Effectiveness of cash-plus programmes on early childhood outcomes compared to cash transfers alone: A systematic review and meta-analysis in low- and middle-income countries.

BACKGROUND: To strengthen the impact of cash transfers, these interventions have begun to be packaged as cash-plus programmes, combining cash with additional transfers, interventions, or services. The intervention's complementary ("plus") components aim to improve cash transfer effectiveness by targeting mediating outcomes or the availability of supplies or services. This study examined whether cash-plus interventions for infants and children <5 are more effective than cash alone in improving health and well-being. METHODS AND
FINDINGS: Forty-two databases, donor agencies, grey literature sources, and trial registries were systematically searched, yielding 5,097 unique articles (as of 06 April 2021). Randomised and quasi-experimental studies were eligible for inclusion if the intervention package aimed to improve outcomes for children <5 in low- and middle-income countries (LMICs) and combined a cash transfer with an intervention targeted to Sustainable Development Goal (SDG) 2 (No Hunger), SDG3 (Good Health and Well-being), SDG4 (Education), or SDG16 (Violence Prevention), had at least one group receiving cash-only, examined outcomes related to child-focused SDGs, and was published in English. Risk of bias was appraised using Cochrane Risk of Bias and ROBINS-I Tools. Random effects meta-analyses were conducted for a cash-plus intervention category when there were at least 3 trials with the same outcome. The review was preregistered with PROSPERO (CRD42018108017). Seventeen studies were included in the review and 11 meta-analysed. Most interventions operated during the first 1,000 days of the child's life and were conducted in communities facing high rates of poverty and often, food insecurity. Evidence was found for 10 LMICs, where most researchers used randomised, longitudinal study designs (n = 14). Five intervention categories were identified, combining cash with nutrition behaviour change communication (BCC, n = 7), food transfers (n = 3), primary healthcare (n = 2), psychosocial stimulation (n = 7), and child protection (n = 4) interventions. Comparing cash-plus to cash alone, meta-analysis results suggest Cash + Food Transfers are more effective in improving height-for-age (d = 0.08 SD (0.03, 0.14), p = 0.02) with significantly reduced odds of stunting (OR = 0.82 (0.74, 0.92), p = 0.01), but had no added impact in improving weight-for-height (d = -0.13 (-0.42, 0.16), p = 0.24) or weight-for-age z-scores (d = -0.06 (-0.28, 0.15), p = 0.43). There was no added impact above cash alone from Cash + Nutrition BCC on anthropometrics; Cash + Psychosocial Stimulation on cognitive development; or Cash + Child Protection on parental use of violent discipline or exclusive positive parenting. Narrative synthesis evidence suggests that compared to cash alone, Cash + Primary Healthcare may have greater impacts in reducing mortality and Cash + Food Transfers in preventing acute malnutrition in crisis contexts. The main limitations of this review are the few numbers of studies that compared cash-plus interventions against cash alone and the potentially high heterogeneity between study findings.
CONCLUSIONS: In this study, we observed that few cash-plus combinations were more effective than cash transfers alone. Cash combined with food transfers and primary healthcare show the greatest signs of added effectiveness. More research is needed on when and how cash-plus combinations are more effective than cash alone, and work in this field must ensure that these interventions improve outcomes among the most vulnerable children.

Introduction

Compared to adults, children are disproportionately affected by poverty and its consequences, with nearly one-fifth of all children living in extreme poverty [1]. Social protection interventions, including cash transfers (direct monetary provision), aim to mitigate the risk and effects of poverty and social exclusion [2]. Cash transfers often do not have restrictions on how the cash is spent but may or may not have conditional requirements to receive the transfer (e.g., regular child growth monitoring). A well-established evidence base indicates that cash transfers improve many aspects of children’s lives, including increased food security and improved school attendance [3]. Nevertheless, impacts have been more mixed and less overwhelming in more challenging areas of child development, including nutrition [4] and health [5].

In efforts to strengthen the impact of these interventions, cash transfers have begun to be packaged as cash-plus programmes, combining cash transfers with other interventions or services (e.g., behaviour change communication (BCC), psychosocial support, or cross-sectoral linkages) [6]. Cash-plus programmes can be classified as multisectoral interventions that seek to improve development outcomes rather than a time-restricted grant to exit poverty [7]. These interventions are argued to be more effective than cash alone because the “plus” component specifically targets factors that are necessary for cash transfers to have impact but that cash alone does not change (e.g., improved mediating outcomes and/or availability of supplies or services). Multisectoral interventions such as cash-plus programmes are conceptually at the forefront of efforts to improve outcomes for infants and young children. Notably, the Nurturing Care Framework for Early Childhood Development proposes an integrated response for young children across health, nutrition, child protection, social protection, and education sectors [8].

Intervening in early childhood is vital for moral, economic, and social reasons. Ensuring that children can reach their full development potential means that they have better health and higher economic earnings as adults, and, in turn, reduce intergenerational transmission of poverty [9]. The multisectoral nature of cash-plus programmes speaks to multiple Sustainable Development Goals (SDGs). Cash transfers are central to poverty alleviation (SDG 1) and in achieving national social protection floors (SDG Target 1.3). The plus-components aim to promote child development through other mechanisms than poverty alleviation, such as through direct nutrition support (SDG 2), health services (SDG 3), early child education (SDG 4), or child protection (SDG 16). The 2 elements of cash-plus programming link across multiple SDGs, which is supported by frameworks that emphasise that achieving each SDG is interdependent with the achievement of other SDGs [10]. This interdependence suggests that multisectoral interventions can have a greater impact than vertical, siloed interventions alone.

Cash-plus interventions hold much promise in accelerating achievement of multiple SDG targets for children. However, to the best of our knowledge to date, there has been no synthesis to evaluate if these multisectoral interventions are more effective than cash transfers alone.

This review sought to answer the question: Are cash-plus interventions for infants and children under the age of 5 more effective than cash transfers alone in improving child health and well-being outcomes across the SDGs?

It is important to emphasise that this review is not comprehensive of all cash-plus programmes for young children. Rather, this review assesses the effectiveness of cash-plus interventions compared to cash alone as opposed to a pure (no-intervention) control. It aims to provide evidence on whether and when combining cash with plus-components can have an accelerating impact in achieving child-focused SDGs.

Methodology

This paper presents a systematic review and meta-analysis for cash-plus interventions compared to cash transfers alone, which are targeted to families with infants and children in low- and middle-income countries (LMICs). This study is reported as per the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines (S1 PRISMA Checklist) [11]. The protocol was preregistered with PROSPERO (CRD42018108017).

Information sources and search

For this review, we searched 11 electronic databases, 27 grey literature sources, and 4 trial registries. Four journals were hand-searched, and 10 experts in the field of development economics and cash-plus interventions were contacted to identify unpublished literature. The information sources and sample search strategy are provided in S1 Text. All searches were completed through 06 April 2021. The search strategy was informed by literature on the review topic and contains categories for the regions and individual countries, the population (infants and children), and terms for social protection interventions, which were replicated and modified from Owusu-Addo and colleagues [12].

Eligibility criteria

Population/Participants

The study had to evaluate an intervention package implemented in one or more LMICs, as defined by the World Bank. The intervention package had to aim to improve outcomes (below) among infants and young children aged 0 to 59 months and be implemented during this age period.

Intervention

The intervention package had to contain at least 2 components. The first component requirement was a cash transfer intervention (SDG 1.3) that met 4 requirements [12], specifically that the programme (1) provide financial assistance at the individual or household level; (2) be noncontributory (i.e., individuals have not paid into the system) and in the form of a nonrepayable, unrestricted grant (i.e., no requirement for how cash was used); (3) aim to reduce the impacts of, or vulnerability to, poverty (monetary or multidimensional); and (4) be disbursed in consistent, predictable intervals.

The second component requirement was having at least 1 “plus” intervention targeting SDG 2 (No Hunger), SDG 3 (Good Health and Well-being), SDG 4 (Education), or SDG 16 (Violence Prevention). Specific plus-components include nutrition support programmes (i.e., food transfers and BCC) to reduce malnutrition (SDG 2.1 to 2.2), interventions to control communicable diseases and reduce infant/child mortality (SDG 3.2 to 3.3), coverage with health insurance (SDG 3.8), early childhood development, care, and preprimary education (SDG 4.2), or violence prevention and parenting interventions (SDG 16.2).

Preliminary searches found that water, sanitation, and hygiene (WASH) interventions were integrated as part of these plus-components (rather than as stand-alone plus-interventions), thus were not assessed as its own plus-component. No restriction was placed on behavioural conditions for receiving cash (i.e., both unconditional and conditional cash transfers were included). Studies comparing the effects of unconditional cash transfers (UCTs) versus conditional cash transfers (CCTs) were excluded because condition monitoring was not conceptualised as a plus component.

Comparison

Studies had to include at least (1) one group receiving the cash-plus intervention and (2) one group receiving cash-only. Because this review aimed to evaluate the added benefit of the plus-component to cash transfers, studies that evaluated no treatment versus cash-plus only (without having a separate cash-only group) were excluded.

Outcomes

Cash transfers alone have had limited impact on distal outcomes for child well-being (e.g., on nutritional status and health [4,5]), despite strong evidence that cash transfers can improve more proximal outcomes, such as improving access to food [3,13,14]. Thus, cash-plus programmes were developed in hopes of achieving improvements in these distal, “third-order” outcomes. This review concentrates on these third-order outcomes that contribute to SDG indicators related to children under 5. These include measures of poverty (including multidimensional poverty), malnutrition (including stunting, wasting, underweight, and obesity), morbidity or mortality (neonatal, infant, and children under 5) including from unsafe water or lack of sanitation/hygiene or infectious disease, psychosocial and cognitive development, and violence against children. Within the included studies that examine these outcomes of interest, we also comment on the impact of proximal outcomes and their contribution in the causal pathway.

