Droperidol for Agitation in the ED - No Danger to the Dangerous?
/Droperidol is a versatile medication with a number of potential uses for patients in the Emergency Department. It is also a medication surrounded in some degree of mystique because of the decision by the FDA in 2001 to issue a black box warning for its use in response to reports of QT prolongation and torsades de pointes. Many at the time (and since) have argued that, despite these case reports, droperidol is a safe and effective medication that can be used for the treatment of agitation, nausea and vomiting, and migraine. We have previously covered much of this background in a previous blog post. In our most recent journal club, we discussed 3 articles that looked at the safety and efficacy of droperidol for treating acutely agitated patients. Take a read and listen below for an in depth look at each of these papers.
Calver, L. et al. The Safety and Effectiveness of Droperidol for Sedation of Acute Behavioral Disturbance in the Emergency Department. Ann Emerg Med 66, 230-238.e1 (2015).
Background and Methods
The authors in this study sought to “investigate the frequency of QT prolongation and torsades de pointes” in patients administered droperidol for the treatment acute agitation in the Emergency Department. Their primary outcome was the proportion of patients with an abnormal QT interval on an EKG done within 2 hours of droperidol administration as assessed by a QT nomogram that has been used previously in the toxicology literature. Secondary outcomes included proportion of patients with torsades de pointes, time to sedation, over/undersedation, and other adverse events.
Prospective multicenter observational study
Included patients - Acutely agitated patients at risk to themselves or others
All study sites had a standardized protocol to administer these patients droperidol 10 mg IM or IV
If incomplete sedation, an additional 10 mg droperidol was administered
If additional sedation needed, other medications were at the discretion of the treating provider
EKGs obtained as soon as reasonably possible
QT determination was made by hand by a single reviewer who had a subset of QT intervals reviewed by a second author (noted 96% agreement for interpretation of normal or abnormal by QT)
Prospective data collection form was used to collect sedation scores and record adverse events
Results
Primary outcome - 6/1009 patients with otherwise unexplained “at-risk” QT prolongation
Secondary outcomes
0 cases of torsades de pointes in the trial
70 adverse events - 22 patients with oxygen desaturation, 28 patients with hypotension, and 7 patients with extrapyramidal symptoms
Time to Sedation - 20 min with 69% of patients being effectively sedated with the initial dose
Oversedation - 109 (7.8%) patients had a Sedation Assessment Tool score of -3 (unresponsive to verbal or physical stimuli)
Impact/Applying Results to Our Patients
The global takeaway from this paper is that, in acutely agitated ED patients, the administration of rather large dosages of droperidol was not associated with a significant risk of QT prolongation or torsades de pointes. With that being said, there are some questions as to the direct applicability of this study to the patients we see in the ED. There was a relatively large number of patients who became oversedated as a result of this study protocol. In addition to the demographic information provided in Table 1 of the paper, it would have also been helpful to see a breakdown of the medical comorbidities of the patients (particularly BMI, rates of OSA or other chronic respiratory illness, and pre-existing medications). It is possible that adverse respiratory events like hypoxia would be higher in a larger BMI more medically ill patient population. Having that additional information would help in an assessment as to the similarities of this patient population to your own. Also, note that the study population is very different from the patient presenting to the ED with nausea and vomiting, who is on or has been on other anti-emetics (many of which are known to cause QT prolongation).
Taylor, D. McD. et al. Midazolam-Droperidol, Droperidol, or Olanzapine for Acute Agitation: A Randomized Clinical Trial. Ann Emerg Med 69, 318-326.e1 (2017).
Why we chose this article
Agitation is commonly encountered in the Emergency Department.
The need for effective and safe sedation for these patients is important.
Droperidol is starting to be more frequently used in this setting.
Why it’s important
Sedation for acute agitation is not uncommon in the ED.
Risk of sedation is real and includes airway compromise, hypoxia, hypotension, and extrapyramidal events.
ED sedation guidelines are often inconsistent, not evidence based, and frequently not followed.
