Validation of the Autism Detection in Early Childhood (ADEC) As a Level 2 Screening Tool for Autistic Disorder

Background:  

The prevalence rate of Autistic Disorder (AD) alone is estimated to be as many as two to four out of every 1000 children. This rate is in stark contrast to previous prevalence rates of about 0.5/1000 during the early 1990s and about 1.2/1000 during the early 2000s. The significance of this increase is that clinicians and paediatricians are likely to see an increase in children presenting with AD in their practice settings and thus need sufficient tools and training to identify them. Screening for ASD is the first step to improve early identification of children who might be considered at risk of the disorder and in need of further assessment, intervention and services. There is a critical need for further research to develop and validate screening tools for ASD in young children referred for developmental difficulties. The Autism Detection in Early Childhood (ADEC) was developed as an effective screening tool for AD for children from 12 to 37 months old.

Objectives:  To investigate the psychometric properties (reliability and construct, concurrent, diagnostic and predictive validity) of the ADEC. Specifically, to examine how well the ADEC classifies children with AD as compared to the ADOS, the ADI-R and clinical judgment based on DSM-IV. In addition, to examine the screening properties of the ADEC using receiver operating characteristic (ROC) curve analyses to identify sensitivity and specificity and to determine the optimal cut-off score.

Methods:  Parents and health care professionals who were concerned that their child presented with possible risk of developing an ASD participated in this screening study and were assessed with a battery of tests (ADEC, ADOS, ADI-R, Mullen Scales and Vineland). The ADEC was administered independent from the diagnostic assessor and blind to the results of the diagnostic evaluation. Likewise, the diagnostic assessor who administered the ADOS and ADI-R was blind to the ADEC assessment result.  A best estimate clinical (BEC) DSM-IV diagnosis was made by the first author using all available information and assessment results, with the exclusion of ADEC data, to generate independent diagnoses.

The resulting sample consisted of 72 children with a BEC DSM-IV diagnosis of AD, 14 children with PDD-NOS, 49 children with other developmental disorders (ODD) (e.g. language and developmental delay) and 20 typically developing (TD) children aged between 12 to 37 months.

Results:  Internal consistency was high. The ADEC scores are also reliable across examiners and across test-retest administrations. The findings supported the construct, concurrent, diagnostic and predictive validity of the ADEC.  In addition, the ADEC’s factor structures are found to be similar to the proposed DSM-5 criteria. The ADEC also has high sensitivity, specificity and predictive values using a cut-off score of 11 with the validation sample.

Conclusions:  This study proposes the ADEC to be an effective Level 2 screening tool to identify children with AD ranging from 12 to 37 months old and has the potential to be established as an efficient, psychometrically valid and a reliably administered screening tool for infants with AD.