The Effect of Cash Transfer on Maternal and Child Health Outcomes: A Systematic Review in Sub-Saharan Africa

Background: Evidence-based policies are critical to navigate the time post-COVID-19. Improving maternal and child health is core to any recovery strategy. Cash Transfer programs can improve maternal and child health outcomes in low- and middle-income countries. However, studies assessing the effectiveness of these programs on maternal and child health outcomes, beyond service utilization, remain inconclusive. This synthesis �lls a gap in the existing literature on the effectiveness of these programs in improving mothers and children’s health outcomes beyond service utilization and suggests a framework for reporting such health outcomes. We focused on Sub-Saharan Africa because of substantial operational differences between regions, and the need for maternal and child health advancement in this region. Methods: This review searched PubMed Central and Google Scholar, and supplemented with a backward citation search, for studies conducted in Sub-Saharan Africa using a combination of MeSH terms and key words for the period between 2000 – 2021. The review included only studies on cash transfer programs that reported health outcomes beyond service utilization among women of reproductive age and children below 18 years old. Findings: The synthesis pulled data from 17 articles conducted in 6 Sub-Saharan African countries. All studies reported health outcomes measures other than service utilization, and programs targeted women of reproductive age and children under 18 years of age. Of the 17 articles, 1 reported measures of mortality, 9 reported measures of functional status; 3 reported subjective measures of well-being, and 4 reported behavioral health outcomes. Across all categories of reported measures, evidence emerges that cash transfer programs are associated with improved health outcomes (e.g., improved infant and child survival, reduced incidence of illnesses, improved cognitive and motor development, improved general health, delayed sexual debut, lower transactional sex, etc.). Conclusion: Cash Transfer programs are effective, with great potential to improve maternal and child health outcomes in Sub-Saharan African countries. However, further research is needed to address implementation challenges, which include data collection, and program management.


Background
Populations in Low-and Middle-Income Countries (LMIC) continue to face daunting challenges. Mothers and children are the most affected by these challenges, which include a double burden of communicable and non-communicable diseases, higher infant and maternal mortality, malnutrition, etc. These Although the full impact of the COVID-19 on maternal health and child survival remains unknown, recent studies have predicted devastating impacts. For instance, if routine healthcare is disrupted (e.g., 9.8% to 51.9% reduced coverage) and access to food is decreased (e.g., 10% increased prevalence of wasting), the resulting consequences will be devasting. Such disruptions could translate to 253,500 additional child Cash Transfer (CT) programs are one of the public health interventions that policymakers and funders can rely on to improve MCH outcomes in LMICs. These programs were initially launched in Latin America (i.e., Brazil and Mexico) in the mid-90s to provide education and health to historically "excluded" populations, before being adopted in other LMICs (Turley, 2020).
CT programs represent a "set of public and private policies or programs aimed at preventing, reducing and eliminating economic and social vulnerability to poverty and deprivation" (UNICEF-ESARO, 2015, p. 1). They are often implemented as a direct transfer payment of money to eligible persons and can be either conditional or unconditional. Unconditional Cash Transfer (UCT) programs provide cash to all eligible and registered bene ciaries; whereas Conditional Cash Transfer (CCT) programs require an eligible person to take a speci ed course of action, also known as co-responsibilities or conditionalities, to receive a bene t ( In Africa, major CT programs such as the Productive Safety Net Program's Direct Support (PSNP-DS) in Ethiopia and the Hunger Safety Net Program (HSNP) in Nigeria were designed to address food insecurity. Leveraging on the success of these programs, African governments and external donors, mainly the World Bank, expanded CT programs to address other challenges, including access to healthcare. Still, many CT programs are geared toward relief or development goals (Garcia et