Time

Studies conducted between 2000 and 2021 were included. The start year (2000) reflects the year that the Millennium Development Goals were adopted, which began a global roll out of evidence-based interventions, including cash transfers. Additionally, pairing cash transfers to another intervention is new conceptual thinking that is not likely to have been investigated prior to the start date.

Study design

Included studies had to have used an experimental or quasi-experimental design. Studies that did not measure the outcome at both baseline and post-intervention were excluded.

Other exclusion criteria

Studies not published in English, interventions related to pregnancy/childbirth, interventions on farming/agricultural productivity practices, or microfinance/savings interventions.

Study selection and data extraction

Studies were imported into Rayyan and deduplicated prior to screening [15]. Studies were double screened blind to reduce potential bias. Initial agreement between raters was >96%, and reviewers agreed on the final included studies. When multiple publications were available for the same evaluation (i.e., baseline, midline, and endline reports), the endline report is cited, and when endline findings are published in both report and journal article form, the journal article was cited. A standardised data extraction form was used, with categories for population characteristics and context, intervention design and components, outcomes and effects, pathways, and equity evidence [16-19]. Two authors extracted meta-analysis data independently to minimise errors.

Risk of bias

Assessing for risk of bias at the study level, randomised studies were evaluated using the Cochrane Risk of Bias Tool and nonrandomised studies using the ROBINS-I tool [20,21]. All studies were double coded, and disagreements discussed. Piloting of the risk-of-bias assessment on 6 studies had 100% agreement between raters and >95% for the full set of studies. Visualisations were created using robvis software [22]. The quality assessment was used to comment on the strength and limitations of the evidence base and the confidence in recommendations from the synthesis [23]. Due to a limited number of retrieved studies across a range of cash-plus combinations and outcomes, no meaningful analysis could be conducted to assess for publication bias.

Data synthesis

Studies were meta-analysed when there were at least 3 trials for an outcome with similar follow-up times for the same cash-plus intervention; this included outcomes in Cash + Nutrition BCC, Cash + Food Transfers, Cash + Psychosocial Stimulation, and Cash + Child Protection. In instances when a study used a pure (no-intervention) control with 2 treatment arms (cash-only and cash-plus), results were reanalysed using the cash-only arm as the control. The analysis corrected for clustering to address possible unit of analysis error by adjusting the sample size as necessary [24]. Specifically, in cases when studies allocated treatment at the cluster level but presented standard errors that were not adjusted for this clustering, the effective sample size was used in analysis. This value was calculated by dividing the original sample size by the design effect, which is computed using the average cluster size and intraclass correlation [25].

Standardised mean differences (d) for continuous outcomes (i.e., cognitive development and anthropometric z-scores) and log odds ratios for binary outcomes (i.e., anthropometric outcome, violence against children, and positive parenting) were calculated before running the random-effects meta-analyses in R using the Knapp and Hartung adjustment, which accounts for few data points [26-28]. In addition to anthropometric z-score values, odds ratios were calculated from the proportion of children experiencing stunting, wasting, and underweight status. Cognitive development effect sizes were calculated from cognition subscales (Bayley Scale of Infant Development) or aggregate cognitive scale score (McCarthy Scales of Children’s Abilities–General Cognitive Index); general measures of child development (Ages and Stages Questionnaire) were synthesised narratively. Parenting practices were assessed in meta-analyses using the prevalence of harsh discipline and exclusive positive parenting. Results from meta-analyses are displayed using forest plots and heterogeneity results using funnel plots (S1 Fig). Quantitative measures of consistency/heterogeneity are reported (I2 and τ2). Confidence intervals for I2 are also provided for context on the uncertainty of the value [29]. However, there were too few studies to explore heterogeneity further.

Study effects were synthesised narratively when there were too few studies to meta-analyse. When studies were narratively synthesised, effect sizes were transformed to cash-only versus cash-plus comparisons. Intervention design was classified as 1 of 4 models set forth in a previous review of Cash + Parenting Programmes [30]. Quantitative equity effects in impact were noted, defined as either subgroup analyses or interaction effects [31].

Results

Study selection

From 5,097 unique articles identified in the search, 80 full-text articles were reviewed for inclusion. Sixty-three were excluded (see Fig 1 and S2 Text); the majority were excluded either because the study did not have a cash-only group or the study did not meet the criteria for a cash-plus programme. Eleven protocols were identified that could meet the criteria for inclusion in an update to the review; study teams were contacted, who all confirmed no intervention impact data were yet available. Seventeen studies were included in the review, of which 11 were meta-analysed (Fig 1 and S2 Text). The studies included in this review represent 11 unique cash-plus programmes.

Fig 1

PRISMA flow chart.

Study characteristics

The majority of studies focus on the first 1,000 days of a child’s life and cite the critical importance of intervening during this period to maximise development potential. The studies included in this review cover countries in sub-Saharan Africa (7), Latin America (6), and South and Southeast Asia (4). CCTs (6 studies), UCTs (7 studies), and public works programmes (4 studies) were studied; all CCTs were implemented in Latin America, where that type of transfer is prominent. In addition to the cash transfer, 5 categories of “plus” components were identified: nutrition BCC (n = 7), food transfers (n = 3), primary healthcare (n = 2), psychosocial stimulation (n = 7), and child protection interventions (n = 4).

There are 2 dominant theories about how these cash-plus programmes work: (1) The package works by improving mediating outcomes on the pathway to impact (e.g., maternal knowledge of proper nutrition) while also addressing the structural deprivations impacting health and development potential [32]; and (2) Supply-side and demand-side interventions must occur in tandem to meet population needs (e.g., providing health services when health checks are a cash condition) [33,34].

Of the 9 studies that reported cash values, standardised transfer amounts (converted from local currency to US dollars at the time of intervention) ranged from approximately $8/month to $75/month and were often distributed monthly. The plus-components were delivered either monthly or weekly. In providing cash value justifications, studies noted either matching values from other cash transfer initiatives in the country or setting the value based on the country’s poverty level and cost of necessary household expenses. When calculating the latter value, interventions covered up to 20% of household expenses [35].

Three out of the 4 intervention designs described by Arriagada and colleagues [30] were identified in the included studies. Six programmes (8 studies) utilised an integrated design, specifically that the plus-component was nested within, and operated by, the cash transfer programme. A further 4 programmes (7 studies) had a convergence design, whereby the 2 components were implemented separately but there was explicit coordination between implementing partners. This contrasts to one programme (2 studies) that used an alignment design, in which there was separate implementation of the 2 components and no coordination between implementers (regardless of whether the same population was reached). No studies were identified that utilised a piggybacking design, which would have the cash component delivered within an already-existing plus intervention.

Only one programme [36-38] was designed for active father involvement throughout the intervention period. Three other programmes [32,39-42] had a portion of the plus component dedicated to father education, of which one programme [40,41] also invited mothers-in-law to participate in hopes of creating a supportive household environment. Further, only 4 programmes exclusively used a home visit model [36-38,43-45].

All studies were done as cluster-randomised controlled trials (cRCTs) except for one randomised at the individual level [44] and 3 using a quasi-experimental design [33,34,46]. Studies were published from 2013 to 2021 and had follow-up times ranging from 4 months to 4 years post-intervention. Details of individual interventions are shown in Table 1 below, and further information is available in S1 Table. Table 2 provides an overview map of the review findings based on the intervention category and outcome.

Table 1

Intervention components and outcomes.

Study	CT and Intervention Design	Nutrition BCC	Food Transfer	Primary Healthcare	Psychosocial Stimulation	Child Protection
Langendorf 2014	Unconditional CT (Integrated)		Anthropometrics, mortality
Field 2021	Unconditional CT (Convergence)	Anthropometrics, child illness
Khan 2019	Unconditional CT (Convergence)	Anthropometrics	Anthropometrics
Guyatt 2018	Unconditional CT (Integrated)	Anthropometrics
Ahmed 2019	Unconditional CT (Integrated)	Anthropometrics, child illness	Anthropometrics, child illness
Ahmed 2020	Unconditional CT (Integrated)	Poverty
UNICEF 2020	Cash-for-Work (Integrated)	Anthropometrics
Premand 2020	Unconditional CT (Integrated)	Anthropometrics			Child development	Violence against children; positive parenting; child illness
Barnhart 2020	Cash-for-Work (Convergence)				Child development	Violence against children; positive parenting
Betancourt 2020	Cash-for-Work (Convergence)					Violence against children; positive parenting; child illness
Jensen 2021	Cash-for-Work (Convergence)				Child development	Violence against children
Attanasio 2014	Conditional CT (Integrated)				Child development, anthropometrics
Andrew 2018	Conditional CT (Integrated)				Child development
Fernald 2016	Conditional CT (Convergence)				Child development
Kagawa 2017	Conditional CT (Convergence)				Child development
Guanais 2013	Conditional CT (Alignment)			Post-Neonatal Infant Mortality
da Silva 2019	Conditional CT (Alignment)			Child Mortality

BCC, behaviour change communication; CT, cash transfer.

Table 2

Overview map of findings from meta-analyses and synthesis.

