230 - Agreement Between Parent Report and Clinician Observation of Language at Time of Autism Spectrum Disorder Diagnosis

Publication Number: 4227.230

Lianna Lipton, Boston Children's Hospital, Brookline, MA, United States; Maya J. Golden, Boston Children's Hospital, Boston, MA, United States; Stephen Camarata, Vanderbilt University, Nashvile, TN, United States; William Barbaresi, Harvard Medical School; Boston Children's Hospital, Boston, MA, United States; Elizabeth Harstad, Division of Developmental Medicine, Boston Children's Hospital, Boston, MA, United States

Background: Assessment of language development often relies on parent report; however, there is limited literature about the validity of parent report of children’s language, particularly in children with neurodevelopmental disorders such as Autism Spectrum Disorder. There are challenges with existing parent-report screening and assessment tools, with accuracy of parent report regarding children’s receptive language (understanding) of particular concern.

Objective: To compare parent-reported and clinician-observed language at time of clinical developmental evaluation for children diagnosed with Autism Spectrum Disorder as toddlers.

Design/Methods: We conducted a retrospective chart review of 135 medical records of patients diagnosed with Autism Spectrum Disorder between 12 and 36 months of age. All children had a multidisciplinary evaluation with a developmental-behavioral pediatrician, who completed a history, and a psychologist, who completed behavioral observations and testing. We abstracted data on parent-report and psychologists’ observations of children’s expressive language, receptive language, and nonverbal/social communication. A second coder abstracted 10% of records for reliability, and inter-rater reliability was >90%. We used Spearman’s Rank Correlation to assess concordance between parent-report and psychologists’ observations as variables had 3 or more levels.

Results: Parent report and psychologist observations of children’s expressive language were moderately to strongly correlated: r = 0.689, p < 0.001, while receptive language (following verbal directions) was weakly to moderately correlated at r = 0.393, p < 0.001 (Table). Concordance was lower for nonverbal/social communication including response to name (r = 0.261, p = 0.007) and eye contact (r = -0.129, p = 0.185).

Conclusion(s): Parent report of expressive language was more strongly correlated with psychologists’ observations than either receptive language or nonverbal communication. These findings highlight the importance of combining multiple sources of information to understand a child’s communication status.

Concordance Between Parent Report and Psychologist Observation of Expressive and Receptive Language
PAS_Table.pdf