We aimed to quantify the clinical utility of continuous performance tests (CPTs) for the diagnosis of attention-deficit/hyperactivity disorder (ADHD) compared to a clinical diagnosis in children and adolescents.
Four databases (MEDLINE, PsycINFO, EMBASE, and PubMed) were screened until January 2023. Risk of bias of included results was judged with the QUADAS-2. We statistically pooled the area under the curve, the sensitivity, and the specificity of three commonly used CPTs subscales: omission/inattention, commission/impulsivity, and total number of errors/ADHD subscales (PROSPERO registration: CRD42020168091).
19 studies using commercially available CPTs were identified. Results from up to 835 control individuals and 819 cases were combined in the summary receiver operating characteristic (ROC) curve analyses (sensitivity and specificity pooling), and up to 996 cases and 1083 control individuals in the area under the curve (AUC) analyses. Clinical utility as measured by AUCs could be considered as barely acceptable (between 0.7 and 0.8) for the most part, with the best results for the total/ADHD score, followed by omissions/inattention, and poorest for commission/impulsivity scores. A similar pattern was found when pooling sensitivity and specificity: 0.75 (95% CI=0.66 to 0.82) and 0.71 (0.62 to 0.78) for the total/ADHD score; 0.63 (=0.49 to 0.75) and 0.74 (0. 65 to .81) for omissions; and 0.59 (0.38 to 0.77) and .66 (CI=.50, .78) for commissions.
At the clinical level, CPTs as a stand-alone tool have only a modest to moderate ability to differentiate ADHD from non-ADHD samples. Hence, they should only be used within a more comprehensive diagnostic process.
Copyright © 2023. Published by Elsevier Inc.
