Supplementary file 1_Multi-method ADHD diagnostics in children: CBCL and TRF lead the way.docx

ObjectiveDiagnosing ADHD in children requires multi-method examinations, yet the efficacy of this approach remains inconsistent. Our case control study investigated which standardized assessment methods most accurately predict an ICD-10 ADHD diagnosis in mixed settings (inpatient, outpatient, school).
MethodsWe examined 125 children aged 6–13 years, thereof 56 with an ADHD diagnosis (M = 9.7, SD = 2.09) and 69 without (M = 9.04, SD = 2.05). Our assessment included a children’s self-report questionnaire (Youth Self Report 11-18R, YSR), which was exploratorily used for younger children, and two objective tests, a Gameboy-administered Go/No-Go-task (QIKtest, 1) and a PC-administered continuous performance test (CPT, 2). Parents were asked to complete some questions on the child’s possible diagnoses and medication, and a parent questionnaire (Child Behavior Checklist 6-18R, CBCL). Teachers received the same questionnaire in adapted form (Teacher’s Report Form 6-18R, TRF). Classification accuracy was determined using receiver operating characteristic (ROC) analyses (Sensitivity, Specificity, Area under the Curve and Diagnostic Odds Ratio). A stepwise combination of indices was used to explore a multi-method procedure and its diagnostic accuracy.
ResultsCBCL and TRF achieved the highest classification accuracy focusing on inattention, followed by the YSR. Omission errors of CPT and QIKtest showed moderate classification performance while commission errors achieved the lowest. Combining CBCL, TRF and YSR showed superior diagnostic accuracy.
ConclusionsOur results emphasize the relevance of multi-perspective questionnaire procedures for ADHD diagnosis despite potential acquisition challenges in clinical practice. Future research should develop more accurate objective test procedures and norm-based scales for children’s self-reports.

研究目的：儿童注意缺陷多动障碍（ADHD）的临床诊断需采用多方法检查，但该策略的诊断效能仍存在不一致性。本病例对照研究旨在探究，在住院、门诊及学校等混合场景中，哪些标准化评估方法能够最精准地预测符合国际疾病分类第10版（ICD-10）标准的ADHD诊断。

研究方法：本研究纳入125名6~13岁的儿童，其中56名确诊ADHD（平均年龄9.7岁，标准差2.09），69名未确诊ADHD（平均年龄9.04岁，标准差2.05）。评估工具包括儿童自评问卷——青少年自评量表11-18R（Youth Self Report 11-18R，YSR），本研究对低龄儿童探索性使用该量表；以及两项客观测试：一项为任天堂Gameboy平台施测的Go/No-Go任务（QIKtest，1），另一项为个人电脑（PC）平台施测的持续性操作测试（CPT，2）。研究要求家长完成关于儿童疑似诊断与用药情况的相关问题，以及儿童行为量表6-18R（Child Behavior Checklist 6-18R，CBCL）；教师则填写适配版本的教师报告量表6-18R（Teacher’s Report Form 6-18R，TRF）。本研究采用受试者工作特征（Receiver Operating Characteristic，ROC）曲线分析评估分类准确率，涉及指标包括灵敏度、特异度、曲线下面积及诊断比值比。通过逐步组合各评估指标，探究多方法评估流程的诊断准确率。

研究结果：在注意力缺陷维度上，CBCL与TRF的分类准确率最高，其次为YSR。CPT与QIKtest的漏报错误呈现中等分类效能，而虚报错误的分类效能最低。联合使用CBCL、TRF与YSR时，诊断准确率最优。

研究结论：本研究结果证实，尽管临床实践中多维度问卷评估流程存在数据收集挑战，但该方法在ADHD诊断中具有重要应用价值。未来研究应开发更为精准的客观测试流程，并针对儿童自评量表制定基于常模的评分标准。