• lithium;
  • manic episode;
  • meta-analysis;
  • short-term treatment

Objectives:  To provide an accurate estimation of the magnitude of effect of lithium in short-term efficacy studies conducted in patients with moderate to severe manic episode.

Methods:  All placebo-controlled randomized studies submitted to the Medicines Evaluation Board (MEB) in which lithium was used as the third study arm were selected for the meta-analysis. The studies were part of registration files submitted to the MEB between the years 1997 and 2005. In addition, Medline and EMbase searches were conducted with the key words ‘manic’ (‘mania’ for the EMbase search) and ‘placebo’ in order to identify additional placebo-controlled studies of lithium. This search was updated until March 1, 2006. Two effect size indicators were used based on the primary outcome measure of each study: Cohen's standardized effect size based on the difference in mean change from baseline, and numbers needed to treat (NNT) based on the difference in treatment response defined as ≥50% improvement from baseline on day 21.

Results:  Six studies were identified. They involved a combined total of 470 patients in the lithium groups and 562 in the placebo groups. The overall standardized effect size was 0.40 [95% confidence interval (CI): 0.28, 0.53] and the overall NNT for response was 6 (95% CI: 4, 13). In the placebo groups response rates varied from 21% to 47%.

Conclusions:  The results indicate that lithium is an effective drug in the treatment of moderate to severe manic episode. The variability in placebo response indicates that a placebo control arm in efficacy studies among patients with moderate to severe manic episode is necessary.