Level 2: Mid-level evidence

Both lithium and valproate have been used to prevent relapses in bipolar disorder. In a recent unblinded randomized trial, 330 patients age ≥16 years were given combination therapy with lithium and valproate during an active run-in period lasting four to eight weeks (Lancet. 2010;375:385-395). Patients were then randomized to lithium, valproate, or combination therapy and followed for 24 months. The primary outcome was a change in management due to a new mood episode, including medication adjustments and hospitalizations. This occurred in 54% of patients taking combination therapy (P = 0.0023 vs. valproate, number needed to treat [NNT] 7), 59% taking lithium (P = 0.0472 vs. valproate, NNT 10), and 69% taking valproate. While use of lithium (alone or with valproate) was statistically superior to valproate alone, this study did not find that combination therapy was significantly better than lithium alone. There were no significant differences among groups in serious adverse events (16 overall, including six deaths).