Intervention	Outcome	Cash vs Cash-Plus
Cash + Nutrition BCC
	Stunting	No difference (6)
	Wasting	No difference (6)
	Underweight	No difference (6)
	Fever	Mixed findings (2)
	Diarrhoea	No difference (2)
	Cough/Cold	Mixed findings (2)
	General Child Illness	Cash-Plus more effective (1)
	Poverty	Cash-Plus more effective (1)
Cash + Food Transfer
	Wasting (short-term crisis)	Cash-Plus more effective (1)
	Stunting (long-term impact)	Cash-Plus more effective (2)
	Wasting (long-term impact)	No difference (2)
	Underweight (long-term impact)	No difference (2)
	Child Illness	No difference (1)
	Mortality	Cash-Plus more effective (1)
Cash + Primary Healthcare
	Mortality	Cash-Plus more effective (2)
Cash + Psychosocial Stimulation
	Cognitive Development	No difference (3)
	Overall Child Development	Cash-Plus more effective (1)
Cash + Child Protection
	Use of Harsh Discipline	No difference (3)
	Exclusive Positive Parenting	No difference (3)
	Child Illness	Mixed findings (3)

Meta-analysis findings are bolded.

BCC, behaviour change communication.

Risk of bias of individual studies

Risk of bias was low for most studies, emphasising a high-quality evidence base and strong methodological rigour. One study was rated as having some concerns in risk of bias because of deviations from the planned intervention, specifically in implementation challenges that led to difficulties in individuals accessing cash [44]. All randomised trials followed children longitudinally throughout the trial period. The 3 quasi-experimental studies were rated as having moderate risk of bias because although the studies account for confounding, there is still greater risk of bias than if the studies could have used a randomised design. Similarly, these 3 studies relied on repeat cross-sectional data, which introduce some bias in potentially uneven exposure to the intervention by assessing within-population change rather than within-person change. No study was rated as having high risk of bias. Individual study assessments are available in S3 Text.

Synthesis of results

Cash + Nutrition Behaviour Change Communication (BCC)

Seven studies were identified that combined cash transfers with nutrition BCC [32,40,41,44-47]. Most BCC components covered UNICEF’s Essential Family Practices on maternal and child nutrition, including exclusive breastfeeding (age <6 months) and complementary feeding (age >6 months), health and hygiene practices, and use of health services preventively and promptly when a child becomes ill [47]. Three anthropometric indicators were assessed (height-for-age, weight-for-height, and weight-for-age) as well as the odds of children experiencing stunting, wasting, or being underweight, respectively, which is defined as having z-scores ≤−2 standard deviations of the WHO Child Growth Standards Median [48].

Meta-analysis findings (Fig 2) suggest that cash-plus programmes are not more effective than cash transfers alone in reducing odds of children experiencing stunting (OR = 0.95 (95% CI 0.83, 1.09), p = 0.40; I2 = 26% (95% CI 0, 87)), wasting (OR = 0.99 (0.93, 1.05), p = 0.64; I2 = 0% (0, 14)), or underweight status (OR = 1.01 (0.84, 1.20), p = 0.93; I2 = 17% (0, 98)). Additional meta-analysis findings (Fig 3) also suggest no added impact on z-score measures of anthropometrics for stunting (d = 0.03 (−0.04, 0.09), p = 0.36, I2 = 24% (0, 88)), wasting (d = −0.03 (−0.12, 0.07), p = 0.51, I2 = 47% (0, 93)), or being underweight (d = −0.03, (−0.11, 0.05), p = 0.45, I2 = 31% (0, 92)). Sensitivity analyses were also conducted by removing the quasi-experimental study [46] from the meta-analyses. No sensitivity test showed a significant z-score change for height-for-age (d = 0.04 (−0.03, 0.11), p = 0.18), weight-for-height (d = −0.03 (−0.15, 0.09), p = 0.57), or weight-for-age (d = −0.02 (−0.13, 0.08), p = 0.61).

Fig 2

Forest plot of Cash + Nutrition BCC vs. cash alone on anthropometric odds ratios.

Fig 3

Forest plot of Cash + Nutrition BCC vs. cash alone on anthropometric z-scores.

One study in Bangladesh found no gendered differences in anthropometric impacts [40], while another in Niger found that impact on wasting was driven by improvements in boys [47]. Compared to cash alone, the intervention effects of the cash-plus intervention in Kenya for wasting/underweight reduction were greater in smaller household sizes and greater for wealthier households in reducing stunting [44]. Another study in Myanmar found the cash-plus intervention to be significant in decreasing moderate stunting compared to cash alone, with no impact on severe stunting [32].

There is also evidence that Cash + Nutrition BCC may reduce poverty. At the end of the intervention in Bangladesh [41], researchers found that cash-plus had greater reduced poverty head count (20% lower) and depth (6% lower) and severity (3% lower) of poverty compared to cash alone. While these differences were no longer statistically significant 4 years post-programme, there is other evidence of sustained reductions in poverty. First, while many families moved out of poverty during the programme, approximately 40% in both cash-only and cash-plus groups fell back into poverty by the 4-year follow-up. However, families in the cash-plus arm experienced greater movement out of poverty during the intervention period than those receiving cash alone; thus at 4 years post-evaluation, there was a 10-percentage point difference between cash-plus and cash-only groups in the proportion of families that had exited poverty and stayed nonpoor. The authors theorise that this was contributable to the evidence that the plus-component further enhanced women’s social capital and agency. Second, 4 years after the intervention ended, there was an added 16% reduction in chronic monetary poverty from cash-plus compared to cash alone, as measured by the Calvo-Dercon Poverty Score [49].

Multiple indicators along the causal pathway between cash and child anthropometrics showed improvement of cash-plus over cash alone. The most notable is improvements in WASH practices (5/5), including improved defecation (2/2), handwashing (4/5), and treating water (1/1) [32,40,44,45,47]. One study also found significant reductions in iron deficiency anaemia (1/1) [45].

However, several other indicators demonstrated no added benefit of cash-plus. This includes in increasing birth registration (0/2) [46,47], complete immunisation (0/2) [44,47], or in reducing vitamin A deficiency (0/2) [45,47] or diarrhoea (0/2) [32,40].

Markers of food security were mixed, specifically for increased food consumption (2/2) [32,40], improved dietary diversity scores or minimally acceptable diet (3/5) [32,40,44,46,47], or increasing number of feedings (0/1) [40]. In contrast, one additional study found that cash-only had better outcomes than the cash-plus group on dietary diversity, iron food consumption, minimum meal frequency, and minimum acceptable diet [45]. There were also mixed findings on increasing ever or early breastfeeding (2/4) [32,44,46,47], increasing women’s decision-making agency (1/2) [32,41], increasing medical care seeking for a sick child (1/2) [32,47] or reducing fever and cough/cold (1/2) or general child illness (1/1) [32,40,47].

Cash + food transfer

Three studies aimed to reduce or prevent undernutrition [40,45,50], all by supplementing cash with additional food transfers. Studies were conducted in impoverished regions with high rates of food insecurity and undernutrition, where approximately half of all study children had stunted growth [40,45,50].

One study was implemented during an acute crisis in the hunger gap in Niger when rates of undernutrition, diarrhoeal diseases, and malaria increase [50]. The trial compared cash alone to cash-plus 1 of 3 different food transfer formulations. The 3 trial arms were meta-analysed to calculate an overall impact of the 3 cash-plus arms in preventing global acute malnutrition. Compared to cash alone, children in households receiving Cash + Food had significantly reduced odds of experiencing acute malnutrition (OR = 0.41 (0.27, 0.63), p = 0.012). The cash-plus groups had up to 8 times lower mortality incidence rates than cash alone. When children died, the leading suspected cause was malaria (76%) and gastroenteritis (14%).

Meta-analysis results of studies in long-term development contexts (Figs 4 and 5) suggest that cash-plus is more effective over cash alone in increasing height-for-age z-score (d = 0.08 (0.03, 0.14), p = 0.02, I2 = 0% (0, 74)), which translated to significantly reduced odds of children experiencing stunting (OR = 0.82 (0.74, 0.92), p = 0.01, I2 = 0% (0, 70)). There was no added impact in improving weight-for-height z-score (d = −0.13 (−0.42, 0.16), p = 0.24, I2 = 87% (57, 99)) or weight-for-age z-score (d = −0.06 (−0.28, 0.15), p = 0.43, I2 = 76% (24, 98)). Similarly, there was no added impact in reducing odds of children experiencing wasting (OR = 0.89 (0.70, 1.14), p = 0.24, I2 = 0% (0, 86)) or underweight status (OR = 0.93, (0.80, 1.09), p = 0.26, I2 = 0% (0, 84)), respectively.

Fig 4

Forest plot of Cash + Food Transfers vs. cash alone on anthropometric z-scores.

Fig 5

Forest plot of Cash + Food Transfers vs. cash alone on anthropometric odds ratios.

The analysis included 2 studies, one in Pakistan with 2 separate trial arms (both receiving Cash + Food, but one also receiving BCC) [45] and one in Bangladesh, which had 2 separate trials (one each in the north-western and southern regions of the country) [40]. To note, this latter study set the cash-plus value to equal the cash-only value (i.e., the cash-plus group received half the cash value of the cash-only transfer) as opposed to providing the same cash value to both groups, but also supplementing with food transfers. There was no difference between cash-only and Cash + Food in reducing fever, cough/cold, or diarrhoea in either the North or South trial in Bangladesh [40].

