Validity of the Autism Diagnostic Observation Schedule-2 (ADOS-2) in South Korean Toddlers and Preschoolers with ASD from 12-47 Months of Age

Background: The gold-standard diagnostic measure, the Autism Diagnostic Observation Schedule-2 (ADOS-2; Lord et al., 2014), has shown strong diagnostic validity in individuals with ASD, including young children from 12 to 47 months of age (Luyster et al., 2009; Gotham et al., 2007). However, the studies of the diagnostic validity of the ADOS have been limited to samples obtained from the Western countries, limiting the evidence for the valid use of the ADOS in the Eastern countries, although initial attempts have been made recently to examine its validity in Chinese (Sun et al., 2015) and Indian (Rudra et al., 2014) children with ASD. A recent translation of the ADOS-2 in Korean (2017) has increased the use of the ADOS in South Korea significantly for both research and clinical purposes, which highlights the urgent need to examine the validity of the ADOS for South Korean children.

Objectives: To provide the initial validity of the ADOS Toddler Module and Module 1-2 in South Korean toddlers and preschoolers.

Methods: The ADOS-2 was administered by a research team led by a research reliable clinician to 143 children clinically referred for ASD (68 children with ASD, 39 with non-spectrum (NS) disorders [e.g., developmental delays]), and 36 typically developing (TD) children recruited as a comparison group in a metropolitan area of South Korea. The mean age was 33.1 months (SD=9.1) and a mean nonverbal IQ, 81.5 (SD=25.1). Fifty-six children received the Toddler Module, 67 children, Module 1, and 20 children, Module 2. Algorithm and item scores were compared based on the ANOVAs for the ASD vs. NS/TD groups. Sensitivity and specificity using the ADOS algorithm cutoffs ("ASD cutoffs"; Lord et al., 2014) for each Module and age/language group were computed for the ASD vs. NS/TD diagnostic comparisons. Correlations (Pearson’s r) between algorithm totals and developmental factors (e.g., nonverbal mental age, chronological age) were computed to examine the effects of these factors on the measurement of autism symptom severity.

Results: The ASD group showed significantly higher ADOS algorithm totals, and social affect and repetitive and restricted behavior (RRB) domain scores across all Modules (p<0.05) compared to the NS/TD groups. Likewise, the ASD group scored significantly higher on the majority of item scores on the diagnostic algorithms across all Modules, with the greatest variability in Toddler Module (p<0.05). Sensitivities were excellent, ranging from 94% to 100% across different Modules. Specificities varied more, ranging from 75% (Module 1, Few to No Words algorithm) to 100% (Module 2). Algorithm totals had mild-to-moderate correlations with NVMA and age (range=r=0.1 to 0.5).

Conclusions: The current study demonstrates promising, initial evidence for the diagnostic validity of the ADOS-2 for clinically ascertained, South Korean toddlers and preschoolers with ASD from 1-4 years of age. Results suggest that the ADOS-2 could be validly implemented as a diagnostic tool by highly trained clinicians, with minimal cultural adaptations, in research and clinical settings for South Korean children. This study is one of the first steps in current efforts for the cross-cultural validation of the ADOS-2 in non-Western countries.