Effectiveness of early intervention and prediction of outcomes in taiwanese children with developmental delay

Background/purpose: In the 1990s, the Taiwanese government introduced a nationwide program for enhancing the development of children with developmental delay (DD) at the earliest point. As a part of this program, children with suspected DD or those at a risk of DD are referred to child assessment centers, where they undergo comprehensive multidisciplinary assessments (MDAs) using standardized screening tools to determine their developmental status. This study investigated the correlation between developmental outcomes and child-level factors by using serial MDA data, and we also aimed to develop an MDA result-based predictive model for identifying factors facilitating functional improvement in children with DD. Methods: This study included children who underwent at least two MDAs at Taipei Medical University Hospital between 2011 and 2020. DD and borderline DD were defined as scores more than1.5 and 1.0–1.5 standard deviations (SD), respectively, below the mean score in age-appropriate standardized norm-referenced tests. MDA results across various developmental domains were converted into a total developmental score. A machine-learning-based predictive model was constructed to differentiate between children with functional improvement and those without it. Results: The final analysis included 684 children (1368 MDAs). Of them, 58.9 % exhibited improved total developmental scores. Children who initially exhibited normal development in socioemotional skills, fine motor abilities, or language skills during the initial assessment were likely to exhibit improved outcomes. Conclusion: This study underscores the effectiveness of early intervention services in Taiwan. Consistent reassessments may facilitate subsequent educational interventions. Future studies should explore effective and efficient early intervention models for addressing DD.