Cash + primary healthcare

Only 2 studies were identified that evaluated provision of cash and primary healthcare services (e.g., growth monitoring, preventive health services, or medical treatment for illness). Both were quasi-experimental studies that examined Brazil’s independent scale-ups of the CCT, Bolsa Família Program (BFP), and primary healthcare, Family Health Program (FHP), which provided care from community health workers to doctors. Although the programmes were each scaled up separately (an alignment design), the findings from both studies demonstrate the interdependence of the 2 programmes in achieving success in reducing post-neonatal infant [33] and child [34] mortality rates (both measured as deaths per 1,000 live births). Both study equations showed that BFP and FHP each had statistically significant, independent effects in reducing mortality rates. Guanais [33] also included a statistically significant interaction term for the 2 programmes, indicating that there is also an interdependence of the 2 programmes in further reducing infant mortality. For example, in a municipality with coverage rates for BFP and FHP at 25% and 0%, respectively, the predicted mortality rate was 5.24 deaths/1,000 live births as opposed to coverage of 60% and 100% having a rate of 1.38 death/1,000 live births [33]. Combining cash and primary healthcare had the greatest impact in the highest poverty regions, and this is especially important given the high inequality rate in Brazil [33]. The models indicate that the impact of FHP is higher at higher coverage levels of BFP, thus highlighting the need to scale up the demand-side and supply-side interventions in concert [33].

Cash + psychosocial stimulation

Seven studies, representing 4 unique programmes, evaluate the impacts of cash-plus programming on measures of child development. One programme used group-based social and BCC methods of which child psychosocial development is one of many topics covered [47], one used group-based parenting support [39,42], and 2 programmes exclusively used a home visit model [36,38,43,51].

Studies employed a variety of child development measures, of which 3 included specific subscales or scale measures of cognitive development. After standardising the measures, meta-analysis findings (Fig 6) suggest that these Cash + Psychosocial Stimulation programmes may not be more effective than cash transfers alone in promoting overall cognitive development (d = 0.16 (−0.25, 0.57), p = 0.24, I2 = 85% (47, 100)), although there is substantial heterogeneity among the studies.

Fig 6

Forest plot of Cash + Psychosocial Stimulation vs. cash alone on cognitive development.

Although a pilot study was underpowered to detect improvements in child development [36], the main trial in Rwanda found significant improvements in overall child development from cash-plus over cash alone (d = 0.21 SD (0.09, 0.33)) [38]. The trial in Niger also found added improvements in socioemotional development (using the strength-and-difficulties questionnaire) for children receiving cash-plus over cash alone (d = 0.149 SD (0.03, 0.27)) [47]. One study also tested the addition of micronutrient supplementation, but found no added benefit for cognitive development or child growth [43].

Even for interventions that are found to be effective, however, caution must be raised for the longer-term effect of the intervention package. While one cash-plus programme showed positive effects above cash alone in the short-term evaluation (d = 0.26 SD improvement in cognitive scores in Colombia) [43], the effects dissipated within 2 years after the intervention [51]. This likewise applies to anticipated mediators of child development. Three out of the 4 programmes found significant improvements across the variety of play materials and variety and frequency of play activities [36,37,43,47,52]. However, in the same 2-year post-intervention follow-up study in Colombia, the intervention impacts on these mediators were no longer statistically significant [51].

Comparing different socioeconomic subgroups in cash-plus arms to cash alone, these intervention packages varied in effects. While there was evidence of higher cognitive impact for Mexican indigenous communities in the stratified sample and among children with lowest scores at baseline, the intervention effects also favoured wealthier households [39]. Two studies in Colombia and Mexico found the interventions favoured mothers with greater education [39,51].

Only one study in our review was identified that evaluated 2 different design models against a cash-only control [39], which found that only Mexican children in the convergence intervention design had improved cognitive development and not in the alignment design. Secondary analysis was also done to assess if impacts differed when women became pregnant [42]; the study found that only mothers in the convergence intervention design who became pregnant after adolescence had improved child development.

Cash + child protection

Four studies of 2 unique programmes [36-38,47] were designed using a comprehensive early child development framework, with child protection as one of the main pillars. Both programmes used home visits, but Premand and Barry [47] also had small-group and village assembly components in Niger. A pilot evaluation [36] notes that the larger effectiveness trial [37,38] also included “booster sessions” in efforts to prolong and sustain intervention effects in Rwanda. Studies were done in high poverty and food insecure contexts, and one study found very high levels (almost 50% prevalence) of violent discipline to children and one-third of mothers had experienced violence in the last 3 months [37]. Violent discipline was measured by the prevalence of any harsh discipline from parents to child (e.g. forbidding, shaking, yelling, spanking with/without an object, berating, slapping, and hitting hands with/without an object) and exclusive positive parenting measured by the absence of these discipline methods and, instead, disciplining through explaining or redirecting the child [47].

Although both main programme trials found reductions in violent discipline, meta-analysis findings suggest that Cash + Child Protection is not more effective than cash alone in reducing parental violent discipline of children (OR = 0.83 (0.59, 1.17), p = 0.15, I2 = 40% (0, 99)) or increasing exclusive use of positive parenting (OR = 1.02 (0.82, 1.28), p = 0.69, I2 = 17% (0, 96)) (Fig 7). One study found that the decline in parental violent discipline was driven by reductions in subscales on the parental use of shaking, spanking, berating, slapping, and hand-hitting, but there was no impact on the use of yelling, forbidding, or hitting their child with objects [47].

Fig 7

Forest plot of Cash + Child Protection vs. cash alone on parenting practices.

The programme in Rwanda found that in addition to reductions in violent discipline of children, there was added benefit of cash-plus above cash alone in increasing father engagement in childcare, improving shared decision-making, and reducing maternal depression and female violence victimisation, but no added benefit in reducing male violence perpetration [36-38]. The intervention effects on reduced violent discipline were maintained 1 year post-intervention (incidence rate ratio (IRR) = 0.74 (0.66, 0.84), p < 0.001) [38]. There was no difference in impact for boys and girls.

These comprehensive early child development programmes also evaluated the impact on childhood illness. Despite an almost 2 times higher odds of accessing clean water in the cash-plus group compared to cash-only group at endline, Betancourt and colleagues [37] found no additional impact from the cash-plus intervention on childhood illness, including diarrhoea, fever, and cough, though this may be due to the change from rainy (baseline) to dry (endline) seasons in Rwanda; the pilot study had only found reductions in cough prevalence compared to cash alone [36]. Premand and Barry [47], however, found a 19% reduction in the cash-plus group compared to cash alone in the child being sick in the past month in Niger.

Discussion

Summary of the evidence

This systematic review and meta-analysis investigated whether cash-plus programmes were more effective than cash alone in accelerating outcome improvements in young children. In addition to the evidence synthesis, this review contributes to the growing literature of situating cash transfers as the “control” or standard of care. This is particularly relevant to studies assessing the cost-effectiveness of new interventions and their value compared to if a low-cost cash transfer programme were implemented as opposed to nothing [53,54].

The review found cash-plus programmes with 5 types of plus-components that are focused on young children. These cash-plus programmes linked cash transfers for poverty alleviation (SDG 1) with SDG 2 (Nutrition BCC, Food Transfers), SDG 3 (Primary Healthcare), SDG4 (Psychosocial Stimulation), and SDG 16 (Child Protection). These intervention packages also touch all 5 areas of the multisectoral Nurturing Care Framework for Early Child Development, namely by integrating cash (Security and safety) with efforts to drive (1) Adequate nutrition; (2) Good health; (3) Opportunities for early learning; and (4) Responsive caregiving [8]. Given the large focus in global health on psychosocial stimulation to improve child development and BCC to address undernutrition, it is unsurprising that these 2 interventions represented the majority studies. Nonetheless, each of the intervention categories was limited in the total number of studies identified.

Meta-analysis results concluded that Cash + Food Transfers may be more effective than cash alone in reducing stunting. However, meta-analysis findings suggest no added impact above cash alone in Cash + Nutrition BCC for improving anthropometrics, Cash + Psychosocial Stimulation for improving cognitive development, or Cash + Child Protection for reducing violent discipline or increasing exclusive positive parenting. However, there was potentially substantial heterogeneity across the meta-analyses. The narrative synthesis found preliminary evidence that Cash + Primary Healthcare may have greater impacts than cash alone in reducing mortality, Cash + Food Transfers may have greater impacts than cash alone in reducing acute malnutrition and mortality in crisis contexts, Cash + Nutrition BCC may have greater impacts than cash alone in reducing poverty, and Cash + Child Protection trials suggest a trend towards greater impact than cash alone in reducing violent discipline.

Overall, the studies included in this review had low risk of bias, and while the studies were rigorously conducted, there are still a very limited number of studies examining cash-plus versus cash alone, and therefore, the ability to generalise is limited until more studies are published. This review found that cash-plus programmes are being employed throughout LMICs and more than two-thirds of the included studies were published in the last 3 years. As this is an emerging area in social protection, there is hope that more research will become available in the coming years that evaluate the interventions against cash controls, some of which have been published as protocols (see S2 Text). This review is timely and especially relevant as new cash-plus programmes continue to be planned and scaled up [55].

