RESUMO
OBJECTIVE: To determine the internal consistency, factorial structure, reproducibility and validity of criterion of the Leyton Obsessional Inventory in children and adolescents. DESIGN: Study of validation with probabilistic sampling. SETTING: Secondary schools in Bucaramanga, Colombia. PARTICIPANTS: A total of 581 students selected at random, of whom 38 did not answer, 8 were excluded because they were over 17, and 33 did not attend the clinical interview. MAIN MEASUREMENTS: The Inventory and the Structured Clinical Interview for axis I of the DSM-IV (SCID-I), clinical version, were used. Cronbach's alpha, factorial analysis, sensibility, specificity, probability, and Lin' s coefficient were calculated. RESULTS: The prevalence of obsessive-compulsive disorder was 11.8% (95% CI, 8.9-14.6). Cronbach's alpha was 0.803 and just one factor with 3.60 eigenvalue explained 74.8% of variance. The area under the ROC curve was 0.799 (95% CI, 0.752-0.847) and the best cut-off point was 12, with sensitivity at 69.5% (95% CI, 56.0-80.5) and specificity at 77.7% (95% CI, 72.9-88.9). Its reproducibility was acceptable (Lin's coefficient at 0.752; 95% CI, 0.714-0.790). CONCLUSION: The Leyton Obsessional Inventory is a fair screening tool for detecting obsessive-compulsive disorder in Colombian children and adolescents.
