Multi-domain assessment of childhood attention-deficit/hyperactivity disorder of the combined presentation

Screening children for Attention-Deficit Hyperactivity Disorder (AD/HD) often relies upon subjective behaviour ratings which are susceptible to biases. A brief (35-40 min) post-screening assessment that does not directly address symptoms, increases objectivity, and accurately discriminates children with AD/HD from typically developing peers, could serve as an adjunct to symptom-focussed screening to support decisions about referral for comprehensive specialist assessment. This study examined the classification performance of the subjective parent-reported SNAP-IV screening questionnaire and the multi-domain Neurocognitive Assessment Tool (NCAT), both alone and in combination, in children with and without AD/HD of the Combined presentation (AD/HD-C). The SNAP-IV and NCAT were administered to 46 children with AD/HD-C and 138 typically developing children. The NCAT assessed executive function performance, brain electrical activity, quality of life, psychological needs satisfaction, sleep habits, self-regulation and social behaviour. Discriminant function and receiver operating characteristic analyses evaluated classification performance for SNAP-IV, NCAT, and combined (NCAT+) models. The SNAP-IV (97.8% accuracy; 91.3% sensitivity; 100% specificity; AUC =.990) and NCAT+ (95.1%; 87.0%; 97.8%; AUC =.993) demonstrated high classification performance; the NCAT showed lower but robust performance (88.0%; 78.3%; 91.3%; AUC =.966). The SNAP-IV showed the highest classification performance based on subjective ratings of symptoms, which are susceptible to a range of biases. The NCAT exhibited strong classification, particularly in specificity, highlighting its capacity to discriminate children with AD/HD-C from typically developing children using objective measures and non-symptom-specific subjective assessments. Combining the SNAP-IV and NCAT yielded superior performance to NCAT alone, supporting integrated use as screening and decision-support tools to identify children warranting comprehensive specialist assessment. External validation in diverse clinical populations, including those with co-occurring diagnoses, is needed.