Early Start Denver Model effectiveness in young autistic children: a large multicentric randomised controlled trial in two European countries

Abstract

Background: Evidence regarding early interventions based on the Naturalistic Developmental Behavioral Interventions framework, such as the Early Start Denver Model (ESDM), suggests efficacy for autistic children. However, the effectiveness of ESDM across diverse cultural contexts remains under-researched.

Objective: To assess the effectiveness of ESDM compared with treatment as usual (TAU) on overall development in young children with autism spectrum disorder (ASD).

Method: This parallel, randomised controlled trial, using a modified Zelen design, was conducted in five Child and Adolescent Mental Health Services in France and Belgium. A total of 180 children aged 19-36 months, who met autism criteria and were referred by community professionals, were randomly assigned to either receive 12-hour weekly ESDM+TAU or TAU alone. The primary outcome was the change in developmental quotient (DQ) on the Mullen Scale of Early Learning, assessed blindly from baseline to 24 months post randomisation.

Findings: From September 2015 to March 2019, 180 children were randomly assigned to either ESDM+TAU (n=61, girls=21.7%) or TAU alone (n=119, girls=15.4%). Three children dropped out immediately after randomisation. Compared with TAU alone, children in the ESDM+TAU group did not significantly improve global DQ (endpoint mean difference 3.82 (95% CI -1.25 to 8.89), p=0.14).

Conclusions and clinical implications: Our findings suggest that ESDM+TAU cannot be universally recommended for young children with ASD. Further research is required to evaluate the long-term effectiveness of ESDM and identify subgroups that may benefit more, thereby guiding optimal implementation strategies.

Trial registration number: NCT02608333.

Keywords: Child & adolescent psychiatry.

Figure 1. Between-group differences for changes in outcomes from baseline to 24 months. (A) Principal outcomes from the Mullen Scales of Early Learning. The treatment effect measures on the forest plot are mean differences for each outcome. (B) Other outcomes. The treatment effect measures on the forest plot are mean differences for each outcome. ADOS-2, Autism Diagnostic Observation Schedule, second edition, BOSCC, Brief Observation of Social Communication Change, the higher the score, the more observable the autistic signs are; CSS, Comparison Severity Score, a higher score reflects a higher autism symptom severity; CSBS, Communication and Symbolic Behavior Scales, the higher the score, the higher the child’s skills are; DCMA, Dyadic Communication Measure for Autism, the higher the score, the more observable communication between the child and one parent is; DQ, developmental quotient; VABS-II, Vineland Adaptive Behavior Scales, second edition, the higher the score, the higher adaptive functioning in daily life is.