Study details
Randomized, controlled, double-blinded, triple-dummy, clinical trial conducted in 2 inner-city tertiary-referral hospitals
Inclusion criteria:
Patients 16 to 65 years old
Required IV sedation medication for acute agitation
Exclusion criteria
Previously enrolled
Known hypersensitivity or contraindication to study medications
Reversible cause of agitation (hypotension, hypoxia, hypoglycemia)
Patient’s undergoing EtOH withdrawal
Pregnancy
Methods
Patients were assigned to one of three arms: midazolam-droperidol combination, droperidol monotherapy, or olanzapine monotherapy
Dosing of medications
Midazolam-droperidol arm: 5 mg IV midazolam + 5 mg IV droperidol (up to two additional doses of 5mg IV midazolam allowed)
Droperidol arm: 10 mg IV droperidol (up to two additional doses of 5mg IV droperidol allowed)
Olanzapine arm: 10 mg IV olanzapine (up to two additional doses of 5mg IV olanzapine allowed)
Outcomes
Primary: proportion of patients adequately sedated within 10 minutes of the first dose
Secondary: time to adequate sedation, need for re-sedation (<60 minutes from initial sedation or >60 mins from sedation until ED discharge), medication failure, QTc interval, adverse events
Data analysis:
Utilized intention to treat principal
Proportion of patients adequately sedated at 10 minutes were analyzed with differences in proportions (95% CI)
Time to sedation analyzed with difference in medians (95% CI) and survival time data was plotted with Kaplan-Meier curve
Hazard ratios for adequate sedation were generated with midazolam-droperidol group as the baseline and multivariable Cox regression was used to adjust for regular medications and medications administered before the study.
Results
349 total patient data analyzed
Midazolam-droperidol arm with more patients adequately sedated within 10 minutes
Difference in proportions: droperidol (25%) vs olanzapine (25.4%)
Faster medial time to sedation in midazolam-droperidol arm
Difference in medians: droperidol (6 min) vs olanzapine (6 min)
Fewer additional doses required for midazolam-droperidol arm
No statistical difference in adverse events or length of stay
Limitations
Midazolam-droperidol arm had slightly more less urgent triage categories, patients with a history of mental illness, and disposition to a psychiatric ward
Sedation scale was potentially subject to measurement bias
No external validity
Questionable practicality of IV medications for agitated patients
Take-home points
Combination therapy with midazolam and droperidol appears to be a safe and more effective way to combat the agitated patient.
May not be practice changing but highlighted the safety profile of large doses of droperidol and olanzapine
Isbister, G. K. et al. Randomized Controlled Trial of Intramuscular Droperidol Versus Midazolam for Violence and Acute Behavioral Disturbance: The DORM Study. Ann Emerg Med 56, 392-401.e1 (2010).
Background
Violent and aggressive behavior in the emergency department is unfortunately a common problem that presents a risk to both patients and staff. The medications most commonly used to sedate aggressive patients in the emergency department are benzodiazepines and antipsychotics. However, benzodiazepines often undersedate patients with benzodiazepine tolerance and cause oversedation in higher doses. Antipsychotics have been associated with cardiac dysrhythmias and QT prolongation and are often not sufficiently sedating. Droperidol is a highly sedative antipsychotic which fell out of favor after a black box warning was issued due to the potential for QT prolongation and torsades. Further review later indicated that this risk may have been exaggerated, but likely because of the black box warning, droperidol has not been extensively studied as a potential tool for sedating agitated patients in the emergency department. This study sought to evaluate the safety and efficacy of droperidol for this purpose by comparing it to both midazolam and a combination of droperidol and midazolam.