Rationale for the study
Findings about the effectiveness of these programs to improve MCH outcomes, beyond service utilization, remain inconclusive (Onwuchekwa et al., 2021;Singh et al., 2021). Most evaluations of these programs tend to focus on the utilization of healthcare services by the bene ciaries and other social determinants of health, thereby overlooking speci c health outcomes related to mortality, individuals' capacity to function, or the subjective sense of well-being (Parrish, 2010). The only synthesis we found examined mental health outcomes among young people under 25 years of age (Zimmerman et al., 2021). As suggested by Parrish (2010, p. 1), "positive health outcomes for people include being alive; functioning well mentally, physically and socially, and having a sense of well-being". On this account, it could be argued that service utilization is one underlying pathway through which CT programs affect health outcomes. However, no research synthesis has been done on these outcomes as the end points among In this time of competing priorities across the world, allocation of scarce resources needs to rely on evidence-based science. Both governments, and external funders navigate resource constraints as they allocate funding. Lack of evidence on a given programmatic effort, could undermine not only funding allocation, but also any progress achieved. Hence, it is crucial to ll any gaps in the existing literature with respect to the effectiveness of CT programs and their impact on MCH outcomes, to keep these programs running, and to advocate in favor of scaling to other geographies and demographics.
Finally, some of the major barriers to the successful implementation of these programs include the lack of transparency, an endemic corruption, and lack of valid data (Wright et al., 2018). These limitations and issues represent opportunities for new areas of investigation.

Scope of synthesis
This research synthesis focuses on the relationship between CT programs and MCH outcomes, other than service utilization. It covers studies that evaluated the effect/impact of CT programs on MCH outcomes in Sub-Saharan African countries for the period between 2000 -2021. The focus on these countries stems from the fact that they are facing similar socioeconomic and political situations, and often share the same social norms, which are associated with populations' health behaviors ( This synthesis addressed the lack of evidence on the effectiveness of CT programs to positively impact MCH outcomes, beyond service utilization-based metrics. Outcome indicators de ned in this review were based on mortality, subjective health state, experiential, and psychological state, and on the ability to function. Protective behaviors that promote health were also considered.

Research questions
1. To what extent CT programs can improve MCH outcomes (functional, subjective sense of well-being, experiential state, and death) in Sub-Saharan African countries? 2. What are the potential pathways through which CT programs in uence MCH outcomes?

Methodology
We searched PubMed Central and Google Scholar for published peer-reviewed papers from 2000-2021. This research synthesis followed the Preferred Reporting Items for Systematic Reviews and Meta-Analysis (PRISMA) checklist (Fig. 1).
Based on the above research questions, we used a combination of Medical Subject Headings (MeSH) and keywords, incorporating Boolean operators, truncation, and eld tags. Our search string was de ned as followed: (("cash transfer*" OR "cash incentive*") AND ("maternal and child health" OR "maternal and child health outcome*") AND "Africa".
We also conducted a backward citation search, in which we looked at the reference list on the most recent and relevant systematic review, manually searching Google Scholar for relevant studies.
Inclusion and exclusion criteria Studies were screened for eligibility and irrelevant articles were excluded based on title and abstract. Selected studies were fully reviewed for data extraction. The review included only peer-reviewed studies, whose CT programs targeted women of reproductive age (15-49 years) and children (< 18 years old) and reported health outcomes other than service utilization. Studies of interest were implemented in the Sub-Saharan African countries between 2000-2021.
Health outcomes included a measure of functional status (e.g., according to the International Classi cation of Functioning, Disability and Health); or a subjective sense of well-being, or mortality (survival). In addition, we included studies that reported some protective health behaviors (e.g., administration of nevirapine to prevent mother-to-child transmission of HIV).

Data extraction and coding
Results from the search string in PubMed were exported in a text format le (i.e., .txt) and saved in Excel.
We rst screened all the titles of articles to remove duplicates and studies that were conducted outside of Sub-Saharan African region, protocol, and systematic reviews. We then screen the abstracts of the remaining articles to further exclude irrelevant studies (e.g., studies not using the speci ed study population, or did not report the de ned health outcomes). Included articles were fully reviewed and the following data were extracted: the name of the author, the year of publication, the title of the study, the study design, the name of the intervention if provided, the country of the intervention, the outcome assessed, the method used to evaluate impact, the reported measure of effect and the 95% con dence intervals or the standard errors, and the covariates. Since all studies were RCTs, they were evaluated for their quality, using the Consolidated Standard of Reporting Trials (CONSORT) 2010 check list and a risk of bias assessment was done using the revised Cochrane risk-of-bias tool for RCTs (RoB 2).

