Attention-deficit hyperactivity disorder: agreement between clinical impression and the SNAP-IV screening tool.

INTRODUCTION The CONCOR study aimed to determine the predictive value of the SNAP-IV screening tool for ADHD relative to the pediatricians' clinical global impression of the presence of ADHD in children attending their practice. METHODS In this multicentre, cross-sectional, case-control study participated pediatricians from all over Spain and each pediatrician enrolled at least six children over the age of 6 years who had no previous diagnosis of ADHD, three whom they considered to have symptoms characteristic of ADHD (DSM-IV criteria) and three children who did not. Screening with the SNAP-IV tool was completed by parents in the waiting room prior to a consultation (for any reason) and results were compared with the pediatrician's clinical global impression. RESULTS Results for 7263 Spanish children (>6 years of age) showed 'good' concordance between the SNAP-IV screening tool and the pediatricians' clinical impression of ADHD (kappa concordance index 0.6471; 95% confidence intervals: 0.6296-0.6646), with acceptable sensitivity and specificity (82.3% and 82.4%, respectively). Based on an estimated ADHD prevalence of 3-10%, the negative predictive values of SNAP-IV were 97.7%-99.3% and positive predictive values were 12.6-34.2%. CONCLUSIONS SNAP-IV is a useful screening tool that identifies children who are unlikely to have ADHD, and brings those in whom there is a high possibility of ADHD to the attention of the pediatrician for clinical evaluation.