Methods
This was a triple blinded randomized controlled trial in which agitated patients in the emergency department were sedated with either 10 mg of droperidol IM, 10 mg of midazolam IM, or 5 mg of midazolam and 5 mg of droperidol IM, each of which were administered in identical, de-identified vials. Inclusion criteria were patients requiring physical restraint and parenteral sedation. Exclusion criteria included successful verbal de-escalation or patients who agreed to take a sedative medication, other sedative medication already administered or patient did not remain in the ED, or acute seizures/postictal state. Once patients were identified for enrollment in the study, they were administered a vial of medication as above and then observed in a monitored bed in the emergency department for 6 hours unless they eloped or awoke and no longer required restraint. All patients received an EKG at 30 minutes, 1 hour, and 4 hours after receiving the trial drug. The primary outcome studied was the duration of the violent episode, which was measured as the time that security was required at bedside. In this hospital, it was standard practice to call an “all clear” when security was no longer needed, so this outcome was able to be easily measured without providers being aware of the outcome that was studied. Secondary outcomes included the time until additional sedative medication was administered (chosen and initiated by the provider), reduction in the Altered Mental Status Scale, injuries to patient or staff, further calls for security, and any drug-related adverse effects, including QT prolongation.
Results
91 patients were enrolled in the study; 33 received droperidol, 29 received midazolam, and 29 received the combination. There was no statistically significant difference in duration of behavioral disturbance between the three different medication options (20 minutes for droperidol, 24 minutes for midazolam, and 25 minutes for the combination). There was also no difference in incidence of injury to patient or staff between the three groups. However, patients who received droperidol were less likely to require additional sedation than patients in the midazolam or combination groups. The hazard ratio for additional sedation in the midazolam versus droperidol group was 2.31 (95% CI 1.01-4.71). The hazard ratio for additional sedation in the combination versus droperdiol group was 1.18 (95% CI 0.46-2.5). Adverse effects were more common in the midazolam group (8 patients or 28% in the midazolam group (95% CI 13%-47%) versus 2 patients or 6% (95% CI 1%-22%) for the droperidol group and 2 patients or 7% (95% CI 1%-24%) for the combination group. There was no difference in incidence of abnormal QT among the three groups.
Discussion
This was a well designed, triple blinded, randomized controlled trial with few limitations. The study did not control for prior substances, such as alcohol, ingested by the patients and did not use weight-based dosing. However, both of those interventions would have been unrealistic in the setting of an agitated patient, as providers most often do not know what the patient may have ingested prior to presentation and are not able to obtain an accurate weight prior to administering medication. It is therefore important to study the relative safety and effectiveness of each drug in a setting where such prior knowledge of the patient is unknown. One limitation cited was the primary outcome being the duration of time that security was required rather than time until sedation using the Altered Mental Status Scale. However, in this setting, duration of time that security was required was thought to be a more objective measurement. Use of the Altered Mental Status Scale may also have been somewhat compromised by interobserver bias. Finally, allowing providers to take the lead on administration of additional sedative medication may have influenced the incidence of adverse effects, but provided important information about the need for additional sedation in each group. Overall, this was a valuable study which demonstrated that the use of droperidol for agitated patients is associated with a lower incidence of adverse effects than midazolam, as well as a decreased need for additional sedation as compared with midazolam or the combination.
References
Calver, L. et al. The Safety and Effectiveness of Droperidol for Sedation of Acute Behavioral Disturbance in the Emergency Department. Ann Emerg Med 66, 230-238.e1 (2015).
Taylor, D. McD. et al. Midazolam-Droperidol, Droperidol, or Olanzapine for Acute Agitation: A Randomized Clinical Trial. Ann Emerg Med 69, 318-326.e1 (2017).
Isbister, G. K. et al. Randomized Controlled Trial of Intramuscular Droperidol Versus Midazolam for Violence and Acute Behavioral Disturbance: The DORM Study. Ann Emerg Med 56, 392-401.e1 (2010).
Authorship
Calver et al. - Jeffery Hill, MD MEd, Associate Professor, University of Cincinnati Department of EM
Taylor et al - Shawn Hassani, MD, University of Cincinnati Department of EM, PGY-3
Isbister et al - Christa Pulvino, MD, MD, University of Cincinnati Department of EM, PGY-3
