The evolution of autism research: A study assessing the use of calibrated severity scores and the social responsiveness scale with implications for future data collections

The purpose of this study was to analyze two instruments used in a large scale Autism genetics consortium. One instrument yields a Calibrated Severity Score (CSS) and involves a structured assessment that must be completed by a highly skilled and trained professional. The next instrument, Social Responsiveness Scale (SRS), is a questionnaire that can be filled out by a parent or teacher. Both yield scores associated with autism or autism risk. These two instruments were analyzed in regards to current and proposed ASD diagnostic criteria and compared with each other in order to determine if they can be used interchangeably in the research setting. Participants included a total of 312 families who have participated in the Simons Simplex Collection at Vanderbilt University. Archived data were analyzed using one way Multiple Analyses of Variance and two-way contingency table analysis. Results from this study indicate there were significant differences in CSS among the current diagnostic categories on the autism spectrum (including those who are not on the spectrum) as well as the proposed diagnostic categories of ASD or non-ASD. Additionally, results of this study indicate that there was not a significant difference in SRS scores among the current and proposed diagnostics categories. Results from a two-way contingency table analysis indicate that the current autism/autism risk cutoff on the Calibrated Severity Score (≥ 4) significantly differentiates children on the spectrum from those who are not on the spectrum according to a Clinician\u27s Best Estimate (CBE), while a proposed autism/autism risk cutoff score on the Social Responsiveness Scale (≥ 65) does not significantly differentiate children on the spectrum from those not on the spectrum. These findings suggest that the CSS and the SRS cannot be used interchangeably in future genetic data collections. Furthermore, results indicate that further research is needed regarding the sensitivity and specificity of the SRS