The potential role for IV olanzapine was examined in a post last year. The following conclusions were reached:
- IV olanzapine appears to be safe.
- IV olanzapine has equal potency compared to IV droperidol and about twice the potency of IV haloperidol.
- Olanzapine doesn't affect QT interval or cause torsade de pointes.
Two articles were just published in the Annals of Emergency Medicine about intravenous olanzapine. What does the new evidence show?
Cole et al. A prospective observational study of patients receiving intravenous and intramuscular olanzapine in the emergency department.
This study is from Dr. Marc Martel's group at the Hennepin County Emergency Department in Minneapolis. Their emergency department began officially using IV olanzapine as a replacement for droperidol in January 2014. They previously published a retrospective study on the safety of IV olanzapine among 713 patients (explored in the prior post).
This is a follow-up prospective observational study comparing the safety of IV olanzapine versus IM olanzapine. 784 patients were treated with olanzapine (295 intravenously, 489 intramuscularly) for a variety of indications including agitation, headache, and nausea/vomiting. The primary outcome was respiratory suppression, with results shown below:
Only two patients in the intravenous olanzapine group required intubation. The details of these cases are shown below. These were sick trauma patients who probably would have required intubation regardless.
A post-hoc analysis was performed analyzing patients who received olanzapine without other sedatives. Among 247 patients treated intravenously and 396 treated intramuscularly, only one patient in the intramuscular group required intubation. This patient was intoxicated with an ethanol level of 358 mg/dL. The combination of alcohol intoxication and olanzapine is a known risk factor for over-sedation.
Regarding the risk of over-sedation, it should be noted that fully half of the patients treated with intravenous olanzapine had indications other than agitation (mostly headache and nausea/vomiting). Given that these patients weren't agitated at baseline, one might expect a higher rate of over-sedation in these patients. Nonetheless, olanzapine was safe.
Other side effects were minor. A few patients had extrapyramidal side-effects (akathisia or dystonia) requiring diphenhydramine. Two patients developed bradycardia, but they had numerous other causes of bradycardia and improved with conservative measures only.
One limitation of this study is that the initial dose of IV olanzapine was limited to 5 mg per emergency department protocol (doses used are shown below). Thus, this study doesn't prove the safety of higher doses (e.g. 10 mg). However, among agitated patients the 5 mg IV dose was effective (with 81% of patients not requiring additional treatment within an hour). Thus, going higher than 5 mg IV initially may be unnecessary.
Taylor et al. Midazolam-droperidol, droperidol, or olanzapine for acute agitation: A randomized clinical trial.
This was a two-center blinded RCT comparing the following regimens for patients presenting to the emergency department with agitation:
- Initial bolus of midazolam 5 mg IV plus droperidol 5 mg IV, followed by an additional 5 mg midazolam if needed five minutes later.
- Initial bolus of 10 mg droperidol IV, followed by an additional 5 mg of droperidol if needed five minutes later.
- Initial 10 mg olanzapine IV, followed by an additional 5 mg of olanzapine if needed five minutes later.
349 patients were randomized. The primary result was adequate sedation ten minutes after the first dose. The midazolam/droperidol group responded most rapidly, with no differences between the droperidol group vs. the olanzapine group:
Prior studies have suggested that adding a benzodiazepine to an antipsychotic increases respiratory complications. This study showed a strong trend towards increased transient airway obstruction among patients in the combination group:
Despite the adverse events listed above, all therapies were generally safe. No patients required intubation. This further substantiates the safety of IV olanzapine, despite the use of 10 mg IV boluses and 10-15 mg cumulative doses.
Commentary & data synthesis
Validation of prior evidence
First, these studies confirm that IV olanzapine is safe and equipotent to IV droperidol. We now have four studies documenting the safety of IV olanzapine among 1,237 patients within several centers in two countries (Chan 2013, Martel 2015, Taylor 2016, Cole 2016).
Why bother with IV olanzapine?
