Material and Method: A retrospective cohort based on medical records of both in- and out-patient with bipolar I disorder (DSM-IV-TR) seen at Srinagarind Hospital between January 1, 2009 and December 31, 2010 was done. A recurrence was observed if the patient had fulfilled the remission criteria and valproate was the maintenance drug. Survival analysis and Cox regression analysis were used to analyze the data.

Results: Eighty-five patients with 124 remitted mood episodes met the inclusion criteria. Of the 85 patients, the average age was 41.4±17.1 years (range, 18-89); 49.4% were males; 48.2% married; 42.4% completed secondary school, 30.6% completed a bachelor degree; 35.3% were unemployed, and 34.1% were government employees. Twenty remitted mood episodes (16.1%) were in maintenance treatment with valproate only. The remaining 104 (83.9%) were in maintenance treatment with valproate in combination with other agents. There were 50 recurrences from 36 patients during the two years of study, the recurrence per a patient ranged from 1 to 3 times. The rate of recurrence was 21%/year or 2.2/100 personmonths (95% CI = 1.65-2.93). The average time to recurrence to any mood episode was 33 months (95% CI = 15.06-50.94). With multivariable Cox regression, a statistically significant greater risk for a recurrence was associated with: (a) the previous episode being hospitalized (adjusted hazard ratio = 5.88, 95% CI = 2.76-12.36, p<0.001); (b) blood valproate concentration during maintenance treatment <50 μg/mL (adjusted hazard ratio = 3.07, 95% CI = 1.11-8.53, p = 0.03); and (c) time duration (month) of valproate maintenance treatment (adjusted hazard ratio = 0.98, 95% CI = 0.96-0.99, p = 0.001). With the adjusted hazard ratio 0.98, it could be interpreted in the other way that each additional month of taking valproate was associated with a statistically significant protective factor that decreased the risk of recurrence by 2% from the previous month.

Conclusion: In the authors’ clinical setting, valproate both singly and in combination with other psychotropic agents used for maintenance treatment of bipolar I disorder yielded a recurrence rate of 21% per year or 2.2 per 100 person-months and time to any mood episode recurrence of 33 months. The present result has importance for both clinical treatment decision-making and patient economic status.

Keywords: Maintenance treatment, Bipolar I disorder, Valproate, Recurrent rate, Time to recurrence