A comparison of divalproex and oxcarbazepine in aggressive youth with bipolar disorder.

Background: Divalproex (DVP) and oxcarbazepine (OXC) are used to treat pediatric bipolar disorder (PBPD) with severe aggression but these agents have not been compared in head to head trials.

Methods: Electronic medical records were reviewed for those (age < 18) who received DVP (n = 20) or OXC (n = 11) for the treatment of PBPD with severe aggression. Change in prospectively rated Clinical Global Impressions-Severity (CGI-S) scores that measured global improvement of mental illness from baseline and rates of discontinuation due to adverse effects at approximately 4 months were the primary outcomes. CGI-S specific to aggression (CGI-Ag-S), which was rated retrospectively and blinded to treatment, was a secondary outcome.

Results: Greater reduction in CGI-S scores occurred with DVP compared with OXC at 4 months (p = 0.007). Both CGI-S and CGI-Ag-S scores improved significantly from baseline to the 4-month timepoint with DVP but not OXC. There were no significant differences in weight changes between the groups. Rates of discontinuation due to adverse events were similar. However, more discontinuations due to worsening aggression occurred with OXC (DVP 0%, OXC 27.3%, p = 0.037). More patients receiving DVP had a decrease of 1 point or more on the CGI-S and had not discontinued due to adverse events as of the patient's last recorded visit on the index medication (DVP 13 (65%), OXC 2 (18%), p = 0.023).

Conclusions: OXC appeared less effective than DVP for PBPD with aggression in this study. Limitations included the small sample size and the open, nonrandomized, retrospective study design. Future prospective, double-blind studies are warranted to determine the place of OXC in the treatment of PBPD.