These articles are accompanied by an editorial, which wasn't particularly fond of IV olanzapine:
The use of intravenous olanzapine is more questionable. Its use is off label; there is no evidence to support intravenous being superior to intramuscular administration, similar to droperidol; and the study by Cole et al suggests it is associated with greater respiratory depression. In addition, the use of parenteral olanzapine (particularly intravenous) for nausea, vomiting, headache, and pain is not appropriate. There are numerous other available effective treatments for these conditions, and the risk of respiratory depression is unacceptable when one is treating less severe conditions. –Isbister 2016
A few comments. First, IV olanzapine remains off-label for intravenous use, but so is IV haloperidol. In fact, IM olanzapine is only FDA-approved for use in schizophrenia and bipolar disorder. Thus, our current use of IM olanzapine among undifferentiated agitated patients is already off-label. The reality is that off-label medication use is ubiquitous and often represents the standard of care.
Second, this editorial refers to the fact that in Cole et al., there is a higher rate of respiratory depression among patients treated with IV olanzapine versus IM olanzapine (3.7% vs. 2%; first table in this post). This difference doesn't come anywhere close to reaching statistical significance (using a Fisher's Exact test, p=0.17). Furthermore, the rate of intubation (arguably a more important clinical outcome) was lower in the IV olanzapine group (0.7% vs. 1%).
Third, the editorial points out that for an agitated patient without any intravenous access, inserting an IV line for the purpose of giving IV olanzapine may not have much advantage compared to giving IM olanzapine. Perhaps. But this is missing the point that many patients receiving olanzapine already have an IV line. For example, olanzapine is often used for ICU patients with mild/moderate agitation, all of whom have IV access. For patients with intravenous access, IV administration is faster, painless, and safer for nurses (no risk of needle-stick injury). Finally, intravenous administration promotes the use of lower doses with titrated dosing (as opposed to IM administration, which encourages giving a single big slug of medication).
Combined sedation vs. monotherapy?
A topic of ongoing debate is whether to use antipsychotic monotherapy or a combination of antipsychotic plus benzodiazepine (e.g., the classic “B-52 Bomber” combination of 5 mg haloperidol plus 2 mg lorazepam). Benzodiazepine may increase the risk of respiratory depression, an effect hinted at above and shown previously (e.g. Knott 2006).
Prior studies haven't found much benefit from combination therapy. However, RCTs on this are surprisingly sparse and small. To date, none has evaluated intravenous combination therapy.
The advantage of IV midazolam is very rapid onset (roughly 1-5 minutes, faster than IV lorazepam, haloperidol, or droperidol which may be closer to ~10-15 minutes). Therefore, it shouldn't come as a surprise that IV midazolam hastens sedation. The question is whether this speed is worth a slight increase in respiratory suppression. Risk/benefit balance may depend on the context:
- For a severely agitated, robust patient who provides a risk to self and others, combination therapy may be preferred due to greater efficacy.
- For a fragile patient with mild/moderate agitation who poses no risk of harm, monotherapy may be preferred due to greater safety.
For the profoundly agitated patient who presents a severe danger, a dissociative dose of IV ketamine is unquestionably the fastest agent (effective in ~30 seconds). Thus, when minutes truly matter, a combination of antipsychotic plus midazolam may not be the best strategy anyway.
The debate about the ideal sedative(s) will undoubtedly rage on, with sound arguments and strong opinions on all sides. These studies add support to IV olanzapine as another tool in the proverbial toolbox. Olanzapine is especially attractive among critically ill patients with multiple comorbidities and medications, due to its efficacy and lack of effect on QTc.
- These studies validate the safety of IV olanzapine (now demonstrated in four studies including administration to 1,237 patients).
- IV olanzapine has an efficacy equal to IV droperidol, and probably about twice as potent as IV haloperidol.
- Addition of IV midazolam to IV droperidol hastens achievement of sedation, but this may come at the cost of increased respiratory suppression.
We appreciate your review of our recent study and summary of the current literature on IV olanzapine. We agree that it is important to remember that the only FDA approved indication for parenteral (IM) olanzapine is “agitation associated with schizophrenia and bipolar I mania”(1). Clinicians are using olanzapine off-label whether it is for acute undifferentiated agitation or for any indication via the IV route of administration. But rest assured, you will be in good company with respect to off-label medication use for agitation. A few reasons why:
- Haloperidol is not approved for IV use.
- Haloperidol is only FDA approved to treat schizophrenia.
- Parenteral lorazepam is FDA approved for the treatment of status epilepticus and pre-anesthesia anxiolysis.
- Although droperidol is FDA approved for both IM and IV administration, it is only approved to treat nausea and vomiting associated with anesthesia.
