Correlates of Treatment Outcome for Patients with BN & EDNOS

My previous post described the treatment outcomes of my 21 former patients with diagnoses of Bulimia Nervosa (BN) or Eating Disorder Not Otherwise Specified (EDNOS). Treatment was very successful for these patients, with 100% of those who completed treatment (n = 7) achieving full remission. However, the attrition rate was very high, with two-thirds of patients (n = 14) discontinuing treatment prematurely for various reasons. Nineteen percent (n = 4) of patients were referred to other providers or treatment settings because I was unable to meet their needs, 5% (n = 1) moved to another geographic area, and 43% (n = 9) quit before finishing treatment.

Given that treatment completion always led to full remission, attrition appears to be the biggest challenge for this population. Therefore, my post will focus on the differences between those who completed treatment and therefore achieved full remission (aka “treatment completers”) versus those who did not (“treatment non-completers”).

Neither age at intake nor duration of illness before intake predicted treatment outcome. Both treatment completers and treatment non-completers averaged approximately 19 years old at intake, with an average duration of illness of 4.5 years.

Patients who completed treatment and achieved full remission did so through a variety of treatment modalities: 43% (n = 3) received individual therapy with no family involvement, 29% (n = 2) received Family-Based Treatment (FBT), 14% (n = 1) received individual therapy with a high level of family involvement, and 14% (n = 1) received individual therapy with moderate family involvement.

Patients receiving FBT (n = 7) were significantly less likely than those receiving individual therapy to quit treatment prematurely. Only 14% of patients receiving FBT (n = 1) quit treatment prematurely. I referred 43% of them (n = 3) to other treatment settings due to my inability to help them progress in treatment. Twenty-nine percent (n = 2) completed treatment with me and 14% (n = 1) moved to another geographic area and continued her treatment there.

Of the 14 patients who received individual therapy, 57% (n = 8) quit treatment prematurely. The remainder of patients either completed treatment and achieved full remission (36%; n = 5) or were referred to other treatment settings (7%; n = 1).

Interestingly, treatment completers did not differ from treatment non-completers on duration of treatment or on number of sessions attended. For treatment completers, duration of treatment ranged from 1-38 months, with an average duration of 10 months. For treatment non-completers, duration of treatment ranged from 1-28 months, with an average duration of 10 months. Treatment completers attended an average of 15 sessions, while treatment non-completers attended an average of 20 sessions. Thus, it appears that treatment completers did not necessarily remain in treatment longer, but rather reached remission more quickly. Anecdotally, I recall that a number of patients who quit treatment prematurely actually did quite well in their first 6-9 months of treatment and nearly reached remission, but then began to regress, got discouraged, and quit.

Rates of treatment completion and treatment outcome differed based on diagnosis. Amongst patients with BN, 22% (n = 2) completed treatment and achieved full remission, while 44% (n = 4) made significant progress prior to discontinuing treatment. For patients with EDNOS, 42% completed treatment and achieved full remission, while 17% (n = 2) made significant progress prior to discontinuing treatment.

Prior history of untreated Anorexia Nervosa (AN) was associated with treatment dropout and poor outcome. Only 18% (n = 2) of those with a prior history of AN completed their BN/EDNOS treatment with me, and both of those patients had been successfully treated for AN years before. None of the patients with a prior history of untreated AN completed their treatment with me.

Patients with a co-morbid disorder were much less likely to complete treatment. While two-thirds of patients in this sample (n = 14) had a co-morbid disorder, only 29% of them (n = 4) completed treatment and achieved remission.

Taking psychotropic medication during treatment was not related to treatment completion. Thirty-six percent of those who took medication during treatment (n = 4) completed treatment and achieved full remission, compared with 30% (n = 3) of those who did not take medication during treatment.

This sample of patients was 71% Caucasian (n = 15), 24% Hispanic (n = 5), and 5% multi-racial (n = 1). Hispanic patients were much more likely to complete treatment and achieve remission than Caucasian patients. Eighty percent (n = 4) of Hispanic patients in this sample completed treatment and achieved remission, compared with only 20% (n = 3) of Caucasian patients.

I work on a sliding fee scale, offering reduced rates based on patient’s ability to pay. Those who paid my full rate were more likely to complete treatment and achieve remission compared with those who paid a reduced rate. Forty-six percent of those who paid my full rate completed treatment (n = 6), compared with 25% of those who paid a reduced rate (n = 2).

Individuals with a history of intensive eating disorder treatment were less likely to complete treatment than those without such a history. Amongst treatment non-completers, 50% (n = 7) had been hospitalized for their eating disorder or a related psychiatric issue, 21% (n = 3) had a history of residential treatment, and 21% had been in partial hospitalization, day treatment, or intensive outpatient programs. Amongst treatment completers, 29% (n = 2) had a prior history of hospitalization, 14% (n = 1) had a history of residential treatment, and none had a history of partial hospitalization, day treatment, or intensive outpatient treatment. The need for hospitalization during treatment with me was not related to treatment completion or treatment outcome.

In sum, the only variables that were related to treatment completion and remission were diagnosis of EDNOS and payment of my full fee for services. The following variables were associated with treatment non-completion: presence of a comorbid diagnosis, history of untreated AN, and prior history of intensive eating disorder treatment. Surprisingly, neither age at intake, duration of illness before intake, type of treatment received, number of sessions attended, duration of treatment, nor need for hospitalization during treatment were related to treatment completion.

These findings are specific to my practice and my patients, and are not intended to be generalized to other clinicians or other populations.

My next post will explore similarities and differences in treatment outcome among patients with AN, BN, and EDNOS.