Development of the Ottawa Disordered Eating Screen for Youth: The ODES-Y

Abstract

To develop a concise screening tool that allows for early identification of disordered eating in youth. In this 2-step classification accuracy study, questions for the Ottawa Disordered Eating Screen-Youth, a 2-question screening tool (index test), were conceptualized by clinician-scientists from tertiary care pediatric eating disorder and weight-related clinics, and was validated using retrospective data (2004-2010) from a community-based study, the Research on Eating and Adolescent Lifestyles (REAL) study. Analyses of contrast between the index test and the reference standard using data from 2892 (1714 females) students between grade 7 and grade 12 revealed classification statistics of 67.1% for sensitivity, 85.9% for specificity, 4.7 for positive likelihood ratio, 0.38 for negative likelihood ratio, 50.6% for positive predictive value, and 92.4% for negative predictive value for females and 61.1% for sensitivity, 93.9% for specificity, and 9.9 for positive likelihood ratio, 0.41 for negative likelihood ratio, 32.3% for positive predictive value, and 98.0% for negative predictive value for males. Our findings suggest that the index test has utility as a short and accurate screening tool for earlier detection of disordered eating thoughts and behaviors in youth. Additional research is needed to best determine how the index test can be administered to youth across various health care, school, public health, and surveillance settings in clinically sensitive pragmatic ways.