A previous review of cash transfer programmes found that the evidence of cash impact decreases with each step further in the causal pathway between cash input and changes in child health and well-being [56]. To address the limitations in impact of cash transfers, cash-plus programmes were developed in hopes of reaching and improving these distal, “third-order” outcomes (child health and development), which our review investigated. Noting the limited evidence of impact on these outcomes, we also examined possible pathway contributions of cash-plus over cash alone within these studies. Assessment of causal pathways suggests that cash-plus likewise has little added benefit over cash alone, except for a few noteworthy differences such as in Cash + Nutrition BCC improving WASH practices and reducing poverty, Cash + Psychosocial Stimulation improving child stimulation practices and materials, and Cash + Child Protection reducing intimate partner violence. However, these results are in context of studies that evaluated the distal outcomes, thus there may be selection bias in these intermediary impact findings.

A recent meta-analysis of cash transfers (including cash transfers both with and without Nutrition BCC) found very small, but statistically significant improvements on stunting height-for-age z-scores (0.03+/−0.03 SD) compared to a no-intervention control, which accounts for a 2.1% reduction in stunting prevalence [4]. Among studies delivering Cash + Nutrition BCC, there was a statistically significant 3.1% reduction in stunting prevalence compared to a no-intervention control, but no impact on height-for-age z-score. Our review adds to the evidence base by specifically comparing Cash + Nutrition BCC to cash alone, finding that there was no statistically significant difference between the 2 interventions in reducing stunting; this may suggest that any improvements in stunting reduction are likely driven by the cash component rather than the added Nutrition BCC. In both our review and the review of cash transfers alone, there were very few studies that assessed child illness and more research is needed to assess the impact on this outcome and its contribution to the causal pathway between cash and child growth.

Lastly, there are limited and mixed impacts on vulnerable subgroups, having potential unintended consequences. Only about half of the included studies provided any equity evidence (defined as subgroup or interaction effects), of which only 3 studies conducted sex-disaggregated analyses; findings demonstrated either no difference in intervention effects between boys and girls (2) or the intervention effects favoured boys only (1). Few studies specifically involved fathers, which may place increased caregiving burden on mothers, such as collecting food/cash transfers, attending intervention sessions, or complying with conditions of cash transfers; this has also been identified as an issue in graduation programmes for early child development [57]. More work is needed to investigate and ensure that these interventions benefit the most vulnerable children, thereby narrowing the health equity gap (SDG 10).

Limitations

The main limitation of this review is the few study numbers that were retrieved for each cash-plus programme. Findings from this review should be viewed as preliminary evidence and serve as a guide for future research, particularly given the potentially high heterogeneity in the meta-analyses. Ultimately, more research is needed before definitive conclusions can be made on whether cash-plus is more effective than cash alone in improving outcomes for young children.

Because this is a review comparing cash-plus against cash alone, trials were not included that had only 2 arms (a cash-plus group and a no-intervention group), some of which are found in S2 Text. Thus, this systematic review retrieved a subset of all studies on cash-plus programmes, and notable programmes, such as Ghana’s Cash + Health Insurance [58,59], were not included due to comparison criteria requirements. However, the goal of this review was to assess whether cash-plus programmes are more effective than cash alone rather than compared to no intervention. In addition to the cash-plus programmes included in this review, there are other combination interventions that may have an impact on children; examples include productivity and livelihood interventions that are targeted to households or adults, but have a benefit to improving child outcomes [60], or broader environmental health interventions such as water and sanitation management [61].

Some studies noted inherent difficulties in measurement (e.g., measuring change in standardised z-scores for anthropometrics) and that improvements in outcomes (e.g., stunting) may take longer than the intervention/study duration [44,50]. This review also comments on the impact of cash-plus on the causal pathways to the outcomes of interest. While the discussion of the intermediary outcomes provides context for the impact of cash-plus programmes compared to cash alone, there is a risk of selection bias in only discussing the findings from studies that also examined the distal outcomes and not including studies that only evaluated the impact of cash-plus on the intermediary outcomes.

Due to language abilities, this study only examined English studies; this may have also introduced selection bias through missing articles published through non-English outlets [62]. However, where known national programmes were operating or being planned, experts working on these programmes were contacted to identify potentially missed publications. Only one study was identified as being omitted by the search [63]. The cognitive impacts of this Cash + Psychosocial Stimulation intervention were published in English literature and included in this review [39]. However, the report also included nutritional impacts that were not reported in English.

Recommendations for research

Our review identified a variety of study designs, which impact the conclusions that the individual studies can make on the effectiveness of a cash-plus programme. First, several studies used a cash control when large-scale cash transfer programmes were already operating in-country. Second, other studies used a no-intervention control and had cash-only and cash-plus arms; only a subset of these studies included additional analyses using the cash-only arm as the control to assess for the added impact of the plus-intervention above cash alone. Lastly, 2 studies used a natural experiment to assess the individual component impacts and their interaction. Although introducing more bias because of nonrandom intervention assignment, this design allowed for assessment of cash-plus at scale and its impact on population-level child outcomes.

An ideal study design would be a longitudinal, cRCT with 4 arms (control, cash-only, plus-only, and cash-plus) and which includes analyses using no-intervention as the control as well as each of the component parts. This would provide a complete picture of the impact of cash-plus programmes and the potential synergies of cash and plus components above and beyond either component alone. Only 3 programmes included in this review utilised a study design that would allow for most of these conclusions [40,41,47,50]. However, whether a no-intervention control is possible is impacted by whether a national cash transfer programme already exists and similarly, if no intervention would be unethical, such as in a crisis context.

There were high levels of heterogeneity in the meta-analyses and limited study numbers, thus more studies are needed before overall effectiveness can be determined. Promising areas of investigation include combining cash with food transfers, primary healthcare, and parenting interventions for child protection. Although Cash + Primary Healthcare was found to be effective in Brazil, there was a clear link between the 2 interventions (i.e., cash was conditioned on use of preventive health services). It is unclear if connecting access to health services and poverty alleviation interventions would be as effective if the transfer was unconditional. Other specific evidence gaps include no evaluations having been done in the Middle East and North Africa and East Asia and Pacific regions.

Two other essential areas for future investigation include (1) assessing for long-term effects of the intervention packages and (2) whether the effects are maintained when brought to scale. One study found some indicators of maintained poverty reductions 4 years post-programme [41], while another study noted that the lack of long-term impact on cognitive development was possibly due to implementation challenges at scale as opposed to a controlled and smaller efficacy trial [51].

Lastly, the heterogenous findings of studies included in this review emphasise the need to further investigate the role of implementation and context, cash-plus models, and selection of the plus-component in optimising the impact of cash-plus programmes. For example, only one programme was identified in the review that evaluated multiple cash-plus models, finding differential impacts based on design [39,42]. Only 3 programmes were identified that tested different plus-components in addressing the same child outcomes, each demonstrating that not every relevant plus-component will further improve outcomes for children [40,41,43,45]. Only one study was found to explicitly consider context (i.e., functioning or accessible food markets) in selecting the intervention package [40]. No included studies tested different intensities of the same plus-component. Optimising programme design, selection of the plus-component, and its implementation and intensity—in addition to the overall evidence of effectiveness (the scope of this review)—will be essential in improving outcomes for the most vulnerable children and maximising the cost-effectiveness of these programmes.

Conclusions

Cash transfers alone may not achieve the outcome improvements in children that practitioners and policymakers aim to address, and, in these cases, cash-plus programming may be considered. While debates remain on the optimal package of interventions, this review provides preliminary evidence that the added plus-component in cash-plus may be more effective than cash alone when combining cash with food transfers to prevent acute malnutrition in crises and to reduce stunting in the long-term, primary healthcare to reduce mortality, nutrition BCC to reduce long-term poverty, and child protection interventions to reduce violent discipline. Findings also suggest that cash-plus is not more effective than cash alone when combined with nutrition BCC to improve child anthropometrics. Ultimately, more research is needed on how to optimise the promise of cash transfers and multisectoral intervention packages for young children to achieve the SDGs by 2030.

PRISMA Checklist.

PRISMA, Preferred Reporting Items for Systematic Reviews and Meta-Analyses.

(DOC)

Click here for additional data file.

Funnel plots.

Funnel plots provided for each meta-analysis.

(TIF)

Click here for additional data file.

Study characteristics.

Information provided on country, participants, follow-up period, study design, cash amount, intervention intensity, intervention provider, and plus-intervention descriptions for each study.

(DOCX)

Click here for additional data file.

Information sources and sample search strategy.

Sources of information searched and sample search strategy are provided.

(DOCX)

Click here for additional data file.

Excluded studies at full text and list of included studies.

Table of excluded studies at full-text screening, with explanation, and a list of the studies included in the review.

(DOCX)

Click here for additional data file.

Risk of bias assessments.

Individual study assessments for risk of bias using Cochrane Risk of Bias and ROBINS-I tools.

(DOCX)

Click here for additional data file.

Study protocol.

Study protocol as published on PROSPERO.

(PDF)

Click here for additional data file.

16 Feb 2021

Dear Dr Little,

Thank you for submitting your manuscript entitled "Do cash-plus programmes improve early childhood outcomes more than cash transfers or child development interventions alone? A systematic review and meta-analysis in low- and middle-income countries." for consideration in PLOS Medicine’s Special Issue on Global Child Health.

Your manuscript has now been evaluated by the PLOS Medicine editorial staff as well as by the Special Issue Guest Editors and I am writing to let you know that we would like to send your submission out for external peer review.

However, before we can send your manuscript to reviewers, we need you to complete your submission by providing the metadata that is required for full assessment. To this end, please login to Editorial Manager where you will find the paper in the 'Submissions Needing Revisions' folder on your homepage. Please click 'Revise Submission' from the Action Links and complete all additional questions in the submission questionnaire.

