Longitudinal follow-up of the development of social skills in patients with autism: a problem of reliability of the parental report in a telephone interview
Abstract
Objective: Factors influencing social development of autistic children have not been sufficiently studied yet. The aim of the study was to evaluate the longitudinal social development of children with autism spectrum disorders (ASD) and to verify the hypothesis of latent social skills in autistic children on longitudinal data.
Methods: The study was designed as a follow-up. All but two exceptions of these follow-up contacts were performed by a phone call.
The sample comprised of 47 autistic children (40 boys and 7 girls) with a mean age of 10.0+-2.5 years (range 5.0-15.8 years). The average follow-up interval was 4.1 years from the first examination.
The social domain of the Autism Diagnostic Interview-Revised (ADI-R) was used as an assessment tool. The sample was further subdivided according to the response to ADI-R Item No. 59 "Secure Base" to: (i) those with a normal response, i.e., score 0 (n=20); and (ii) those with an abnormal response, including the original untransformed scores of 1, 2, and 3 (n=27).
From the original research at baseline, parental reports' reliability scores from ADI-R raters, who were able to compare the parental report with their own knowledge of the child, were available. A school grading scale of 1-5 was used for the assessment, provided that the highest reliability was achieved by parents with grade 1.
Results: The decrease in the ADI-R social domain score during follow-up (i.e., compared to baseline) was significantly different between children with normal and abnormal response to the ADI-R item 59 (0.85 vs 5.26; p=0.046). However, if only reliable parental statements were included (grades 1-3), then the difference substantially changed and lost its statistical significance (2.57 vs 5.56; p=0.263).
Conclusions: Our results have not confirmed the hypothesis of latent social skills indicating that autistic children aged 4-5 year with a presence of separation anxiety, would later achieve better results in their social skills. On the contrary, the results pointed at the factor of reliability of parental reports that has not been explored yet.
This factor, if uncorrected, can significantly influence the results of telephone interviews.