518 - Early ASD-screening using M-CHAT in a neonatal high-risk group

Publication Number: 518.4640

Benjamin Lassebro, Department of Biomedical and Clinical Sciences (BKV), Linkoping University, Linköping, Ostergotlands Lan, Sweden; Matilda Morin, Karolinska Institutet, Stockholm, Stockholms Lan, Sweden; Weiyao Yin, Karolinska Institutet, Stockholm, Stockholms Lan, Sweden; Sven Sandin, Icahn School of Medicine at Mount Sinai, New York, NY, United States; ulrika Åden, Karolinska and Linköping, Stockholm, Stockholms Lan, Sweden

Background: Autism Spectrum Disorder (ASD) is a chronic neurodevelopmental condition with both genetic and environmental origins. Early detection has been shown key for successful clinical intervention, improving future health and adaptive functioning for the individuals effected. Being born preterm or having severe illness at birth is associated with higher risk for developing ASD. In Sweden, children being born extremely preterm or having severe illness, called neonatal high-risk children, at birth are routinely assessed for developmental delays such as ASD with the Modified Checklist for Autism in Toddles, M-CHAT. The predictive validity of M-CHAT varies between countries and study populations. Extremely preterm-born children have a high prevalence of positive M-CHAT results (indicating further assessment is needed). The usefulness of the M-CHAT as a screening instrument for the group of neonatal at-risk children should thus be evaluated specifically within the context of neonatal follow-up.

Objective: This study aims to evaluate the predictive value of M-CHAT as a screening instrument for ASD in a high-risk group of Swedish-born children. A secondary aim is to identify divergent subgroups, positive or negative, by examining risk factors potentially influencing the predictive value of M-CHAT.

Design/Methods: All individuals born alive in Sweden between 2013 and 2019 and with a registered M-CHAT result at 16-30 month of adjusted age in the Swedish National Quality registry for neonatal care, SNQ. We will use data on M-CHAT screens, from the SNQ. We will compare screening results on the M-CHAT at the follow-up visit at 24 months of age with presence or absence of a diagnosis of ASD or other neurodevelopmental disorders in the National Patient Register, NPR.

Results: A total of 4705 children born between 2013 to 2019 were registered for neonatal follow-up in the SNQ. Out of these, 2178 children were screened with the M-CHAT. The overall sensitivity of M-CHAT in the study population was 62.4% and the specificity was 91.2%. When predictive validity was estimated in each neonatal high-risk group separately, the proportion of children with later ASD correctly identified by M-CHAT was highest among extremely premature infants with a sensitivity of 69.4%.

Conclusion(s): Our study adds to the understanding of how the M-CHAT performs in different types of populations. Our conclusion is that the predictive values of M-CHAT in the neonatal high-risk population are acceptable but should be interpreted with caution, in conjunction with the clinical team assessment conducted with a psychologist and physician.

Table 1 Subject characteristics
Subject characteristics for individuals with positive and negative m-chat, and total

Table 2 Cross-tabulation of M-CHAT outcomes and ASD diagnoses
Cross-tabulation of M-CHAT outcomes and ASD diagnoses in neonatal high-risk children, in total, by sex, and by reason for inclusion in neonatal follow-up

Table 3 Sensitivity and Specificity of M-CHAT
Sensitivity and Specificity of M-CHAT overall, by sex and by neonatal risk group

Table 1 Subject characteristics
Subject characteristics for individuals with positive and negative m-chat, and total

Table 2 Cross-tabulation of M-CHAT outcomes and ASD diagnoses
Cross-tabulation of M-CHAT outcomes and ASD diagnoses in neonatal high-risk children, in total, by sex, and by reason for inclusion in neonatal follow-up

Table 3 Sensitivity and Specificity of M-CHAT
Sensitivity and Specificity of M-CHAT overall, by sex and by neonatal risk group