Results
Final studies included 5 papers by the same group of authors evaluating a program implemented in Malawi, and 3 papers from the same authors for a program implemented in Kenya. Table 1 below presents results from the synthesis by type of health outcome because many studies relied on the same dataset, while investigating a diverse set of outcomes or end points (Table 1).
Almost all included studies were RCTs implemented in Sub-Saharan African countries (i.e., Malawi, Kenya, South Africa, Nigeria, Tanzania, and Zambia). Assessed outcomes included: illness in past 30 days, illness that stopped normal activities in the past 30 days, sexual debut, risky sexual behaviors (e.g., unprotected sex, multiple partners, transactional sex), depression, incidence of HIV and HSV-2, subjective sense of well-being, child survival, fetal loss, fetal death, stillborn, infant death, child anthropometric measures, prevention of mother to child HIV transmission by the administration of nevirapine, and Early Infant Diagnosis of HIV (EID), happiness, well-being, child cognitive, language and motor development, child anthropometry.

Page 7/18
Outcomes were grouped into three main groups based on the classi cation suggested by (Parrish, 2010): mortality, functional status, and subjective sense of well-being. In addition, an additional category was added to account for the relevance of some protective health behaviors reported. Health-promotion behaviors 4 (23.5)

Prevention of mother to child HIV transmission 3
Tetanus toxoid vaccination 1 Protective sexual behaviors (e.g., condom use, etc.) 1 Of the 17 articles included in the nal review, health outcomes related to individual capacities to function were reported in 9 articles, representing 52.9% of the articles included in the review. Protective behaviors such as administration of nevirapine to prevent mother-to-child transmission of HIV, etc. were reported in 23.5% (n = 4) of the articles. Three articles out of the 17 (17.6%) reported well-being outcomes, including subjective well-being, a mother's general health, happiness, and self-reported quality of life. Only one article reported mortality or child survival outcomes.

Mortality
The only article reporting mortality included outcomes on child survival, fetal death, stillborn, and infant death. The article reported a positive association between CCT programs and child survival or reduced mortality (i.e., a child who was in-utero at enrollment was alive at follow-up). In Nigeria, a CCT program was associated with substantial increases in child survival (ITT: 0.0606; std. error: 0.0098, p-value < 0.01). The increase in overall child survival was driven by a large decrease in fetal losses (29%-point decrease in the treatment group relative to the control group mean). Potential causal mechanisms included conditionalities attached to a cash payment of $14 (i.e., if eligible pregnant women used a package of health services consisting of at least three antenatal care visits, a health facility delivery, and one postnatal visit). Provision of other essential health interventions (e.g., antenatal care, immunization, screening for childhood pneumonia, etc.) at the beginning of life, are also key mechanisms that could improve child survival (Fig. 2).

Behavioral and protective interventions
Although the main focus of the study was to synthesize the existing literature with respect to the three outcome categories suggested by Parrish (2010, p. 1), we expanded the review to accommodate protective behavior promoting health. These outcomes are important given their protective nature and the risk associated with the lack thereof. The present review captured protected sex, tetanus toxoid vaccination, and health visits for nevirapine intake to prevent mother-to-child transmission of HIV infection. Studies reported a positive effect of CCT programs on the incidence of HIV and other sexually transmitted infections (STIs) through behavioral mechanisms such as condoms use, reduced number of sexual partners, reduced incidence of transactional sex, administration of nevirapine to prevent mother-tochild transmission of HIV infection. Another study looked at reception of tetanus toxoid vaccination during the perinatal period. The study showed that CCT programs was associated with increased likelihood of receiving the vaccine and therefore of being protected against maternal and neonatal tetanus (OR for those who received C300: 3.362; 95% C.C.: 2.595-4.354; and for those who received C800: 7.575; 95% C.I.: 5.648-10.160).

