A Meta-analysis of the Diagnostic Accuracy of Autism Diagnostic Observation Schedule Module-1 for Autism Spectrum Disorders

Objective Autism Diagnostic Observation Schedule (ADOS) is considered gold standard for the diagnosis of Autism Spectrum Disorders (ASD). The authors evaluated the cumulative diagnostic accuracy of ADOS-Module 1 (ADOS(M1)) using the original diagnostic algorithm with meta-analysis and meta-regression. Methods The authors, electronically and manually searched for studies from 1999 to 2013 that evaluated the accuracy of ADOS(M1) using the original diagnostic algorithm in detecting ASD. Primary results of Sensitivity (Sn), Specificity (Sp) and Diagnostic Odds Ratio (DOR) for ADOS(M1) were summarized using random-effects model. Summary Receiver Operating characteristic Curves and its Area Under the Curve (SROC-AUC) were used to summarize overall diagnostic accuracy of ADOS(M1). The modifying effects of quality of study and sample size, on the diagnostic odds ratio, were investigated using meta-regression. Results A total of 7 cross-sectional studies provided data on 4,057 children. The pooled Sn, Sp, DOR and SROC-AUC for the overall diagnostic accuracy of ADOS(M1) were: 0.91 (95 % CI=0.88 to 0.93), 0.27 (95 % CI=0.24 to 0.30), 5.40 (95 % CI=1.51 to 19.23) and 0.52 respectively. Meta-regression analysis showed a non-significant relationship between ADOS(M1) and study quality as well as sample size. There were subgroup differences in the DOR. Conclusions The authors conclude that ADOS(M1) with the original diagnostic algorithm lacks overall diagnostic accuracy and pooled specificity to confirm the diagnosis of Autism Spectrum Disorders. ADOS(M1) with the revised diagnostic algorithm should be used instead for the diagnosis of this group of disorders.