Predictors of decertification from involuntary hospitalization for patients with bipolar disorder

Objective: This study examined predictors of decertification (release from involuntary hospitalization after legal hearing) among inpatients with bipolar disorder. Methods: Records from 1992 to 1997 were examined retrospectively for 50 decertified and 48 certified patients with bipolar disorder. The relationship between demographic and clinical variables and decertification was examined using logistic regression analyses. Results: In the overall multiple logistic regression model, participants were significantly more likely to be decertified if they used a mood stabilizer before the decertification hearing (odds ratio [OR]=6.73, 95% confidence interval [CI]=1.78-25.50) or if they had a comorbid substance use disorder (OR=3.45, CI=1.15-10.34). The odds of decertification increased with the number of prior hospitalizations (OR=3.92, CI= 1.73-8.87) and decreased with the length of prior hospitalization (OR=.72 per week, CI=.49-1.04) and number of emergency room visits before admission (OR=.46, CI=.28-.74). Conclusions: Predictors of decertification in bipolar disorder require further research to guide future efforts to improve inpatient treatment outcomes.