A cluster-randomized controlled trial to measure the impact of a parenting intervention on child and parent outcomes in rural China

The project will be run as a randomized controlled trial with one intervention group and a control group with no intervention. Parenting centers will be based in a central location in each village, and all caregivers living in the village with children aged 6-24 months will be welcome to drop by at any time during regular operating hours. The parenting centers will be based in re-purposed office spaces that will be renovated to be child friendly (e.g. colorful walls, non-lead-based paint, soft floors). All parenting centers will include a central playroom, will be the same size, and have the same types of toys. 

The parenting curriculum that will be used in the parenting centers has been carefully developed according both to child development needs, and to the local environment in rural China. It consists of weekly interactive lessons targeting caregivers of babies aged 6 to 36 months (124 lessons total). Each month consists of activities (involving both caregivers and children) designed to teach caregivers how to interact with their children in ways that stimulate development in four dimensions considered essential by child development experts:
1. Cognitive
2. Motor
3. Language
4. Social-emotional development
This parenting curriculum has already undergone preliminary field testing and evaluation in rural China by our government partners at the National Health & Family Planning Commission (NHFPC) and our own evaluation team. It is loosely based on the “Jamaica curriculum” developed by Sally Grantham-McGregor and adapted to many different underdeveloped settings around the world, but it is fully designed and developed for a rural Chinese population, and is owned and copyrighted by the NHFPC.
Work on the full-scale RCT should begin in November, 2015. It is anticipated the parenting intervention will continue for 21 months, through August, 2017. 

A sample of all households in the village with children aged 6-24 months will be selected for participation in the evaluation surveys. All data on primary outcomes will be collected from these households.

2015-12-07

2018-06-08

1. Children’s cognitive, language and motor development (as measured by the Bayley Scales of Infant Development—BSID) 
2. Children’s social-emotional development (as measured by the Ages & Stages Questionnaire)

Measured prior to treatment assignment (baseline), and at the conclusion of all interventions (endline).

The impacts of village-based parenting centers on child development will be tested through a cluster-randomized controlled trial. In this trial, 100 villages will be randomly selected and allocated between an intervention group (50 villages) and a control group (50 villages) with no intervention.

Parenting centers will be based in a central location in each village, and all caregivers living in the village with children aged 6-36 months will be invited to visit centers at any time during operating hours. The parenting centers will be based in repurposed office spaces that will be renovated to be child friendly (eg: colorful walls, non-lead-based paint, soft floors).

All parenting centers will include a central playroom, and will be the same size, and have the same types of toys. All parenting center toys will be split into two types: (1) large, “free play” toys such as a small climbing structure, hula hoops, big plastic bouncy balls, a toy drum set, etc.; and (2) smaller toys associated with a week-by-week parenting curriculum (henceforth, “curriculum toys”).

The parenting curriculum that will be used in the parenting centers has been carefully developed according both to child development needs, and to the local environment in rural China. It consists of weekly interactive training courses targeting caregivers of babies aged 6 to 36 months (124 training courses total). Each month consists of activities (involving both caregivers and children) designed to teach caregivers how to interact with their children in ways that stimulate development in four dimensions considered essential by child development experts: a.) cognitive, b.) psychomotor, c.) language, and d.) social-emotional development. This parenting curriculum is fully designed and developed for a rural Chinese population and has already undergone preliminary field testing and evaluation in rural China by NHFPC and the research team.

Each center will have two staff members: a full-time center monitor, and a parenting instructor. The full-time center monitor will be a local villager who manages the center and takes care of day to day maintenance, such as regular disinfecting of the toys, tidying up, ensuring caregiver and child adherence to center rules, etc. The monitor will also run daily reading programs and assist the parenting instructor as needed. The parenting instructor will be a government worker from the township-level Health & Family Planning Commission (Local HFPC). He or she will undergo comprehensive training in child development and in the implementation of the NHFPC parenting curriculum by the Training & Communication Center of the NHFPC. Once a week, the parenting instructor will teach local caregivers the stage-based (age-appropriate) curriculum-based activity of the week. The parenting instructors’ goal is to teach caregivers how to interact with their child in such a way as to foster the child’s motor skills, cognition, language ability, and social-emotional development.

Randomization done in office by a computer

Village level randomization

Yes

We included 100 village clusters with 12 participating households on average per village.

100 villages will be randomly selected and allocated between an intervention group (50 villages) and a control group (50 villages) with no intervention. In total, allowing for attrition, we expect to enroll 1000 households in the study, amounting to around 2000 individuals (1000 caregivers + 1000 children).

100 villages will be randomly selected and allocated between an intervention group (50 villages) and a control group (50 villages) with no intervention.

To determine the sample size required, we performed power calculations using Optimal Design software. We calibrated the model as appropriate for a Cluster Randomized Trials with a continuous outcome (scores on both the Bayley Scales of Infant Development) and clustering at the village level. The power needed to detect a difference in outcome variables between the treatment and control groups in a Cluster Randomized Trial with a continuous outcome depends on 5 factors: a.) the number of babies per township (n); b.) the number of townships (J); c.) the intra-township correlation in outcome variable (ρ); d.) the minimum effect size that we would expect to be able to detect (δ); and e.) the proportion of variation in the true township-level post-intervention outcome variable explained by township-level pre-intervention outcome variable (R2). Based on our previous study of a household-based parenting intervention (also conducted in poor, rural areas of Shaanxi Province) we assume there will be on average 12 babies aged 6-24 months in each village at endline taking into account attrition from the sample over the study period, R2=0.5, and an intra-cluster correlation of 0.1. Based on these parameters, we estimate that we require 50 villages per arm to detect a 0.2 standard deviation effect on cognitive scores with 80% power at a significance level (α) of 0.05.