A Comparison of ASD Screening in Toddlers By Global Developmental Screening and the Modified Checklist for Autism in Toddlers (M-CHAT)

Poster Presentation
Saturday, May 12, 2018: 11:30 AM-1:30 PM
Hall Grote Zaal (de Doelen ICC Rotterdam)
O. Kerub1, E. J. Haas2, G. Meiri3,4, N. Davidovitch5 and I. Menashe6, (1)Ben-Gurion University, Ministry of Health, Be’er Sheva,, Israel, (2)Ministry of Health, Beer Sheva, Israel, (3)Soroka Medical Center, Beer Sheva, Israel, (4)Soroka Medical Center, Beer Sheba, Israel, (5)Ben Gurion University in the Negev, Beer Sheva, Israel, (6)Public Health Department, Ben-Gurion University, Beer Sheva, Israel

Early diagnosis of autism spectrum disorder (ASD) improves long-term outcomes of existing interventions. Systematic screening for early ASD symptoms can considerably reduce the age of diagnosis of the disorder. The most widely used ASD screening tool is the Modified Checklist for Autism in Toddlers (M-CHAT) which has been shown to have high sensitivity and specificity in populations of various ethnic backgrounds.

In Israel, there is no ASD-specific screening program, but rather toddlers undergo serial health visits at government-funded maternal child health centers (MCHCs), which include global developmental screening (GDS) for age appropriate developmental milestones. While GDS may detect a variety of developmental problems, its efficiency in detecting ASD at early ages is unknown.

Objectives: To assess the screening efficiency for ASD of the GDS and M-CHAT among toddlers from an ethnically diverse population in south Israel.


We applied both GDS and M-CHAT screening tests to 1,600 toddlers of ages 18-30 months in 35 randomly selected MCHCs in southern Israel. Toddlers who failed in either the M-CHAT or the GDS were referred to child neurologist or child psychiatrist for a more rigorous developmental evaluation based on DSM-5 criteria. In addition, we assessed the current GDS-based referral procedure by asking eleven experienced MCHC nurses to review and provide a referral decision for 200 toddlers based on their full GDS result history.

Results: Of the 1600 toddlers in our study, 84 were referred for further developmental evaluation. Of these, 26 toddlers were identified

only by the M-CHAT, 34 only by the GDS, and 24 by both of these tests (Kappa=0.52; p<0.001). Of these, ten toddlers had a confirmed diagnosis of ASD and another 30 toddlers were diagnosed with other forms of developmental delays (DD). Notably, the M-CHAT had a remarkably better sensitivity than the GDS in detecting toddlers with ASD (70.0% vs. 50.0% respectively), and slightly better specificity (98.2% vs. 96.6% respectively). These differences were translated to 2.3 fold difference in positive predictive value (PPV) between these two screening tools. Both methods had an equivalent efficiency in detecting other forms of DD (i.e. sensitivity= 63%; and specificity=98%). There was a remarkable variability between the nurses regarding toddler referral. However, a significant correlation was found between the total number of failed items and the number of referral nurses (Spearman r=0.65; P<0.001) with ≥5 failed items serving as the referral cutoff for the majority (≥6) of the nurses. No association was found between nurses’ referral and the age, sex or race of the toddlers.


Use of the M-CHAT is more effective in detecting children with ASD compared to the GDS approach. Thus, combining these two approaches in the developmental check-up at the MCHCs will improve early detection of ASD and other developmental problems.

See more of: Epidemiology
See more of: Epidemiology