Can we study (treat) borderline patients? Attrition from research and open treatment

Predictors of attrition from research and follow-up treatment were studied in 97 borderline patients participating in a 22-week double-blind placebo-controlled trial of phenelzine and haloperidol. Patient characteristics, baseline symptom severity, response to treatment, and prior and current treatment variables were compared between treatment “completers” and “dropouts.” Overall attrition was 46.4%. Premature termination was related to comorbid drug use disorder and baseline measures of anger, observed hostility, and behavioral impulsiveness. Completers were more depressed at baseline and showed significantly more global improvement with treatment. Length of research hospitalization was predictive of completion whereas severity of hysteroid trait and lifetime inpatient days predicted premature termination. Baseline symptom severity and initial medication assignment had little effect on clinic attrition. Global improvement and length of treatment were directly related.