A multidimensional meta-analysis of psychotherapy for bulimia nervosa

We report a multidimensional meta-analysis of psychotherapy trials for bulimia nervosa published between 1980 and 2000, including multiple variables in addition to effect size such as inclusion and exclusion, recovery, and sustained recovery rates. The data point to four conclusions. First, psychotherapy leads to large improvements from baseline. Approximately 40% of patients who complete treatment recover completely, although 60% maintain clinically significant posttreatment symptoms. Second, individual therapy shows substantially better effects than group therapy for the therapies tested. Third, additional approaches or treatment parameters (e.g., number of sessions) need to be tested for the substantial number of patients who enter treatment and do not recover. Finally, the utility of meta-analyses can be augmented by including a wider range of outcome metrics, such as recovery rates and posttreatment symptom levels.