Discriminant Validity of the ADI-R in Latino Families: Identifying Differences Between Autism Spectrum Disorder and Developmental Delay

Background:  The development of culturally and linguistically appropriate diagnostic tools is crucial in the assessment of Autism Spectrum Disorders (ASD).  The Autism Diagnostic Interview-Revised (ADI-R; Lord, Rutter, & LeCouteur, 1994) is considered a gold-standard parent interview that captures children’s past and current development.  Although the ADI-R has been widely used in clinical and research settings in the diagnostic process, few studies have evaluated the use of the ADI-R in distinguishing between children with ASD and children with other developmental disabilities (e.g., Gray, Tonge, & Sweeney, 2008; Lord, Rutter, & LeCouteur, 1994).  Even fewer studies have assessed the validity of the ADI-R in Latino populations (e.g., Blacher, Cohen, & Azad, 2014; Magaña & Smith, 2013).  Thus, there is a need to appraise the utility of the ADI-R in the accurate identification of children with ASD in the Latino population.  This information is critical in identifying the appropriate supports and resources for parents and their children.

Objectives:  The current study aims to assess whether the use of ADI-R in a USA based Spanish speaking Latino population demonstrates adequate discriminant validity in distinguishing between children with ASD and children with DD. 

Methods:  The current study included Spanish speaking Latino parents (primarily mothers) of children and adolescents who were between 4 and 16 years of age and received a clinical diagnosis of ASD or a Developmental Delay (DD).  The official Spanish version of the ADI-R was administered by trained interviewers.  Medical records were obtained to establish the clinical diagnosis. Children were grouped by the clinical diagnosis and included diagnoses of ASD (n = 28) and DD (n= 21).     

Results:  Discriminant validity was assessed by comparing scores for the three domains on the ADI-R.  Converted scores were used as recommended by the updated algorithms (e.g., Rutter, LeCouteur, & Lord, 2003).  Preliminary analyses found that children with ASD had higher scores than children with DD on the Reciprocal Social Interaction domain, F(1, 47) = 6.61, p = .013, and the Restricted and Repetitive Behaviors domain, F(1, 47) = 6.78, p = .012.  No significant differences were observed between children with ASD and children with DD on the Communication – Nonverbal, F(1, 47) = 1.26, p = .268, and Communication – Verbal domains, F(1, 31) = 1.91, p= .176.  Means are displayed in Table 1.  Additional analyses will evaluate group differences at the subdomain and individual item level to identify specific areas that might be significant in the diagnostic process. 

Conclusions: The preliminary findings indicate that the Spanish ADI-R exhibits adequate discriminant validity for the Reciprocal Social Interaction and the Restricted, Repetitive Behavior domains, but not in the Communication (Verbal and Nonverbal) domain.  These results are in line with other studies finding that the validity of the ADI-R may not be as strong for certain populations (e.g., very young children; Tsuchiya et al., 2013).  Further analyses are needed to determine which behaviors may distinguish between children with ASD and children with DD.