Outlining patient characteristics that predict the outcome of psychotherapy for bulimia nervosa has been an elusive goal. If patients who will or will not respond to psychotherapy could be identified, treatment could be reserved for those who will respond. If those who are not likely to respond to psychotherapy can be identified early, other treatment methods may be more effective. The most effective therapy for patients with bulimia nervosa is cognitive behavior therapy. Even though it is the most effective therapy, only about 50 percent of bulimic patients treated this way recover. Because of this low response rate, it is important to identify patients who will not respond to cognitive behavior therapy to allow an early move to another treatment strategy. Agras and colleagues studied useful predictors of attrition and the outcome of cognitive behavior therapy in patients with bulimia nervosa.

Women who met the criteria for bulimia nervosa in the Diagnostic and Statistical Manual of Mental Disorders-Revised, 3d ed. (DSM-III-R) were enrolled in the study. The pretreatment, course and outcome data were recorded for these patients. All were given 18 sessions of manual-based cognitive behavior therapy. The differences between dropouts and nondropouts and recovered and nonrecovered patients were examined descriptively. Signal detection analyses were used to evaluate clinical cutoff points predicting attrition and abstinence.

Of the 140 patients who completed the treatment course, 41 percent had stopped binge eating or purging according to the Eating Disorder Examination. Dropouts were more likely to have severe bulimic cognition and greater impulsivity, but these factors were not useful as clinical predictors. Poor social adjustment and lower body mass index characterized patients in whom treatment failed. The signal detection analysis identified poor outcomes in patients who had not reduced purging by 70 percent or more by the sixth treatment session. Using this cutoff provided a method of identifying patients who were prospective failures of cognitive behavior therapy.

The authors conclude that patients with bulimia nervosa who have not significantly reduced their purging behavior by the sixth session of cognitive behavior therapy are not likely to respond to this treatment. These patients should be transferred to a second-line treatment.