نتیجه رها کردن و نتیجه درمان سرپایی درمان شناختی رفتاری برای بی اشتهایی عصبی و بولیمیا

In the present study, drop-out-analyses were carried out for a manual-based cognitive–behavioral therapy for 104 females with anorexia nervosa (AN) and bulimia nervosa (BN), in the service setting of a university outpatient clinic (naturalistic setting). A total of 22.9% of patients with AN terminated therapy prematurely (drop-outs), compared to 40.6% of patients with BN. Group differences between drop-outs and completers show that the group of drop-outs with BN had higher values in the depression score at the start of therapy and was almost two times more likely to have a comorbid disorder (odds ratio 1.69), whereas drop-outs with AN had higher values in the outcome-scale drive for thinness and the odds ratio for being employed or living in a partnership was slightly lower. Completers and drop-outs did not differ significantly within groups in regard to age, body mass index at the start and end of therapy, or the number of comorbid disorders. On the whole, the therapy effect in the group of drop-outs was relatively moderate. For patients with AN, even higher therapy effects were observed among the drop-outs than among the completers. These data suggest that moderate therapy effects and responses can be achieved even among the drop-outs.

Current research on therapy for the treatment of eating disorders has demonstrated that premature therapy termination is a common problem. Previous randomized controlled studies (RCTs) have shown that patients with a primary diagnosis of an eating disorder have a more than twofold higher drop-out rate (e.g. Ref. [1]), in comparison to patients with other mental disorders. In cognitive–behavioral therapy, (CBT) termination rates of up to 57% have often been reported [2]. As the greater part of outpatient psychotherapeutic treatments takes place in the office settings of routine health-care services (naturalistic setting) and not within the framework of controlled research studies, the question arises as to the size of the drop-out rate in routine settings. This is of particular relevance because these naturalistic studies are not subject to the strict inclusion criteria that are employed in randomized controlled trials (e.g. Ref. [3]). To date, only a few studies have been published on the treatment of eating disorders in a naturalistic setting (e.g. Refs. [4], [5] and [6]), which have shown that good to very good effect sizes for the various eating disorder diagnoses can also be achieved in a naturalistic setting. However, therapy drop-outs have mostly not been investigated separately in these studies. Tuschen-Caffier et al. [4] and [5] studied a sample of non-selected patients (n = 67) who participated in a 10-week cognitive–behavioral intensive program (2 weeks with daily sessions of 8 h, then 6 to 8 further weeks with one session per week). Only one person (1.5%) dropped out before completion of the study program. RCTs with adult patients who have the primary diagnosis of anorexia nervosa (AN), report different drop-out rates. For example, in a comparative study of three different outpatient treatment procedures (CBT, interpersonal psychotherapy, non-specific supportive sessions) with 20 sessions, McIntosh et al. [7] observed premature terminations in 37.5% of the patients after an average of seven sessions. With regard to possible differences between subjects who completed therapy (completers) and those who terminated prematurely (drop-outs), the authors stated that the two groups differed only in body mass index (BMI) at start of the therapy, with the drop-outs having a significantly lower weight than the completers. The following reasons were given for dropping out: hospitalization necessary due to loss of weight, moving away, rejection of therapy, early improvement, and non-appearance. In another outpatient setting, Pike et al. [8] found a drop-out rate of 20% in the group of patients with a primary diagnosis of AN, who were undergoing CBT. In a parallel patient group within this trial, who only received nutritional advice, the drop-out rate was 53%. Halmi et al. [2] carried out a comparative study of patients with a diagnosis of AN using three treatment modalities: (1) CBT, 1-year duration, with 37 sessions; (2) CBT plus drug treatment with fluoxetin; and (3) Only drug treatment with fluoxetin. In the CBT group, 57% of the participants dropped out before completion of the regular program. In the other two groups the drop-out rates were even higher. In comparison to these results from outpatient settings, the drop-out rates of patients with the primary diagnosis of AN in an inpatient setting, ranged from 20.2% to 49.6% [9]. Here, widely different reasons were given for dropping out, such as a higher BMI at the start of therapy or, in another study, a lower BMI or binge-eating/purging behaviors were identified as risk factors of a high drop-out probability [9]. Tagay et al. [10] also identified loss of weight in the first half of the treatment as a possible predictor for premature termination of therapy in the inpatient treatment of AN. Woodside et al. [11] found a relationship between higher levels of depressive symptoms at the start of therapy and therapy termination in patients with a diagnosis of AN, in an inpatient setting. This observation is in contradiction to the results of Wallier et al. [9]. In a compilation of RCTs of patients with the diagnosis AN, the authors describe the lack of a comorbid depressive disorder increases the probability of therapy termination. For female patients with the diagnosis of bulimia nervosa (BN), and AN of the bulimic subtype, undergoing CBT in an outpatient setting (n = 50), Waller [12] determined a drop-out rate of 30%. Reasons for the premature termination were found to be more pronounced bulimic symptoms, based on the bulimic investigatory test (BITE [13]) and increased borderline symptoms, as measured with the borderline syndrome index (BSI; [14]). In a study of patients with the primary diagnosis BN, in which the therapeutic effects of CBT and interpersonal psychotherapy (IPT) were investigated in an outpatient setting, with 19 individual sessions in each, Agras et al. [15] found drop-out rates of 28% in the group receiving CBT and 24% in the IPT group (initial sample sizes of n = 110 each). In another therapeutic study of CBT for patients with BN in an outpatient setting, Agras et al. [16] found a drop-out rate of 26% after an average of 4.6 weeks. At the start of their therapy, the drop-outs showed a more pronounced bulimic cognition (no higher frequency of binge eating) and a more pronounced impulsivity. Agras et al. [16] could not determine any further clinical predictors. In addition, the authors examined predictors for response and non-response at the end of the therapy. The non-responders exhibited a lower social adaptation and a lower BMI, accompanied with a higher drive for thinness. On the whole, it seems that an early improvement of symptoms is the best predictor for therapeutic success [16] and [17]. A further predictor of therapy outcome seems to be the level of perfectionism at the start of therapy. Several different studies have shown a relationship between higher levels of perfectionism and increased eating disorder symptoms, as well as a poorer therapy result (e.g. Refs. [18] and [19]). No details were given in these studies about whether or not values for the outcome measure may possibly have become poorer during the course of the therapy. In this context it needs to be mentioned that the research into deterioration during behavioral therapy is unsatisfactory and this field has so far only been sparsely explored [20]. Our literature search revealed that there are currently no RCTs available that have analyzed deteriorations during cognitive–behavioral therapy in particular, for the two eating disorder diagnoses AN and BN. In sum, it can be seen that the current data situation on drop-out rates in eating disorders originates mainly from RCTs. The findings so far have been inconsistent and stable predictors to indicate therapy termination are lacking. In addition, the above-mentioned studies do not demonstrate to what extent therapy termination affects the therapeutic results, or if a therapy termination is even accompanied by a deterioration of the symptoms. These findings led to the three main research aims of the present study. Firstly, we were interested in finding out what proportion of subjects with an eating disorder (AN, BN) prematurely terminated their participation, in a naturalistic therapeutic study (outpatient routine health-care service). Secondly, it was also of interest to find out how subjects who regularly completed the therapy (completers) differed from those who terminated it prematurely (drop-outs). Thirdly, it was also of interest to find out what effect the therapy had on both the completers and drop-outs. In particular, it was of interest as to whether there were subjects whose values deteriorated in the course of therapy and if the proportion of deteriorations would be higher among the drop-outs.