For decades, clinicians have prioritized patient safety, provider wellbeing, and the efficient medical evaluation and treatment of patients with undifferentiated agitation over approval from the FDA (2). This patient population in particular, needs us to continually seek out innovative pharmacologic therapies over physical restraint when verbal de-escalation techniques fail.
The primary outcome measure of our study was the incidence of respiratory depression. We defined events of respiratory depression as any incident of endotracheal intubation, bi-level positive airway pressure support, bag-valve-mask ventilation, loss of protective airway reflexes, airway repositioning, stimulation to induce respiration, airway suctioning or the addition of supplemental oxygen. Our intent was to be highly sensitive, including events associated with essentially no morbidity or mortality when managed by emergency physicians and intensivists.
If agitated patients are excluded from our findings, a total of 148 patients received IV and 59 patients received IM olanzapine for either headache, abdominal pain, nausea/vomiting, vertigo or another indication. These are the patients we should examine in more detail for evidence of respiratory depression specifically related to olanzapine, as they are the most vulnerable to the effects of over-sedation and respiratory compromise. Four of 148 (2.7%) IV and zero of 59 (0%) IM had any event of respiratory depression. No patients were intubated, and the four who were treated for respiratory depression in the IV group were managed simply with the placement of supplemental oxygen. Obviously, these events of respiratory depression are not comparable to those requiring airway manipulation or intubation.
Any combination of CNS depressants has the potential to result in over-sedation and respiratory depression. As you highlighted though, our post-hoc analysis looked at the 643 patients who only received olanzapine (no additional or combined sedatives) and there was a very low rate of respiratory depression. None of the IV olanzapine patients were intubated, 9 of 247 had respiratory depression. In contrast, four of the 396 IM olanzapine patients had respiratory depression and one was intubated.
Other adverse effects were limited. Although we only reported 2 cases of bradycardia that resolved with conservative management, anecdotally in our subsequent experience, we feel this may be more common, but unlikely clinically significant.
We believe (and liberally practice under the auspice) that IV olanzapine is safe for a number of off-label indications. In agitated patients with IV access, the use of IV olanzapine avoids additional needle sticks for patients. It also eliminates the risk of needle stick injuries and exposure to blood-borne pathogens for the nursing staff who are asked administer IM olanzapine in these patients. The risk and benefit profile seems too far in favor of IV administration. While it is true a number of cheap effective drugs exist for many of our off-label uses, such as GI complaints and acute headache, these drugs, such as metoclopramide and prochlorperazine, have significant side effects such as akathisia and dystonia, and have both proven inferior to droperidol. Our observation (and data) suggest olanzapine has similar efficacy to droperidol with fewer adverse events than metoclopramide or prochlorperazine. While it is true we currently have cheap effective medications for headache and GI complaints, their presence does not mean we should cease striving to improve on therapies for these common and troublesome complaints.
Ultimately, we feel that emerging evidence will support the use of olanzapine for a variety of emergency department indications (3,4). But, to suggest that the use of IM olanzapine is restricted to use as a “reasonable alternative to droperidol in regions in which there is limited availability of droperidol, and IV olanzapine is “not appropriate” for nausea, vomiting, headache and pain (5) is far too restrictive based on the current literature in our opinion.
Marc L Martel and Jon B Cole
- http://www.fda.gov/downloads/advisorycommittees/committeesmeetingmaterials/pediatricadvisorycommittee/ucm191898.pdf. Accessed January 5, 2017.
- Clinton JE, Sterner S, Stelmachers Z, Ruiz E. Haloperidol for sedation of disruptive emergency patients. Ann Emerg Med. 1987;16(3):319-22.
- Hill CH, Miner JR, Martel ML. Olanzapine versus droperidol for the treatment of primary headache in the emergency department. Acad Emerg Med. 2008;15(9):806-811.
- Martel ML, Klein LR, Rivard RL, Cole JB. A Large Retrospective Cohort of Patients Receiving Intravenous Olanzapine in the Emergency Department. Acad Emerg Med. 2016;23(1):29-35.
- Isbister GK. Droperidol or Olanzapine, Intramuscularly or Intravenously, Monotherapy or Combination Therapy for Sedating Acute Behavioral Disturbance. Ann Emerg Med.