Please re-submit your manuscript within two working days, i.e. by .

Login to Editorial Manager here: https://www.editorialmanager.com/pmedicine

Once your full submission is complete, your paper will undergo a series of checks in preparation for peer review. Once your manuscript has passed all checks it will be sent out for review.

Feel free to email us at plosmedicine@plos.org if you have any queries relating to your submission.

Kind regards,

Caitlin Moyer, Ph.D.

Associate Editor

PLOS Medicine

1 Apr 2021

Dear Dr. Little,

Thank you very much for submitting your manuscript "Do cash-plus programmes improve early childhood outcomes more than cash transfers or child development interventions alone? A systematic review and meta-analysis in low- and middle-income countries." (PMEDICINE-D-21-00432R1) for consideration in PLOS Medicine’s Special Issue on Global Child Health.

Your paper was evaluated by a senior editor and discussed among all the editors here. It was also discussed with the special issue Guest Editors, and sent to three independent reviewers, including a statistical reviewer. The reviews are appended at the bottom of this email and any accompanying reviewer attachments can be seen via the link below:

[LINK]

In light of these reviews, I am afraid that we will not be able to accept the manuscript for publication in the journal in its current form, but we would like to consider a revised version that addresses the reviewers' and editors' comments. Obviously we cannot make any decision about publication until we have seen the revised manuscript and your response, and we plan to seek re-review by one or more of the reviewers.

In revising the manuscript for further consideration, your revisions should address the specific points made by each reviewer and the editors. Please also check the guidelines for revised papers at http://journals.plos.org/plosmedicine/s/revising-your-manuscript for any that apply to your paper. In your rebuttal letter you should indicate your response to the reviewers' and editors' comments, the changes you have made in the manuscript, and include either an excerpt of the revised text or the location (eg: page and line number) where each change can be found. Please submit a clean version of the paper as the main article file; a version with changes marked should be uploaded as a marked up manuscript.

In addition, we request that you upload any figures associated with your paper as individual TIF or EPS files with 300dpi resolution at resubmission; please read our figure guidelines for more information on our requirements: http://journals.plos.org/plosmedicine/s/figures. While revising your submission, please upload your figure files to the PACE digital diagnostic tool, https://pacev2.apexcovantage.com/. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email us at PLOSMedicine@plos.org.

We expect to receive your revised manuscript by . Please email us (plosmedicine@plos.org) if you have any questions or concerns.

***Please note while forming your response, if your article is accepted, you may have the opportunity to make the peer review history publicly available. The record will include editor decision letters (with reviews) and your responses to reviewer comments. If eligible, we will contact you to opt in or out.***

We ask every co-author listed on the manuscript to fill in a contributing author statement, making sure to declare all competing interests. If any of the co-authors have not filled in the statement, we will remind them to do so when the paper is revised. If all statements are not completed in a timely fashion this could hold up the re-review process. If new competing interests are declared later in the revision process, this may also hold up the submission. Should there be a problem getting one of your co-authors to fill in a statement we will be in contact. YOU MUST NOT ADD OR REMOVE AUTHORS UNLESS YOU HAVE ALERTED THE EDITOR HANDLING THE MANUSCRIPT TO THE CHANGE AND THEY SPECIFICALLY HAVE AGREED TO IT. You can see our competing interests policy here: http://journals.plos.org/plosmedicine/s/competing-interests.

Please use the following link to submit the revised manuscript:

https://www.editorialmanager.com/pmedicine/

Your article can be found in the "Submissions Needing Revision" folder.

To enhance the reproducibility of your results, we recommend that you deposit your laboratory protocols in protocols.io, where a protocol can be assigned its own identifier (DOI) such that it can be cited independently in the future. Additionally, PLOS ONE offers an option to publish peer-reviewed clinical study protocols. Read more information on sharing protocols at https://plos.org/protocols?utm_medium=editorial-email&utm_source=authorletters&utm_campaign=protocols

Please ensure that the paper adheres to the PLOS Data Availability Policy (see http://journals.plos.org/plosmedicine/s/data-availability), which requires that all data underlying the study's findings be provided in a repository or as Supporting Information. For data residing with a third party, authors are required to provide instructions with contact information for obtaining the data. PLOS journals do not allow statements supported by "data not shown" or "unpublished results." For such statements, authors must provide supporting data or cite public sources that include it.

We look forward to receiving your revised manuscript.

Sincerely,

Caitlin Moyer, Ph.D.

Associate Editor

PLOS Medicine

plosmedicine.org

-----------------------------------------------------------

Requests from the editors:

1. Please revise your title according to PLOS Medicine's style. Your title must be nondeclarative and not a question. It should begin with main concept if possible. "Effect of" should be used only if causality can be inferred, i.e., for an RCT. Please place the study design ("A randomized controlled trial," "A retrospective study," "A modelling study," etc.) in the subtitle (ie, after a colon).

2. Abstract: Please structure your abstract using the PLOS Medicine headings (Background, Methods and Findings, Conclusions; Please combine the Methods and Findings sections into one section, “Methods and findings”)

3. Abstract: Please provide the dates of search, types of study designs included, eligibility criteria including language of publication. Please briefly mention the countries represented in the studies, as well as some information on types of included studies, participants, and relevant characteristics of studies.

4. Abstract: Please remove italics used for emphasis.

5. Abstract: As the final sentence of the Methods and Findings section, please describe the main limitation(s) of the study's methodology.

6. Abstract: Conclusions: In the second sentence, we suggest phrasing this as the overall finding of your study, for example: “In this study, we observed that few package combinations showed evidence of effectiveness.” or similar.

7. Author Summary: At this stage, we ask that you include a short, non-technical Author Summary of your research to make findings accessible to a wide audience that includes both scientists and non-scientists. The Author Summary should immediately follow the Abstract in your revised manuscript. This text is subject to editorial change and should be distinct from the scientific abstract. Please see our author guidelines for more information: https://journals.plos.org/plosmedicine/s/revising-your-manuscript#loc-author-summary

8. Throughout: Please use numbers within brackets for in-text reference citations, like this [1]. Please avoid the use of italics for emphasis.

9. Introduction: Page 5, Lines 9-10: Please temper this with “... to the best of our knowledge,to date there has been no synthesis to evaluate…” or similar.

10. Methods: Page 5: Line 24: Please include the sentence "This study is reported as per the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guideline (S1_Checklist)."

11. Methods: Page 6, Lines 5-6: Please update your search to the present time.

12. Methods: Page 6-7: It would be preferable to format the description of eligibility criteria for included studies in paragraph form, rather than as an outline/list.

13. Methods: page 7, Lines 8-9: Please ensure that all abbreviations are fully spelled out at their first appearance in the main text (for example, UCTs, CCTs)

14. Methods: Page 8: Please comment on assessment of publication bias.

15. Methods: Page 9, lines 9-10: Please clarify the description of correction for clustering- “The analysis corrected for clustering to address possible unit of analysis error”

16. Methods: Page 10, line 6-7: “Eight protocols were identified that could meet the criteria for

7 inclusion in an update to the review.” Was an attempt made to contact the authors to determine if the study findings have since been published?

17. Results: Page 15, Line 7-8: Please present separate the results for the randomized vs. non-randomized studies, if possible.

18. Results: Throughout (for example page 23, line 19) please give the exact p value rather than p<0.05 for the meta analysis results (however please report p<0.001 when appropriate).

19. References: Please use the "Vancouver" style for reference formatting, and see our website for other reference guidelines https://journals.plos.org/plosmedicine/s/submission-guidelines#loc-references

20. PRISMA Checklist: Thank you for including the PRISMA checklist. Please update the checklist, using section and paragraph numbers to refer to locations within the text, instead of using page numbers.

21. Figures 2, 3, 4: For the forest plots, please present the overall result from the meta-analysis, with 95% CI and p values.

22. Table 1: Please note in the legend that “CT” stands for cash transfer.

Comments from the reviewers:

Reviewer #1: See attachment

Michael Dewey

Reviewer #2: This paper explores an interesting topic by systematically examining results from published studies about cash transfers targeting infants or young children development outcomes. The originality of this study, compared to previous published reviews of the impact of cash transfers, is to address the delicate question of cash transfers combined with an additional intervention targeting child development and to explore whether this combination (named "Cash+") gives better results than one of the components alone (cash alone or "plus" alone). This kind of information is really needed to help decision makers, planners and various stakeholders design the best interventions possible in each particular context.

The study is rigorously conducted, the analysis includes meta-analyses when doable and the paper is well written. However, due to the small number of studies available, the variety of contexts and a large heterogeneity of the studies included in the review, both in terms of outcomes measured, in the kind of "plus" interventions and in the intervention designs, results are a bit disappointing. In the end, very limited evidence can be shown and no firm conclusion can be drawn from this work. The authors are not to be blamed for this, of course, and it's obvious that they did their best to assemble the available information, but the practical impact of such a study remains limited.

One wonders if other choices could have been made to draw more information from the corpus of 63 articles that were assessed for eligibility, by applying less strict inclusion and/or exclusion criteria. It is out of the scope of the review of this paper to fully examine all 43 excluded articles, and I know that when undertaking such a systematic review the right balance is really difficult to establish between being very inclusive (and having to deal with more heterogeneity) or very exclusive (and ending up with a too few number of studies). Nevertheless, it seems to me that some criteria could have been less strict in order to increase the number of studies to be included, at least in the narrative analysis (the qualitative synthesis).

