Abstract
Delirium affects up to 80% of critically ill patients, worsens outcomes, and is frequently treated with antipsychotics despite uncertainty regarding their efficacy and safety. We identified published, English-language, randomized, controlled studies evaluating antipsychotics in ICU patients either with delirium or at risk for developing delirium. In 105 mechanically ventilated patients, the number of days spent alive without delirium or coma was similar between haloperidol (median, 14.0 days; interquartile range [IQR], 6.0–18.0 days) or ziprasidone (median, 15.0 days; IQR, 9.1–18.0 days) prophylaxis, and placebo (median, 12.5 days; IQR, 1.2–17.2 days) groups (p = 0.66). Treating delirium with quetiapine, compared to placebo, in 36 ICU patients was associated with a quicker resolution of delirium (median for quetiapine, 1.0 days; IQR, 0.5–3.0 days/median for placebo, 4.5 days; IQR, 2.0–7.0 days [p = 0.001]). In a third study, a similar decrease over time in delirium severity was noted between fixed-dose oral olanzapine and oral haloperidol in patients with delirium. None of the studies identified serious safety concerns with administering the antipsychotics that were studied. Published prospective, randomized clinical trials evaluating antipsychotic therapy for preventing or treating delirium in the ICU are few in number. The conclusions that can be drawn from them are limited by their heterogeneity, inconsistency in incorporating non-antipsychotic strategies known to reduce delirium or in maintaining patients in an arousable state, their size, the lack of ICU and non-ICU clinical outcomes evaluated, and the lack of placebo arms. A research framework for future evaluation of the use of antipsychotic therapy in the critically ill is proposed.