Discussion
This review synthesizes evidence on the impact of CT programs on MCH outcomes other than service utilization, in Sub-Saharan African countries. Although initially designed to address nutritional needs in Latin American countries (Shei, 2013), CT programs have gained momentum around the world, including among African countries.
A lack of evidence on the effectiveness of CT programs to address MCH outcomes other than service utilization persists. This synthesis provides evidence that CT programs effectively improve MCH outcomes beyond service utilization. For instance, the study found that CCT programs are effective in averting fetal and infant deaths and improving child survival. This synthesis suggested that the conditionalities attached to most of the CCT programs also represent pathways through which CT programs in uence health outcomes ( However, nancial constraints (e.g., fees for healthcare services, transportation, payment of medicines, etc.) and cultural beliefs remain signi cant barriers to healthcare access in most of these countries (Bright et al., 2017;Geleto et al., 2018). Hence, providing nancial incentives to encourage women and their children to bene t from those interventions during critical moments (e.g., gestational period, postdelivery and during early childhood) has a signi cant impact on their survival and functional experiences.
However, performance of this behavior could be modi ed by lack of motivation, and other cognitive factors. Therefore, attaching conditionality to such programs can contribute to improving adherence to promoted behaviors.
The synthesis also found evidence that CCT programs were associated with improved functioning capacities (e.g., lack of illness in the past 30; no report of being unable to perform usual activities because of illness, reduced depressive symptoms, and incidence of HIV and HSV-2). Across studies, there seems to be a consensus that CCT programs impact these outcomes through two underlying mechanisms, namely education ( UCT programs were associated with improved emotional health (e.g., subjective sense of well-being, selfreported quality of life). UCT programs impact people's sense of well-being, by providing nancial support to people, whose lives are subjected to daily concerns over the basic needs of food, shelter, clothing, etc. Given that emotions in uence health directly (e.g., through the activation of the hypothalamic-pituitaryadrenal axis) and indirectly (e.g., through health behaviors), fewer episodes of stress imply improved health among participants (Berkman, Kawachi, & Glymour, 2014). For instance, a study reported that the UCT program was associated with positive feelings about children's future, and generally happiness among children in recipient households (Natali et al., 2018).
Although we found supporting evidence on the positive impact of these CT interventions on MCH outcomes, it should be underscored that one of the key issue plaguing the implementation of these programs is the lack of transparency and auditability at all levels (Jaffer et al., 2020; Wright et al., 2018). For instance, the decision on where to implement these programs (e.g., selecting a given state or province rather than others), and who should bene t from such programs cast questions on its transparency, and accountability. Quality data and information are fundamental for effective implementation and evaluation of programs (Owino, 2020). The lack thereof exposes the program to shortcomings (e.g., corruption, lack of transparency, and accountability), and to missed opportunities for improving its robustness. In the context of CT programs, data collection remains a very expensive and often complex operation, plumbing their effective implementation.
Hence, the need for a tamper-proof system, such as blockchain technology, that provides security, privacy, con dentiality and most important, a decentralization of data collection activities (Shahnaz, Qamar, & Khalid, 2019). The relevance of this approach in the context of Sub-Saharan African countries stems from the fact that most of these countries rely on international aid, often tied to stringent conditionalities, to implement, and sustain such programs (McShea, 2019). When the lack of transparency and accountability becomes an issue, most of the external donors become hesitant to fund the program. Hence, given the abovementioned evidence on the effectiveness of CT programs on improving MCH outcomes, it is crucial to improve the implementation of these programs to ensure current and potential stakeholders regarding the actual use of the funds and the impact thereof. Blockchain technology, through its smart contracts, could be one possible and sustainable solution.
Also, future studies should consider using a standardized framework for assessing health outcomes, such as the one proposed by Parrish (2010, p. 3). Also, the International Classi cation of Functioning, Disability and Health could be used to inform selection and reporting of measures of functional status.
The scope of this study was to synthesize the existing literature in a rather nascent eld. Further research synthesis should expand this research to draw more statistical conclusions, not only in Sub-Saharan African countries, but across LMICs.

Conclusion
We found compelling evidence that in Sub-Saharan African countries, CT programs are effective interventions with a positive impact on MCH outcomes, including survival experience, functional status, and emotional health. The quality of evidence is higher because most studies were randomized controlled trials. The underlying mechanisms through which CT programs operate have been discussed. Policymakers and funders should tailor such interventions based on the mechanisms discussed. However, more needs to be done to improve transparency in the implementation of these programs.

Declarations
Ethics approval and consent to participate Not applicable

Consent for publication
Both authors gave consent for publication of the nal manuscript.
Availability of data and materials Data sharing is not applicable to this article as no datasets were generated or analyzed during the current study. However, references to the original papers are provided in appendix. Pathways through which CT programs impact MCH outcomes

Supplementary Files
This is a list of supplementary les associated with this preprint. Click to download. Appendix.docx