- For example, they considered as outcomes only what they call "third order SDG-indicators" related to children <5y old. The rationale behind this choice is not well explained and doesn't sound as the most effective choice. Indeed, when discussing their results, the authors themselves note that reaching "distant" outcomes is far more difficult than more proximal ones (discussion section, page 2 lines 10-18). Why not, then, having considered more proximal outcomes? Indeed, in this part of the discussion the authors could have elaborated a bit more about what implications does that have; in particular because many authors tend to think that there has been a too strong focus on ultimate (nutritional indices) outcomes in recent years, while many interventions achieve good results on more proximal outcomes. Also, to me the reference to SDGs is not very appropriate since SDGs target populations as a whole, not necessarily individuals (e.g. SDG targets 1.1.1 to 1.2.2, 2.1.1, 3.2.1, 3.2.2, 3.3.1-3.3.5… that the authors mentioned in their paper are all SDG indicators that are measured at the population level).

- Also, the authors restrained the selection of studies to a list of "plus" components, referring again to the SDGs. First, in their protocol published on Prospero the authors mentioned SDG6 (water & sanitation) as one of the SDGs considered (Supplement 7, page 2) but they didn't mention that in the list of considered SDGs in the paper. If there was no study adding a WASH component to a cash transfer, which I doubt, they should at least have said so. Besides this, anyway, I think that referring to conceptual models of the young child development would have been a far better option than the SDGs (that don't have any conceptual model behind; and are not specifically targeting the child development per se).

- Still in terms of studies selection, the authors could also have thought back of the final objective of their work, which might have been to give useful information for planners rather than highlighting a lack of evidence on a narrower question: indeed, by strictly focusing on interactions between cash and plus components they excluded studies with a 'cash-alone' and a 'cash+' arms, but without a 'plus alone' arm. At the end of the day, this is likely to have prevented them to examine more in-depth the impact of cash+ versus cash-alone interventions, which is a very interesting question. Interestingly enough, the first sentence in the conclusions says "Questions in social protection have remained as to what complementary interventions are necessary for cash transfers to have maximal impact" (page 31, lines 1-2). Which is exactly my point, but not exactly the question they addressed in their review.

- I also checked one or two excluded studies and failed to find an obvious reason for exclusion, even with the strict criteria applied in the study (e.g. the first in the alphabetical order: Adubra et al. 2019).

Again, I know how harsh it is to conduct such a systematic analysis. All the above is not to say that the study was badly conducted (which is not the case) but I wanted to highlight that other choices could have led to a more informative review. This can't be for sure, of course, but this deserves at least to be discussed quite extensively in the discussion section.

Another major point I'd like to make regards the absence of information about the evaluation designs. While there is information about duration, age range, etc., it's obvious when reading the abstracts of the included studies that some of them used longitudinal designs (same children are followed-up and surveyed at baseline and endline) while others are based on repeated cross-sectional surveys at baseline and endline, on children of the same age range. This makes a huge difference in terms of interpretation, potential biases, duration of exposition to the intervention (for repeated cross-sectional designs) and statistical power. I found it very strange that this was not accounted for in the study and that this point is not even mentioned I the discussion section.

Also, regarding studies for which the outcome is the child undernutrition, the authors analysed the impact in terms of rates of undernutrition. This is not the most powerful analysis, in particular for studies using a longitudinal evaluation design. Why not analysing the results in terms of changes of the continuous underlying variables (e.g. height-for-age or weight-for-length/height z-scores)?

My last major comment regards the discussion section. The paper is already quite long; results are well and quite extensively described and the discussion starts with a summary of the results (page 25), which is good. As said above, and as acknowledged by the authors, at this point we don't have much evidence, no firm conclusions, nor much information we don't already know. I understand that the authors made some efforts to comment about some characteristics of cash or cash+ interventions that could increase impacts on some outcomes. But let's face it: they don't have much to say that really comes from their study.

- About the context, for example, they draw some comments on the role of food transfers (page 26, first paragraph) but all what that says is that in emergency contexts, when dealing with acute undernutrition in infants and young children, it is key to give access to some nutritious foods (that are indeed most often ready-to-use supplementary foods) on top of the cash transfer. We already know that but it's good to have some combined evidence. However, the rest of the paragraph is made of more general statements that are not really related to the work presented.

- Similarly, when commenting on the importance of the age of children for interventions targeting stunting (page 26, lines 14-18), the authors refer to the "wider literature in cash transfers" while in fact this has been demonstrated years ago in the nutrition literature. And in a few other parts of the discussion, the authors err on the side of making general comments that are not directly related to their results. This is likely because results are poor, of course, but then the discussion can be shorter.

- The first paragraph of the recommendations for research, again, describes good but already well known points about cash or + intervention components, but those statements don't come specifically from this review.

Minor comments:

- Please avoid the term "malnutrition" when it obviously refers to "undernutrition" (which is the preferred term, since "malnutrition" encompasses also overweight, obesity and related noncommunicable diseases). Page 7 line 18, on the contrary, "malnutrition" should replace "malnourishment".

- Please note also that "underweight" is not a type of undernutrition (such as stunting and wasting); it is, indeed, only an anthropometric indicator of one of the two types above, or both.

- UCT and CCT acronyms are used without being defined

- Page 10, lines 8-9: this last sentence of the paragraph belongs to the methods section (where it is already said: page 9, line 5) but not to the results section.

- Page 11, lines 3-4: "CCTs (7 studies), UCTs (11 studies), and public works programmes (3 studies)" add up to 21 studies, not 20.

- Page 11, lines 7-9: even if the information is in the table, it would be good to have the number of studies for each category of 'plus' components.

Reviewer #3: This is a timely and important topic, as more countries and development partners increasingly implement cash plus or integrated programming to address multidimensional poverty. This paper is well motivated and the background is well written. I have the following comments to improve the manuscript.

1. I find the comparison to either cash alone or plus alone somewhat problematic. If the authors stick to cash plus v. cash only, then you can make conclusions about whether 1) the plus component is effective or 2) "cash plus" is more effective that cash alone (if there are no differences you would conclude it is not). However, the combination of comparing cash plus to either cash alone or plus alone makes it difficult to understand what the conclusion is. I agree with their decision to do meta-analysis on cash plus v. cash only (avoiding mixing in of plus only with cash only as a comparison) but question the descriptive inclusion of cash plus to plus only. In an ideal world, you would want to compare studies that had the following three arms: 1) cash only, 2) plus only, 3) cash plus and 4) controls. This would allow you to understand impacts of "cash plus" and the synergies of cash and plus components (above and beyond either component separately). However, the authors could not do this because a sufficient number of studies with this design does not exist. Thus, the authors had to make a choice and in the background/framing, they seem to choose not to understand overall impacts of cash plus, but rather they start with cash as "standard of care" and seek to understand whether cash plus is more effective than cash alone. Following from this logic and the theoretical framing, they should drop the comparison to plus only, which just tells you whether cash is effective by itself or whether the combination is more effective than services/other interventions alone.

2. Page 21: Lines 6-11: It's not clear whether the discussion of heterogenous impacts follows the current paper's model of cash plus v. cash only, or whether they are citing findings from the original studies being reviewed, which may have compared to a pure control, and therefore should be removed from the results section.

3. This line seems out of place and doesn't connect to the sentence or larger paper: "one study emphasises that the burden of high fertility rates puts pressure on already strained social services in meeting the demands for early child development (Premand & Barry, 2020)."

4. The language on page 26, line 27 should be tempered. While the studies examined are rigorous, there are still a very limited number of studies examining cash plus v. cash alone and therefore the ability to generalize is still limited until more studies are published.

5. In the comparison to Manley et al. (2020), can the authors add a sentence to explicitly state how this study adds to the literature (i.e., what the combined findings suggest)? They appear to be saying that since Manley et al. found impacts of cash plus or cash plus on stunting, and the current authors find no difference between cash only and cash plus, that cash alone appears to be driving the impacts. It would help if the authors could clarify their conclusion.

6. This conclusion should be tempered and contexualized within the limitations of the study: "Findings also suggest limited or no added impact of nutrition behaviour change and psychosocial stimulation interventions and possibly worsening inequality from some cash-plus combinations compared to cash alone." In fact, Roy et al. (2019) have found that in terms of sustained impacts in Bangladesh, transfers + BCC had sustained impacts whereas transfers only did not on several pathways and outcomes important for child health (i.e., maternal IPV, economic resources, agency, social control and household poverty). The limitations of the current paper should be more clearly stated, that they are simply comparing effectiveness of the "plus" components and in fact have not compared cash v. control as compared to cash plus v. control. The statements/conclusions they are making would require the latter, which the authors have not done.

Any attachments provided with reviews can be seen via the following link:

[LINK]

Submitted filename: little.pdf

Click here for additional data file.

12 May 2021

Submitted filename: Response to Reviewers.docx

Click here for additional data file.

11 Jun 2021

Dear Dr. Little,

Thank you very much for re-submitting your manuscript "Effectiveness of cash-plus programmes on early childhood outcomes compared to cash transfers alone: A systematic review and meta-analysis in low- and middle-income countries." (PMEDICINE-D-21-00432R2) for consideration in PLOS Medicine’s Special Issue: Global Child Health: From Birth to Adolescence and Beyond.

I have discussed the paper with my colleagues and the academic editor and it was also seen again by three reviewers. I am pleased to say that provided the remaining editorial and production issues are dealt with we are planning to accept the paper for publication in the journal.

