Objectives: We compared the sensitivity and positive predictive value (PPV) of the Modified Checklist for Autism in Toddler (M-CHAT) with those of the Parents Evaluation of Developmental Status (PEDS) in a sample of young children who completed a clinical evaluation.
Methods: Participants were identified from an ongoing screening study at Georgia State University (GSU) that involves administration of the M-CHAT and PEDS during a routine 18- or 24-month well child visit. If ASD risk was indicated on the M-CHAT, a member of the study team called the family and administered a follow-up interview. If ASD risk was still indicated, the family was invited for a clinical evaluation that included a cognitive and adaptive measure and several autism diagnostic measures. A licensed clinical psychologist diagnosed the children with ASD or nonASD given all available data. Performance on the PEDS did not influence whether a family was invited for the clinical evaluation.
Results: A total of 3998 children were screened with the M-CHAT; the sample was limited to 52 participants who completed both the M-CHAT and PEDS and received a clinical evaluation. Forty-seven children failed the M-CHAT and five received an evaluation because the pediatrician noted concerns on the M-CHAT form. Forty children qualified for PEDS Path A, defined as two or more predictive concerns, and eight children qualified for PEDS Path B, defined as only one predictive concern. The mean age at the time of screening and evaluation were 21 and 26 months, respectively. Thirty children (58%) received an ASD diagnosis after the clinical evaluation. The sensitivity and PPV of the PEDS were .73 and .55 for Path A, .14 and .33 for Path B, and .93 and .58 for Paths A+B combined. The sensitivity and PPV of the M-CHAT were .93 and .60.
Conclusions: The M-CHAT detected more children with ASDs than the PEDS Path A or Path B and as many children with ASDs as the PEDS Paths A+B combined. The PPV of the M-CHAT was higher than any PEDS Path, suggesting use of the PEDS alone may result in an over-referral of ASD assessments. This point is highlighted by the fact that 968 children who passed the M-CHAT failed the PEDS Path A or B; our sub-sample of children who had concerns noted on the M-CHAT may have artificially inflated the PEDS PPV. These results support autism-specific screening in addition to general developmental screening during 18- and 24-month well child visits, to streamline referrals to ASD specialists. Moreover, given the fairly low predictive value of the M-CHAT in this study, we encourage comprehensive evaluation of all children who fail an ASD-specific screen to verify ASD diagnosis.