I read with interest the article by Fisher & Broderick regarding the differences between sodium valproate and valproate semisodium in the management of bipolar disorder (Psychiatric Bulletin, December 2003, 27, 446-448). I would like to comment on both the accuracy of the information presented and highlight important information not included in the article.

The vast majority of trials use valproate semisodium, with over 4000 bipolar patients involved in these studies. The authors erroneously state that the Pope et al (Reference Pope, Mcelroy and Keck1991) study used sodium valproate, when in fact valproate semisodium was used in this placebo controlled trial.

The search criteria used did not include data published in 2003 and consequently missed a further large, randomised, controlled trial which compared valproate semisodium with olanzapine in the management of bipolar disorder, and showed no difference in rates of bipolar relapse between both agents (Reference Tohen, Ketter and ZarateTohen et al, 2003).

With regard to tolerability, it should be noted by the authors that the most commonly used preparation of sodium valproate in the UK, Epilim Chrono, is not enteric coated and will therefore break down to valproic acid in the stomach and small intestine. The studies showing improved tolerability of valproate semisodium v. valproic acid therefore have some particular relevance to the UK.

Finally, the National Institute for Clinical Excellence (2003) has recently reviewed extensively the evidence regarding the use of valproate semisodium in the management of bipolar disorder, and the guidelines specifically state that this compound is recommended for the treatment of acute mania associated with bipolar I disorder.