Objective

To identify the factors associated with hospital readmission in patients with bipolar disorder between 2011 and 2017 in a psychiatric hospital in order to develop an explanatory model of the risk of rehospitalization.

Method

Retrospective longitudinal observational study based on a secondary database provided by the Hospital Psiquiátrico Departamental Universitario del Valle. Statistics: Survival analysis using a Cox regression model. The risk of readmission was defined as a dependent variable and some sociodemographic and clinical variables were defined as independent variables.

Results

From the multivariate model of Cox Regression, a statistically significant association was obtained between the risk of readmission and the variables: lack of support network due to lack of retentive HR=1.33 [95% CI 1.02-1.72; (p=0.034)] and not having adherence to the treatment HR=2.33 [95% CI 1.87-2.90; (p=0.000)].

Conclusion

In order to reduce hospital readmission due to bipolar disorder, interven-tions should be prioritized in cases where patients do not have a guardian and there is inadequate adherence to the treatment; the latter perceived by the specialist doctor. Focusing on these two factors can influence readmissions.