To the Editor: Dr. Sachs et al. conducted a 3-week double-blind, randomized, placebo-controlled study of risperidone or haloperidol augmentation of a mood stabilizer in bipolar patients with a current manic (N=123) or mixed (N=33) episode. They found that the antipsychotic drugs conveyed benefits in patients with mania but not in those with mixed illness. From my experience, the latter finding is hard to accept. I suggest that it arose because the authors did not examine early benefits with antipsychotic drugs and/or used insensitive measures to assess treatment gains.

With reference to the former possibility, the authors compared treatment gains at the 3-week endpoint but not, for instance, during the first week; this is not a trivial issue because every additional day that a patient is disturbed adds to the risk of illness-related harm to the self or environment. With reference to the latter possibility, the authors did not examine whether nonviolent self- and environment-damaging acts resulting from impaired judgment were reduced by antipsychotic augmentation. If the sizes of individual groups were too small for analysis, the drug groups could have been combined to determine whether antipsychotic treatment, per se, is helpful early or otherwise during a mixed episode.

Of note, if the authors are right that risperidone or haloperidol augmentation of mood stabilizers is unhelpful in mixed illness, then olanzapine augmentation may be worth considering. Tohen et al. (1) found that olanzapine attenuated both manic and depressive symptoms and that the gains were greatest in valproate-treated patients with mixed illness.