not require IRB review.

Search results

The initial broader literature search, which included maternal, newborn and child health, and other health areas (e.g., (i) non-communicable diseases, (ii) neglected tropical diseases, (iii) infectious diseases [21], and (iv) horizontal, integrated CHW programs [22] - some of which are currently under review or in press), yielded 9,790 articles. This was reduced to 5,663 after the removal of duplicates. A further 5,345 studies were excluded following abstract screening and an additional 170 after full-text review.

After coding studies by health/disease area, this process resulted in 53 RMNCH studies being included in this review. Each of the 53 studies contained multiple scenarios (for example data reported from various countries, various intervention models, or varying study perspectives) which are described in the findings tables. Further details can be found in the PRISMA flow chart (Fig 1).

Download:

• PNG
  larger image
• TIFF
  original image

Fig 1. PRISMA-ScR flow diagram.

A total of 9,790 records were identified through database searches. After removing 4,127 duplicates, 5,663 records were screened. Of 318 full-text reports assessed for eligibility, 53 studies met inclusion criteria. Exclusions were based on relevance to CHWs, presence of economic evaluation, and RMNCH focus.

https://doi.org/10.1371/journal.pgph.0004893.g001

This review includes several articles that report summarized data from individual studies, which are also part of our review [23–30]. These summary papers sometimes took the form of a meta-analysis, and sometimes simply re-analyzed the original data differently, producing different results. To avoid duplicating cost data across these summary papers, we conducted a comprehensive screening of each situation where both summarized data and an individual article reporting on the same study were included. The extracted data from these papers were exported and compared row by row in an Excel spreadsheet, a process carried out by three of the authors. After a thorough comparison, we found no overlap in the reported cost data between the summary and individual papers, and as a result, we included all of them in this review.

The following two subsections present summarized findings for cost, cost-effectiveness and affordability findings separately for (1) RMNH (2) Child health. Each section describes the CHW programs and alternatives assessed and reports on the relevant cost, cost-effectiveness and affordability findings. For more detailed findings, please see the supplementary files (S1 Data and S2 Data).

1. 1. Reproductive health, maternal and newborn care

We found a total of 21 reproductive, maternal and newborn health-focused studies, covering 77 scenarios across reproductive health (n = 4 studies, n = 5 scenarios) [31–34] and maternal and newborn care (n = 17 studies, n = 72 scenarios) [23–29,35–44]. These were conducted in 12 countries: Bolivia [24,25], Ethiopia [24–26,32], Ghana [24,25,27], India [35,41,42], Kenya [34], Malawi [24,25,39], Mozambique [35], Pakistan [35,40,43], Sierra Leone [37], South Africa [24,25,36,38], Tanzania [24,25,28,31], and Uganda [24,25,29,33]. CHWs were mostly compared to unspecified “standard” care. Table 1 summarizes the findings from the reproductive health, maternal and newborn care studies.

Download:

• PNG
  larger image
• TIFF
  original image

Table 1. Reproductive, maternal and newborn health findings.

https://doi.org/10.1371/journal.pgph.0004893.t001

Cost-effectiveness.

Cost-effectiveness was assessed for 51 scenarios, with the majority (40/51) [25–28,33,34,37,40–44] concluding CHWs were cost-effective, with another 4 scenarios [39] concluding that the CHW intervention was “likely” cost-effective. This conclusion was based on GDP per capita (29/40) [25–27,37,40–42] or willingness to pay (2/40) [34,43] thresholds, comparing CHWs with alternatives (8/40) [28,33] and return on investment results (1/40) [44]. Examples of cost-effective CHW interventions included the distribution of contraceptives (compared to facility-based distribution and self-injections) in Burkina Faso, Uganda and Senegal [33], CHW-led interventions for newborn health through home visiting (compared to standard facility-based care) in Ethiopia and Pakistan [26,40], and an intervention to prevent and treat perinatal depression (compared to standard facility based care) in Pakistan [43]. For two scenarios, the conclusion was that the CHW intervention was not cost-effective, based on the cost per life year saved far exceeding both the GDP per capita threshold (n = 1) [37] and the cost per life year saved of other exclusive breastfeeding interventions implemented in the same context though evaluated in other studies (n = 1) [38]. Three scenarios found the CHW intervention as implemented not to be cost-effective, but predicted it would be cost-effective if the number of contacts between CHWs and pregnant women increased [35].

