KeySteps@JC - Early Child Intervention

Socioeconomic disadvantage permeates multiple levels of environmental contexts in which children are raised, including the family environment and schools. Disparities in life outcomes can be observed in early childhood and developmental gaps widen over time. Continued exposure to adverse conditions propagates the developing child on an increasingly risky trajectory. This study aims to mitigate the adverse effect of socioeconomic disadvantage on children and their families. This is a clustered randomised controlled study recruiting 32 kindergartens in two underprivileged districts (16 in each district) in Hong Kong. District is a blocking factor. Within each district, 8 kindergartens will be randomised in comprehensive intervention group while the remaining 8 will be in health support group. The families in the comprehensive intervention will receive a holistic package of interventions, including child interactive intervention, family empowerment, and health support. The families in the health support group will only receive health support intervention. In addition, a stepped wedge cluster randomised control trial design will be used to deliver and evaluate the health intervention package (health seminars/workshops, dental check‐up/treatment, visual check‐up/treatment). The schematic diagram of the design is shown in the figure below. Another randomisation will be conducted to determine the sequence in which the health intervention takes place. 4 KGs will be provided with the health intervention in each month sequentially from Nov 2018 to Jun 2019 (totally 8 months). The randomisation should ideally be stratified by district (SSP/TSW) and the original group allocation (Comprehensive/Health). In other words, in each month, 1 KG from SSP Comprehensive, 1 from SSP Health, 1 from TSW Comprehensive, 1 from TSW Health will receive the health intervention package. In analysis, the time exposed to intervention (e.g. Gp1 exposed to 8 months at Jun 2018) will be used to evaluate the effect of the intervention. Since the time exposed to intervention (i.e. the starting time of the intervention) is randomised, internal validity should be ensured. In addition, the stratification of district and the original group allocation should eliminate the influence of these factors.