Predicting Who May Drop Out of Inpatient Treatment for AN

Reprinted from Eating Disorders Review
November/December 2001 Volume 12, Number 6
©2001 Gürze Books

Patients with anorexia nervosa (AN) are often extremely resistant to outside intervention, and frequently refuse treatment. It is no surprise that many drop out of treatment early, and that as many as 50% of patients will have relapses.

In past studies, predictors of early dropout included: later age at onset of AN, older age when treatment began, less education, and lower socioeconomic status. Results of a recent study indicate that successful completion of inpatient treatment for AN may not be related to the severity of the disease (Int J Eat Disord 2002; 30:237).

Drs. Kathleen M. Pike and Caroline Kahn studied 81 severely underweight individuals with AN who received free inpatient treatment for AN at a clinical research unit. The patients were divided into those who were discharged before reaching their target weight of 90% of ideal body weight (IBW). All others were categorized as program completers.

The treatment dropout group was then divided into two subgroups, early and late dropouts, based on the percent IBW at discharge. Subjects discharged at or below 80% of IBW were categorized as early dropouts, while those discharged at or above 81% of IBW as late dropouts.

Study participants were assessed with clinical interviews, the EAT-26, the Body Shape Questionnaire, the Beck Depression Inventory, and the Symptom Checklist 90-Revised (a 90-item self-report measure designed to assess a wide range of symptoms). Other measures included the Inventory of Interpersonal Problems and the Rosenberg Self-Esteem Scale.

Group profile

The subjects were between 16 and 45 years of age, and all were severely underweight at admission. Mean BMI at admission was 15.1, at 69% of IBW. The mean duration of illness was nearly 8 years, during which the patients had been hospitalized for a mean of 1.9 times. They had first become ill with AN from 12 to 33 years of age. Thirty patients (37%) met criteria for restrictive AN and 51 (63%) met criteria for binge-purge subtype of AN. The mean hospital stay was 83 days.

Who dropped out, and why?

One-third of patients dropped out of treatment before reaching 90% of their IBW. Those who dropped out prematurely were slightly more likely to have a binge/purge subtype of AN (77.8%) than those who completed the program (55.6%). Compared with the late dropout group, the early dropout group had slightly more previous hospitalizations. No other differences were noted.

The overall dropout rate of 33.3% is substantially higher than the average rate of 16% reported in the general psychiatric literature. Inpatients with AN seem to be twice as likely to drop out of treatment than inpatients on a general psychiatric ward; the authors report this is not surprising given the well-known treatment resistance seen among many AN patients.

Early and late dropouts

The only remarkable difference between early and late dropouts was that those who dropped out in the earliest phase of treatment had more prior hospitalizations than those who dropped out later.

One striking finding was that patients who dropped out of treatment were hard to differentiate from patients who completed the program on all baseline measures except subtype. The absence of significant differences suggests that for persons with AN, compliance with treatment may be unrelated to illness severity and character disturbances.

Contrary to what Drs. Kahn and Pike expected, the more severely ill patients were not more likely to drop out of treatment early. One drawback to the study was that a continuous variable (% IBW at discharge) was artificially divided into group variables.

No Comments Yet

Comments are closed