Validation of the Arabic Version of the Social Communication Questionnaire (SCQ)

Background:  

There is a scarcity of validated autism screening and diagnostic tools for Arabic-speaking individuals which presents a major challenge to clinicians as well as researchers in many countries around the world. 

Objectives:  

The objective of the study is to evaluate the sensitivity and specificity of the Arabic Version of SCQ.

Methods:  

As part of the first country-wide Autism Prevalence Study that is ongoing in Qatar, we translated the Social Communication Questionnaire (SCQ) to Arabic and worked with Western Psychological Services (WPS) to have the back-translation reviewed by author-assigned reviewer. Multiple review cycles were needed prior to agreeing on the final version. The ASD sample was recruited from the Shafalah Autism Center in Qatar (N=35) and form 10 Autism Centers and classes in Saudi Arabia (N=97). The control sample were recruited from 8 schools in Qatar ( N=778 recruited by mailing SCQ forms to the families) and 10 schools in Saudi Arabia (N=247 recruited by distributing forms to schools by two of the authors). The study was carried between February and September 2015. We evaluated the sensitivity and specificity of the Arabic SCQ using the published cut-off value of 15.

Results:  

The pilot sample included 1060  ( 132 ASD and 928 Control). We are continuing to recruit more children.  The boy to girl ratio was 2.14:1 (92/42; 68.2% male) in the ASD group. In the control group, the corresponding values were 0.60:1 (346/579; 37.3% male). The mean SCQ total score was significantly higher in cases as compared to controls (20.79 (SD=6.4) vs 7.26 (SD=5.0); p<.0001). As expected, the variability was larger in cases than in controls as illustrated by the standard deviations. A total of  184 children (109 cases, 75 controls) had scores equal or above the cut-off of 15; the remaining 23 cases (17.4% of the cases) had scores below the cut-off. An ROC analysis was performed to examine the overall performance of the SCQ. The area under the curve showed excellent discrimination between cases and controls (AUC=0.937; 95%CI: .915-.959).  For the established cut-off of 15, sensitivity and specificity were of .826 and 0.919, respectively.  Adjacent values for the cut-off were associated with slightly lower overall performance as defined by the sum of the sensitivity and 1-specificity (Yougen index).

Conclusions:

The findings from this study validates the use of the Arabic SCQ as screening instrument with the original published cut-off values.