Less than half of scenarios (n = 32) [23,25,26,29,36,39] explicitly considered the affordability of CHWs, and only three of these [23,29] concluded CHWs to be affordable based on program costs as a share of public health expenditure per capita (0.63% to 1.8%). These affordable scenarios involved CHWs conducting pregnancy visits, delivering Integrated Management of Childhood Illness, and occasionally attending home deliveries across varying coverage levels in Bolivia. The remaining scenarios (n = 29) had no affordability conclusions [25,26,36,39].

1. 2. Child health

We found a total of 32 child health studies covering 84 scenarios across preventing child mortality and promoting survival (n = 9 scenarios) [46–49], childhood development (n = 7) [50–52], nutrition (n = 30) [53–65], and child-specific infectious disease prevention and management (n = 38) [30,66–76]. These were conducted in 18 countries: Bangladesh [53], Burundi [72], Ethiopia [30,48,73], Ghana [30,46,67], India [57,66], Kenya [50,56,65,71], Madagascar [74], Malawi [32,50], Mali [30,48,59,62], Mozambique [30], Nepal [64], Niger [30,49,54,63], Nigeria [55], Pakistan [58], Rwanda [52], South Africa [47,51], Tanzania [61] and Uganda [60,66–70,75,76]. CHWs were mostly compared to alternatives such as facility-based services, unspecified “standard” care or drug sellers (trained or untrained workers at private drug shops). Table 2 summarizes the findings from the child health studies.

Download:

• PNG
  larger image
• TIFF
  original image

Table 2. Child health findings.

https://doi.org/10.1371/journal.pgph.0004893.t002

Methodological findings across all studies

In this section, we summarise selected methods-related findings across all included studies (n = 53) and scenarios (n = 161).

Across studies in the two focus areas, the most commonly reported outcomes were cost per beneficiary (72% of scenarios), cost per consultation (36%) and cost per capita (34%). Nine percent of scenarios reported cost/DALY averted.

The majority (55%) of scenarios took a provider perspective, either explicitly stated (n = 67 scenarios) or inferred (n = 22). Fifty-three percent of scenarios employed a one-year time horizon, which is helpful for comparability across studies. Recurrent costs were included in 98% of scenarios. There was considerable variability in reporting on training costs (76% of scenarios), non-training capital items (meaning items used over one year, such as equipment) (76%), indirect costs or overheads (66%), and out-of-pocket and opportunity costs (56%). A limited number of studies reported costs averted (12%).

Only nine studies (representing 11% of scenarios) used the Consolidated Health Economic Evaluation Reporting Standards (CHEERS) checklist [12], despite it being the leading guidance for authors to follow to ensure that health economic evaluations are identifiable, interpretable, and useful for decision-making.

Cost findings and variability

Across reproductive health, maternal and newborn care and child health, CHW-led interventions in our review were associated with widely varying cost estimates, as reflected in common metrics such as cost per beneficiary (ranging from $0.02 to $1,547), cost per capita (ranging from $0.09 to $20.25), and cost per consultation (ranging from $0.26 to $52.91). The wide ranges reflect substantial differences in program scale, intensity, local resource costs, and the scope of CHW responsibilities. For example, CHWs in the included studies performed a variety of tasks—from providing basic home-based care and health education to screening for complications and referring patients for facility-based care—complicating attempts to derive standardized “typical” costs. There are significant contextual differences in CHW employment, integration, and support across countries. Moreover, compensation strategies for CHWs varied from salaries (with amounts as high as $3,216 per month in some maternal care scenarios) to stipends of only a few dollars per day, and in some cases, non-remunerated volunteer labor. These differing remuneration models do not always align with global policy recommendations to formally compensate CHWs and make cross-study comparisons challenging [77]. Finally, there are also methodological differences between studies which may be driving cost findings, such as the inclusion of different cost items or the use of different time horizons. Together, these issues make it challenging to compare cost findings across studies and deem it unlikely that cost findings from this review are useful for planning and budgeting purposes at sub-national, national or global levels. We caution against overgeneralizing findings without considering the broader structural social and health determinants that drive maternal and child health outcomes.