The remaining issues that need to be addressed are listed at the end of this email. Any accompanying reviewer attachments can be seen via the link below. Please take these into account before resubmitting your manuscript:

[LINK]

***Please note while forming your response, if your article is accepted, you may have the opportunity to make the peer review history publicly available. The record will include editor decision letters (with reviews) and your responses to reviewer comments. If eligible, we will contact you to opt in or out.***

In revising the manuscript for further consideration here, please ensure you address the specific points made by each reviewer and the editors. In your rebuttal letter you should indicate your response to the reviewers' and editors' comments and the changes you have made in the manuscript. Please submit a clean version of the paper as the main article file. A version with changes marked must also be uploaded as a marked up manuscript file.

Please also check the guidelines for revised papers at http://journals.plos.org/plosmedicine/s/revising-your-manuscript for any that apply to your paper. If you haven't already, we ask that you provide a short, non-technical Author Summary of your research to make findings accessible to a wide audience that includes both scientists and non-scientists. The Author Summary should immediately follow the Abstract in your revised manuscript. This text is subject to editorial change and should be distinct from the scientific abstract.

We expect to receive your revised manuscript within 1 week. Please email us (plosmedicine@plos.org) if you have any questions or concerns.

We ask every co-author listed on the manuscript to fill in a contributing author statement. If any of the co-authors have not filled in the statement, we will remind them to do so when the paper is revised. If all statements are not completed in a timely fashion this could hold up the re-review process. Should there be a problem getting one of your co-authors to fill in a statement we will be in contact. YOU MUST NOT ADD OR REMOVE AUTHORS UNLESS YOU HAVE ALERTED THE EDITOR HANDLING THE MANUSCRIPT TO THE CHANGE AND THEY SPECIFICALLY HAVE AGREED TO IT.

Please ensure that the paper adheres to the PLOS Data Availability Policy (see http://journals.plos.org/plosmedicine/s/data-availability), which requires that all data underlying the study's findings be provided in a repository or as Supporting Information. For data residing with a third party, authors are required to provide instructions with contact information for obtaining the data. PLOS journals do not allow statements supported by "data not shown" or "unpublished results." For such statements, authors must provide supporting data or cite public sources that include it.

To enhance the reproducibility of your results, we recommend that you deposit your laboratory protocols in protocols.io, where a protocol can be assigned its own identifier (DOI) such that it can be cited independently in the future. Additionally, PLOS ONE offers an option to publish peer-reviewed clinical study protocols. Read more information on sharing protocols at https://plos.org/protocols?utm_medium=editorial-email&utm_source=authorletters&utm_campaign=protocols

Please review your reference list to ensure that it is complete and correct. If you have cited papers that have been retracted, please include the rationale for doing so in the manuscript text, or remove these references and replace them with relevant current references. Any changes to the reference list should be mentioned in the rebuttal letter that accompanies your revised manuscript.

Please note, when your manuscript is accepted, an uncorrected proof of your manuscript will be published online ahead of the final version, unless you've already opted out via the online submission form. If, for any reason, you do not want an earlier version of your manuscript published online or are unsure if you have already indicated as such, please let the journal staff know immediately at plosmedicine@plos.org.

If you have any questions in the meantime, please contact me or the journal staff on plosmedicine@plos.org.

We look forward to receiving the revised manuscript by Jun 18 2021 11:59PM.

Sincerely,

Caitlin Moyer, Ph.D.

Associate Editor

PLOS Medicine

plosmedicine.org

------------------------------------------------------------

Requests from Editors:

1. Abstract: Please define the abbreviation “SDG” and “LMIC” at first use.

2. Abstract: Methods and Findings Line 26-27: Please provide the result for “...but had no added impact in improving weight-for-height or weight-for-age”

3. Author summary: Please rename the “non-technical summary” as “Author Summary”

4. Author summary: What do these findings mean?: In the first point, please qualify this with “There are few studies to date…” or similar. In the second point, please be more specific or clarify what is meant with “...improving the lives of vulnerable children.“

5. Introduction: Page 5 Line 9: We suggest revising this sentence to clarify/replace the word pernicious.

6. Results: If feasible, please include a sensitivity analysis omitting the 3 non-randomized studies.

7. Results: Page 17: Here and throughout the text, please use person first language where possible (avoid referring to children as “stunted” or “wasted”).

8. Results: Page 20 Line 1-2: Please revise to : “...were no longer statistically significant four years post-programme.” or similar.

9. Results: Page 24 Line 18: We would suggest providing the specific reference here, to clarify.

10. References: Please check to ensure that the "Vancouver" style is used for reference formatting (including Journal Title abbreviations, for example PLOS Med should be PLoS Med for reference 11), and see our website for other reference guidelines https://journals.plos.org/plosmedicine/s/submission-guidelines#loc-references

11. Supporting information files: Please rename/refer to files as S1 text, S1 Checklist, S1 Table, etc. or similar. On page 31 where the files are listed, please provide descriptive titles and legends, where appropriate.

12. PRISMA checklist: Please remove all references to page numbers (please refer only to section and paragraph locations).

13. Supplement 3: Please more clearly indicate the study designs and/or otherwise indicate the three non-randomized studies.

14. Supplement 7: Please provide a clean version of this document (without the comment).

Comments from Reviewers:

Reviewer #1: The authors have met most of my points. It is hard to believe that Google Translate could not cope with the article not in English nor, assuming it is the one in Spanish, that nobody at any of the institutions to which the authors are affiliated can read Spanish but if the authors wish to leave it out there is nothing much to be done.

Michael Dewey

Reviewer #2: This paper has been greatly improved as compared to the previous version. The comments I made have been satisfactorily taken into account. In particular, the discussion section is much more concise and "to the point". The authors have to be commended also for having to re-run analyses. Good job !

I have only a few additional (very)minor comments to make:

- In the abstract (page 2 line 5) then in the introduction section (page 5 lines 18-19) there are two expressions that are not readily understandable: "individual behavioural mediators" and "supply-side moderators". Indeed, the explanation of their meaning comes only in the result section (page 13, lines 2 and 4). I suggest to re-write (e.g. "improved mediating outcomes and/or enhanced supplies")

- In the last bullet of the non-technical summary, I suggest to explicitely mention the main limitation of the paper, i.e. the small number of studies that were available for analysis (despite all efforts); indeed, saying "more research is needed" is not enough, in my opinion (and can be said for almost all papers!)

- The expression "caloric-food transfer" (page 7 line 27 and page 21 line 6) is not a standard one and has no real meaning from a nutritional point of view; indeed, all foods are "caloric" by nature (except water); if the authors want to insist that the food-transfer used products that are high in calories, the correct expression would be "energy-dense foods", but even that wouldn't be correct regarding the 3 studies that are analyzed, since at least one of them used "normal" foods (not specificaly designed energy-dense products). Please replace by "food transfers".

- Page 20, line 20, the Calvo-Dercon Poverty Score is not known by everyone; adding a bibliographic reference would be welcome.

Yves MARTIN-PREVEL

Reviewer #3: The authors have addressed my concerns and I recommend publishing.

Any attachments provided with reviews can be seen via the following link:

[LINK]

14 Jun 2021

Submitted filename: Response to Reviewers.docx

Click here for additional data file.

15 Jun 2021

Dear Dr Little,

On behalf of my colleagues and the Academic Editor, Zulfiqar Bhutta, I am pleased to inform you that we have agreed to publish your manuscript "Effectiveness of cash-plus programmes on early childhood outcomes compared to cash transfers alone: A systematic review and meta-analysis in low- and middle-income countries." (PMEDICINE-D-21-00432R3) in PLOS Medicine’s Special Issue: Global Child Health: From Birth to Adolescence and Beyond.

Before your manuscript can be formally accepted you will need to complete some formatting changes, which you will receive in a follow up email. Please be aware that it may take several days for you to receive this email; during this time no action is required by you. Once you have received these formatting requests, please note that your manuscript will not be scheduled for publication until you have made the required changes.

In the meantime, please log into Editorial Manager at http://www.editorialmanager.com/pmedicine/, click the "Update My Information" link at the top of the page, and update your user information to ensure an efficient production process.

PRESS

We frequently collaborate with press offices. If your institution or institutions have a press office, please notify them about your upcoming paper at this point, to enable them to help maximise its impact. If the press office is planning to promote your findings, we would be grateful if they could coordinate with medicinepress@plos.org. If you have not yet opted out of the early version process, we ask that you notify us immediately of any press plans so that we may do so on your behalf.

We also ask that you take this opportunity to read our Embargo Policy regarding the discussion, promotion and media coverage of work that is yet to be published by PLOS. As your manuscript is not yet published, it is bound by the conditions of our Embargo Policy. Please be aware that this policy is in place both to ensure that any press coverage of your article is fully substantiated and to provide a direct link between such coverage and the published work. For full details of our Embargo Policy, please visit http://www.plos.org/about/media-inquiries/embargo-policy/.

To enhance the reproducibility of your results, we recommend that you deposit your laboratory protocols in protocols.io, where a protocol can be assigned its own identifier (DOI) such that it can be cited independently in the future. Additionally, PLOS ONE offers an option to publish peer-reviewed clinical study protocols. Read more information on sharing protocols at https://plos.org/protocols?utm_medium=editorial-email&utm_source=authorletters&utm_campaign=protocols

Thank you again for submitting to PLOS Medicine. We look forward to publishing your paper.

Sincerely,

Caitlin Moyer, Ph.D.

Associate Editor

PLOS Medicine