Psychometric evaluation of SAFA P test for eating disorders in adolescents: comparative validation with EDI-2

This study evaluates the psychometric properties of self-administered psychiatric scale for children and adolescents with psychogenic eating disorders (SAFA P)--a brief self-report designed to screen and assess eating disorders (ED) in children and adolescents. Although SAFA P belongs to a broad battery of tests (SAFA) that explores different psychiatric conditions, it has not undergone appropriate validation until now.

We administered SAFA P and Eating Disorder Inventory 2 (EDI-2) to 87 ED patients, with an average age of 15.4 ± 1.6 years.

The internal reliability of SAFA P is good (Cronbach α = .776). Convergent validity with EDI-2 was assessed: both SAFA P subscale P1 (p < .005) and EDI-2 subscale bulimia (p < .001) showed a statistically significant difference among the three diagnostic categories (anorexia nervosa, bulimia nervosa and eating disorder not otherwise specified). Sensibility and specificity range from 62 to 91%, depending on the subscales. McNemar's test did not reveal statistically significant differences in assessing the concordance of the two measures. Statistically significant correlations were found between specific couples of subscales (p < .001).

Cross-validation with EDI-2 showed good results. SAFA P may be an alternative, useful and reliable instrument for assessing cursory ED in childhood and adolescence.