Cost-effectiveness

This review documents 100 scenarios where authors assessed cost-effectiveness, over three-quarters of which found CHWs to be cost-effective compared to different alternatives or thresholds. Within child health, the evidence indicates that CHW interventions are cost-effective, based on favorable findings in 37 of 49 scenarios where cost-effectiveness was assessed. In 22 of 37 scenarios, CHWs were found to be more cost-effective than alternatives such as health facility-based service delivery. Examples of cost effective child health interventions included parenting groups, screening and treatment of malnutrition, and iCCM. For maternal health related interventions, examples of cost effective interventions included distribution of contraceptives and newborn care through home visiting. This does not necessarily mean they were less expensive than alternative care: cost-effectiveness considers both costs and effectiveness, and in some cases CHWs were more expensive than alternative delivery mechanisms but also more effective, driving their cost-effectiveness. Other scenarios found CHWs to be cost-effective when compared against thresholds such as GDP/capita or willingness to pay. Although our scoping review did not include assessing the quality or robustness of authors’ reported conclusions about cost-effectiveness, we do find it important to note that the GDP/capita threshold in particular is no longer widely accepted for concluding about cost-effectiveness. Multiple critiques have noted that it fails to consider local resource constraints and results in too many interventions being considered cost-effective, meaning findings cannot be useful for supporting decision-making. It is possible that CHWs in these scenarios would still be found to be cost-effective if findings were considered against new, alternative thresholds that are lower than those based on GDP/capita, but this analysis was outside the scope of our review. Nevertheless, the predominance of “cost-effective” findings based on comparison with an alternative does reinforce the general promise of child health interventions delivered by CHWs, especially in settings where resource constraints limit access to facility-based care and where CHWs can help overcome barriers such as travel distance and shortage of skilled providers.

There are some differences by sub-area. Within child health, the evidence around CHWs involved in nutrition and child-specific infectious disease prevention and management is sizeable (n = 68) and overall positive about the cost-effectiveness of CHWs, while the evidence base is much smaller (n = 16) and less conclusive around interventions on preventing child mortality or promoting child survival as well as child development.

For reproductive health, maternal and newborn care, while the majority of scenarios where authors assessed cost-effectiveness were reported to be cost-effective or likely cost-effective (44 of 51 scenarios), a large number of these are based on the GDP/capita threshold, subject to the limitations mentioned above.

Amongst all 81 scenarios concluding CHWs were cost-effective or likely cost-effective, CHWs received a salary or stipend or were valued hypothetically in around half of these, with other scenarios representing unpaid (volunteer) CHWs or situations where CHWs received non-monetary incentives such as raincoats. This brings into question whether the cost-effectiveness evidence would still be positive in the remaining scenarios if the CHWs were remunerated in alignment with global policy recommendations.

