Factors affecting dropout rate from cognitive-behavioral group treatment for bulimia nervosa

Abstract

The aim of this study was to retrospectively identify clinical variables assessed prior to treatment which were predictive of patients' dropping out versus completing a 10-week group cognitive-behavioral treatment program for bulimia nervosa. Following a lengthy initial assessment, 81 women meeting DSM-Ill-R criteria for bulimia nervosa (BN) were referred to one of twelve 10-week groups of 8 to 12 patients having bulimic symptoms. The dropout rate for those meeting full DSM-Ill-R criteria for BN was found to be 28.7%. A series of seven discriminant function analyses were performed to determine whether dropouts differed from completers in terms of depression, anxiety, difficulties in trust and relating to others, bulimic symptom severity, family environment, weight history and symptom duration, and severity of bulimic cognitions. Of these, only the factor assessing difficulties trusting and relating to others was found to significantly discriminate dropouts from completers. Implications of the findings are discussed in terms of clinical and research relevance in the field of eating disorders. 0

Ten-year stability and predictive validity of five bulimia-related indicators

“…Earlier studies (11)(12)(13) showed support for the Eating Disorders Inventory interpersonal distrust subscale as a prognostic indicator. While this variable may predict treatment response over the short term, our findings do not support it as a long-term predictor of later bulimic symptoms.…”

“…Interpersonal variables may also be useful prognostic indices: Steiger et al (9) found support for pretreatment social adaptation, Blouin et al (10) reported positive findings for family environment, and Blouin et al (11), Olmsted et al (12), and Sohlberg et al (13) found that interpersonal distrust was a significant prognostic indicator. Several other variables have been identified as prognostic indices, for example, attitudes related to weight and shape (14), low self-esteem (14-16), comorbid depressive symptoms (17,18), and low levels of social support (13).…”

Selection practices of group leaders: A national survey

“…However, they did find significant differences between dropouts and completers on amount of participation as rated by group leaders and on client ratings on whether they felt able t o express themselves in the group. In another study, Blouin et al (1995) found that only one factor-concern with difficulties trusting and relating to others-significantly discriminated dropouts from completers in group cognitive behavioral treatment for bulimia nervosa. Similar results were found in a study comparing dropouts and completers in open-ended interpersonal therapy groups (MacNair & Corazzini, 1994).…”

Do broken relationships in childhood relate to bulimic women breaking off psychotherapy in adulthood?

“…Dropouts may be more likely to abuse laxatives (Herzog, Keller, Lavori, & Sacks, 1991;Weltzin, Bulik, McConaha, & Kaye, 1995). Where differences in personality measures have been found, bulimic dropouts have been more likely to meet diagnostic criteria for borderline personality disorder (Coker et al, 1993;Waller, 1997), to be more hostile (Lee & Rush, 1986), to have diminished capacity for trust (Blouin et al, 1995), and to have lower self-esteem (Coker et al, 1993). These results indicate that younger patients with difficulty making and maintaining relationships may be more likely to drop out.…”

“…Attitudes of patients to forming a trusting relationship with a therapist would be an important part of this transaction. It was hypothesized that patients' early family experience, more than their current family environment (Waller, 1997;Blouin et al, 1995), would be relevant. Therefore, childhood abuse and parental loss were considered.…”