MSD urged to adopt predictive tool to identify at risk children