Affordability and budget impact

A program can be cost-effective yet still remain unaffordable for governments if the total budget required exceeds available financial resources. Thus, an affordability or budget impact assessment - and not just cost-effectiveness results - are critical for translating evidence into policy. Despite the positive cost-effectiveness findings, only 18 child health scenarios examined affordability or budget impact. For reproductive health, maternal and newborn care, only 29 scenarios examined affordability. Importantly, only two child health scenarios (and no reproductive health, maternal and newborn care scenarios) concluded that CHWs would be affordable given that they generated cost savings compared to alternative cadre (nurses), while five concluded that program costs as a percentage of public health expenditure per capita would be considered affordable; only two of the five reported this percentage (0.6% and 1.8%). Other studies reported the percentage without drawing a conclusion about the actual affordability of the reported percentage (0.3% to 7.4%). As with the cost-effectiveness results, the type and amount of CHW remuneration are crucial here. Further, when CHWs are paid, these workforce costs can appear significant relative to tight government budgets, especially if savings generated as a result of task shifting from more expensive cadres is not considered. Overall we find too little evidence to draw conclusions about CHW affordability in LMICs.

Implications for policy and practice

Taken collectively, these findings suggest that certain CHW-led RMNCH programs can be a cost-effective strategy to improve outcomes compared to alternative delivery modalities, such as health facility-based service delivery, particularly for child health. However, given the heterogeneity of interventions, study methods and study contexts, these findings should be interpreted cautiously. To facilitate evidence-informed policy decisions, future economic evaluations must adopt rigorous and transparent methodologies, explicitly address affordability, and use consistent, comparable outcome measures such as DALYs. Policymakers should interpret cost-effectiveness and affordability findings in conjunction with local budget realities and other decision-making criteria, including equity, population coverage, and the health system’s capacity to absorb and supervise CHWs at scale.

Strengths and limitations of this review

A strength of this review is its comprehensive scope, including both published and gray literature across multiple regions and a wide range of RMNCH interventions. The systematic approach and disaggregation of 161 scenarios allow for a granular analysis of how RMNCH-focused CHW-led programs have been evaluated economically. However, the inherent limitations of a scoping review remain: we did not exclude studies based on quality (nor did we conduct a quality analysis) [78], and our synthesis is constrained by the heterogeneity of methods and reporting standards. Moreover, we did not attempt any reinterpretation of data reported using current cost-effectiveness thresholds, but instead reported them as the authors did in respective papers. Finally, while we have brought all cost-related findings to 2024 US$ for comparability, we have not adjusted for purchasing power parity.

Reflexivity statement

We adhered to the consensus statement on equitable authorship in international research collaborations as outlined by Morton et al., (2021) [79]. The following reflexivity statement is provided in that context. This research was conducted by a multidisciplinary, global team comprising researchers and practitioners from LMICs where CHW programs are implemented, such as Malawi, Rwanda, Uganda, Kenya and Liberia. Team members hold positions in academic institutions, non-governmental organisations, and frontline health services (and included CHWs), enabling us to integrate both theoretical and practical insights into our study. All members who contributed to the study design, implementation, analysis, and writing of this paper have been included as co-authors. We acknowledge that the authorship team does not include LMIC government stakeholders, who are a key audience for this research. While many team members have extensive experience engaging with government health systems, this absence may limit study findings. That said, this study was undertaken by the Community Health Impact Coalition (CHIC), a collective of thousands of CHWs and dozens of global health organisations spanning over 60 countries in five WHO regions. The research questions, data collection methods, and analysis were shaped by CHIC’s commitment to understanding the drivers of impact and quality in CHW-delivered care globally.

Importantly, this work was also shared with CHWs to explore their opinions and solicit their feedback. Their insights were integral to refining our approach and ensuring the relevance of our findings to those most directly impacted by CHW programs. For a detailed reflexivity checklist, please refer to the Supplementary Material (S2 File).

Future research

Several key areas warrant further exploration. First, consistent use of a standardized reference case - specifying the analytic perspective, time horizon, and outcome measures - would greatly enhance comparability. Second, studies should incorporate budget impact analyses to provide more rigorous assessments of affordability, addressing a critical gap faced by policymakers in determining whether and how to scale CHW programs. As health service delivery shifts to integrated models, future research is likely to consider more horizontal, integrated CHW programs as opposed to the vertical ones assessed in this review.