Olanzapine Versus Quetiapine In The ICU

Olanzapine (Zyprexa) and quetiapine (Seroquel) are often discussed as if they are interchangeable sedating antipsychotics used to help with addressing facets of delirium (insomnia, emotional lability, etc.). It often feels like dealer’s choice when deciding which one to use. Here are some of my considerations as an intensivist.

Olanzapine is available as an orally disintegrating tablet, intravenous injection, and intramuscular injection, making it uniquely valuable when agitation is severe or oral access is unreliable. Quetiapine is oral-only, which immediately limits its use to patients who can take and absorb enteral medication.

Olanzapine tends to deliver more consistent antipsychotic and anti manic effects at relatively modest doses, whether given orally or intramuscularly. The IM formulation can be effective for rapid calming, but it comes with important cautions, especially when combined with benzodiazepines, where respiratory depression has been reported. In contrast, quetiapine often behaves primarily as a sedative at low doses, with actual antipsychotic effects emerging only at higher dosing, which can make it worthwhile for nighttime agitation but less reliable for acute psychosis.

As far as side effects, olanzapine has a reputation for metabolic issues like weight gain and glucose dysregulation. Quetiapine is usually lighter on extrapyramidal symptoms but brings orthostasis, QT prolongation, and a pronounced hangover effect that can interfere with daytime recovery, especially in medically ill or postoperative patients.

Overall, the route and goal should lead the decision. If one needs rapid control of agitation and oral dosing is not an option, olanzapine (or really, haloperidol) may be the only realistic choice with careful attention to co-administered sedatives. If the patient is stable enough for oral therapy and the goal is sleep or mild agitation rather than frank psychosis, quetiapine can be reasonable. My preference is to stop thinking of this as a brand-versus-brand debate and instead view it as a risk-profile match. I pick one, start low, reassess early, and am willing to stop when the acute problem has passed.

Here are some PubMed IDs to review the data: 16172203, 17329466, 16974184, 19915454, 36845642