Validation of the Spanish version of the SCOFF questionnaire for the screening of eating disorders in primary care

Abstract

Objective To assess the performance of the Spanish version of a new screening tool (the SCOFF) for the detection of eating disorders (EDs) in primary care. Methods Design: validation study. The psychiatric interview Schedules for Clinical Assessment in Neuropsychiatry (SCAN) was used as the “gold standard.” Blinding was applied to administration of the SCOFF and the clinical interview. Setting: six primary health care centers in Zaragoza and Huesca, Spain. Participants: 203 female patients, between the ages of 15 and 53 years, with a probable diagnosis of ED. Main outcome measure: validity as assessed by sensitivity, specificity, positive predictive value and negative predictive value. In addition, test–retest reliability and the receiver operating characteristics (ROC) curve were calculated. Results The best threshold point in the Spanish version was 2+ positive answers, the same as that recommended in the original British study. This cutoff point provided 97.7% sensitivity and 94.4% specificity for the detection of EDs in primary care. For each specific ED, sensitivity and specificity with this threshold point were, respectively, the following: bulimia, 97.8% and 94.4%; anorexia, 93.1% and 94.4%; and ED not otherwise specified (EDNOS), 100% and 94.4%. Conclusions The Spanish version of the SCOFF questionnaire shows excellent psychometric properties for the early detection of EDs in primary settings.