Main findings

The authors include a total of 56 studies assessing the effectiveness of 30 different interventions in 24 countries across Africa, Latin America and the Caribbean, Europe, Asia and the Pacific. The review includes randomised trials and quasi-experimental designs using difference-in-difference, propensity score matching, instrumental variables or other simultaneous equations estimation techniques.

Overall, the study results show a positive impact of cash-transfer, nutrition-only, nutrition and care, and nutrition and education interventions on cognitive, behavioural, health and schooling outcomes. The authors report the following findings:

Impact by duration of intervention: interventions lasting between 1 and 3 years yield slightly stronger results than shorter interventions.Impact by intervention type: cash transfers, nutrition interventions and mixed interventions all have a small positive impact on the measured outcomes. Mixed interventions have the strongest impact.

Impact by study design: studies with more rigorous research designs (randomisation or propensity score matching) yield smaller effects than less rigorous studies.Impact by age group: interventions targeting younger children (infants, toddlers, pre-kindergarten) yield slightly better results than interventions targeting children in all age ranges at the same time.

Impact by economic status of country: intervention impacts in low- and middle-low-income countries are smaller than impacts in middle- and middle-high income countries.

Impact by target outcome:

• Cognitive outcomes: cash-transfer interventions yield the smallest impact, nutritional interventions an intermediate impact, and educational interventions the largest impact on cognitive outcomes. Short-term impact is slightly higher than long-term impact on average cognitive outcomes.Impact by target outcome:
• Behavioural outcomes (social and emotional): cash transfer and early educational interventions have a small positive impact, and nutritional interventions have a larger positive impact on behavioural outcomes. There is no difference in short-term versus long-term intervention impact on behavioural development outcomes.
• Health outcomes: cash transfers, nutritional interventions and combined nutritional and educational/child-care interventions all have a positive impact on health outcomes. Short-term impacts are stronger than long-term impacts.
• Schooling outcomes: nutrition interventions have a small positive impact, and mixed nutrition and education/stimulation interventions have a larger impact on schooling outcomes. There are no differences in short-term versus long-term impacts of interventions on schooling outcomes.

The authors also find that intervention effects are context-dependent and smaller in low-income countries. Mixed interventions are particularly effective for cognitive outcomes, and interventions targeting infants or preschoolers were more effective than untargeted interventions for behavioural outcomes. Health outcomes depend on intervention dosage, and (unexpectedly) longer intervention durations had less impact on schooling outcomes. The authors recommend that future research should aim to include information on costs of interventions, as well as information on the design and implementation of the interventions.

Background

Environmental influences on child development have received growing attention in the past few decades. There is also an increased recognition that child development needs to be promoted early in order to improve future developmental and educational outcomes. Different types of early childhood intervention have been proposed to promote child development, including cash transfers and educational, nutritional and mixed interventions. Yet different bodies do not agree on which interventions are the most effective and what the necessary duration of such interventions should be. Compared with the developed world, intervention coverage in the developing world is sparse and irregular. Given monetary constraints, cost-effective interventions are needed, along with reliable information about whether such interventions should have a single or multi-purpose focus. This systematic review presents the first meta-analysis of early-childhood intervention evaluations outside the United States, considering the impact of research design, context, services and other intervention factors on short- and long-term intervention outcomes.

Research objectives

The objective of this systematic review is to examine evidence from a range of early childhood interventions from outside the USA and Canada, and to summarise this evidence according to short- and long-term effects with respect to cognitive, behavioural, health and schooling outcomes.

Methodology

The authors included randomised trials and rigorous quasi-experimental studies, assessing the effect of cash transfers, nutrition-only, nutrition and care, and nutrition and education interventions on cognitive gains, behavioural change, health gains and amount of schooling of children. The included population was not specified, but study locations had to be outside the USA and Canada.

The authors conducted computer-aided searches of the published and unpublished literature, including databases like JSTOR, ScienceDirect and PubMed. The authors also reviewed reference lists of identified studies.The authors do not report assessing study quality. They analysed the included studies along several dimensions (by intervention type, by outcome, etc.) and reported average Cohen’s D effect sizes for all four outcomes. Additionally, they estimated the effects of a range of factors on the various effect sizes, using maximum-likelihood random-effects estimation.The authors conducted computer-aided searches of the published and unpublished literature, including databases like JSTOR, ScienceDirect and PubMed. The authors also reviewed reference lists of identified studies.

Quality assessment

The review uses appropriate methods to reduce risk of bias, in terms of a reasonably comprehensive search strategy and well-reported criteria for study inclusion. However, the review has some major limitations. The most important limitation is that although the review explore whether the results differ by study design, it does not include an assessment of the quality of the included studies. Although the authors only include study designs considered as appropriate for causal inference, they do not systematically explore whether potential flaws in the included studies affect the validity of the findings. Moreover, the review does not report much information about the characteristics of the participants, interventions and outcomes reported in the included studies. There are also limitations in the
reporting of the search, and the study does not include information on the screening or data-extraction process. Nevertheless, the review is limited to evidence from randomised trials and rigorous quasi-experimental studies and does not draw strong policy conclusions, and this mitigates some of the limitations.

Source link