Introduction: In recent years, parenting interventions for reducing child maltreatment in low- and middle-income countries have risen sharply up the global policy agenda. If policy is linked to evidence, we would expect to see a corresponding rise in rigorous evaluations of parenting programs in these countries, and hence, need for critical synthesis of this new evidence. Accordingly, we aimed to update our well-cited 2013 systematic review (Prevention Science; k=12 RCTs) on this topic, and to re-examine the evidence base in low-resource countries, including main effects and country-level moderators.
Methods: Results are synthesized from randomized trial and high quality quasi-experimental evaluations of parenting programs for families with children aged 2-17. Searches included global and regional electronic databases unrestricted by language, including language-specific databases in Chinese, Thai, Farsi, Spanish, as well as unpublished grey literature, trial registries, and communication with experts. We will use well-powered robust variance estimation meta-analysis to examine intervention effects on multiple indices of primary outcomes of child maltreatment and harsh parenting, as well as secondary outcomes of positive parenting, and child behavior. We examine country level moderators, by investigating variation in outcomes by world region, cultural and economic factors, including country-level poverty and family welfare policy indicators.
Results: Searches identified 80 randomized trials of parenting programs. Very few high quality QED evaluations were found. Data extraction and meta-analysis is underway and will be completed by February 2019.
Conclusion: Our finding of a more than 6-fold increase in number of RCTs (from k=12 in 2013 to over 80), is likely to reflect a surge of interest in obtaining rigorous evidence to inform policy on parenting and maltreatment prevention in LMICs. Since the majority of trials were dated post-2013, it suggests that the huge increase is due primarily to new trials being conducted, but also partly to our additional foreign language search methods. We discuss findings in relation to study quality/ risk of bias; in terms of extent to which predictors of effect heterogeneity at country level can be found; and lessons that can be learned for policy makers and prevention researchers.
Registered protocol: PROSPERO 2018: CRD42018088697