Pino Maria Chiara, Masedu Francesco, Vagnetti Roberto, Attanasio Margherita, Di Giovanni Chiara, Valenti Marco, Mazza Monica
Department of Applied Clinical Sciences and Biotechnology, University of L'Aquila, L'Aquila, Italy.
Regional Centre for Autism, Abruzzo Region Health System, L'Aquila, Italy.
Front Psychol. 2020 Feb 5;11:4. doi: 10.3389/fpsyg.2020.00004. eCollection 2020.
The current study evaluated three social cognition (SC) tests for their clinical utility in aiding autism diagnosis. To do so, we compared the performance of 86 children with autism spectrum disorder (ASD) and 68 typically developing (TD) children, all aged from 4 to 10 years old, on three SC tasks [the Social Information Processing Interview (SIPI), the Comic Strip Task (CST), and the children's version of the Eyes Task] and calculated threshold scores that best differentiated the two groups. While difficulties in these abilities appear to represent the "central core" of ASD, services have largely ignored SC tests when supporting autism diagnoses. Therefore, this study attempted to validate and evaluate the diagnostic potential of these three tasks for children with ASD. To investigate the accuracy of these SC tests, we used the receiver operating characteristic (ROC) curve. As expected, the ASD group performed worse than the TD group on the SIPI and CST, but contrary to our prediction, the groups did not significantly differ on the Eyes Task. Specifically, the overall area under the curve (AUC) for the SIPI was 0.87, with a sensitivity of 73.5% and a specificity of 83.9% at the best cutoff point (score range 0-36; best cutoff = 31). The overall AUC for the CST was 0.75, with a sensitivity of 71.1% and a specificity of 77.0% at the best cutoff point (score range 0-15; best cutoff = 11). The overall AUC for the Eyes Task was 0.51, with a sensitivity of 50.3% and a specificity of 40.2% at the best cutoff point (score range 0-54; best cutoff = 45). In conclusion, the results showed that the SIPI test has good predictive power for classifying children with ASD. It should provide substantial supplementary clinical information and help to consolidate diagnostic procedures based on standard tools. Moreover, the results of the study have substantial implications for clinical practice: the better the knowledge of SC functioning in children with ASD, the more effective the intervention program for rehabilitation.
本研究评估了三项社会认知（SC）测试在辅助自闭症诊断方面的临床效用。为此，我们比较了86名自闭症谱系障碍（ASD）儿童和68名发育正常（TD）儿童（年龄均在4至10岁之间）在三项SC任务[社会信息处理访谈（SIPI）、连环漫画任务（CST）和儿童版眼睛任务]上的表现，并计算了能最佳区分两组儿童的阈值分数。虽然这些能力方面的困难似乎代表了ASD的“核心”，但在支持自闭症诊断时，服务机构在很大程度上忽略了SC测试。因此，本研究试图验证并评估这三项任务对ASD儿童的诊断潜力。为了研究这些SC测试的准确性，我们使用了受试者工作特征（ROC）曲线。正如预期的那样，ASD组在SIPI和CST上的表现比TD组差，但与我们的预测相反，两组在眼睛任务上没有显著差异。具体而言，SIPI的总体曲线下面积（AUC）为0.87，在最佳截断点（分数范围0 - 36；最佳截断值 = 31）时，灵敏度为73.5%，特异度为83.9%。CST的总体AUC为0.75，在最佳截断点（分数范围0 - 15；最佳截断值 = 11）时，灵敏度为71.1%，特异度为77.0%。眼睛任务的总体AUC为0.51，在最佳截断点（分数范围0 - 54；最佳截断值 = 45）时，灵敏度为50.3%，特异度为40.2%。总之，结果表明SIPI测试在对ASD儿童进行分类方面具有良好的预测能力。它应能提供大量补充临床信息，并有助于巩固基于标准工具的诊断程序。此外，该研究结果对临床实践具有重大意义：对ASD儿童的SC功能了解得越好，康复干预计划就越有效。
