Assessing the accuracy of the Modified Checklist for Autism in Toddlers: a systematic review and meta-analysis.

Abstract

The Modified Checklist for Autism in Toddlers (M-CHAT) could be appropriate for universal screening for autism spectrum disorder (ASD) at 18 months and 24 months. Validation studies, however, reported differences in psychometric properties across sample populations. This meta-analysis summarized its accuracy measures and quantified their change in relation to patient and study characteristics. Four electronic databases (MEDLINE, PsycINFO, CINAHL, and Embase) were searched to identify articles published between January 2001 and May 2016. Bayesian regression models pooled study-specific measures. Meta-regressions covariates were age at screening, study design, and proportion of males. On the basis of the 13 studies included, the pooled sensitivity was 0.83 (95% credible interval [CI] 0.75-0.90), specificity was 0.51 (95% CI 0.41-0.61), and positive predictive value was 0.53 (95% CI 0.43-0.63) in high-risk children and 0.06 (95% CI <0.01-0.14) in low-risk children. Sensitivity was higher for screening at 30 months compared with 24 months. Findings indicate that the M-CHAT performs with low to moderate accuracy in identifying ASD among children with developmental concerns, but there was a lack of evidence on its performance in low-risk children or at age 18 months. Clinicians should account for a child's age and presence of developmental concern when interpreting their M-CHAT score. The Modified Checklist for Autism in Toddlers (M-CHAT) performs with low-to-moderate accuracy in children with developmental concerns. There is limited evidence supporting its use at 18 months or in low-risk children.