03. IV Haloperidol: Are Cardiac Safety Concerns Overblown?

Text version

Haloperidol Remains Cardiologists’ Least Favorite Drug

If you ask 100 cardiologists what their least favorite psychiatric medication is, my guess is that a large majority of them would say it’s haloperidol, particularly IV haloperidol. Since even before the FDA issued its black box warning in 2007, no psychiatric medication has been more associated in the minds of internists and other medical teams with adverse cardiac outcomes, particularly ventricular arrhythmias.

Several times per week, I’m asked by inpatient teams to recommend a medication to help manage the sequelae of delirium but with the additional caveat that I need to choose something other than IV haloperidol because of their concerns about QTc prolongation or torsades de pointes. I can’t think of another medication that I am so frequently asked to avoid using.

For years, my colleagues and I have been attempting to educate our peers about the differences between the myths associated with IV haloperidol and the actual evidence base. A few more recent studies and reviews have been helpful in this regard, but nothing has had the potential impact of a recent meta-analysis in the journal PLOS One, which is the subject of today’s Quick Take.

Meta-Analysis Challenges Haloperidol Fears

The authors of today’s study reviewed the literature on major adverse cardiac events associated with haloperidol, focusing on the highest quality evidence. 84 randomized controlled trials involving over 12,000 patients were included, which together reported over 1,100 cardiac events. The vast majority of these events, nearly 98%, were deaths.

Indications for haloperidol varied across studies, including:

• Psychiatric illness
• Delirium prophylaxis and management
• Dementia

A third of the included trials utilized the intravenous route of haloperidol. The mean daily dose of haloperidol was about 7 mg, so fairly low to moderate, with a range of 1 to 40 mg.

A quarter of the studies reported a mean or median age above 65 years, suggesting that an older population who may be at increased risk for adverse cardiac outcomes was well represented.

No Differences Found Between Haloperidol and Placebo

The authors found that there were no differences in adverse cardiac outcomes or death between those patients who had received haloperidol and those receiving placebo in the studies.

• No sudden cardiac deaths occurred among patients receiving haloperidol.
• There was a total of 2 episodes of torsades and 22 other ventricular arrhythmias in the haloperidol arm.
• A post hoc analysis looking specifically at IV haloperidol also found no differences in outcomes between IV haloperidol and placebo.

It’s hard to overstate the importance of this article for psychiatrists, and particularly for consultation liaison psychiatrists. It calls into question one of the most pervasive pieces of lore in all of medicine, namely that haloperidol and particularly IV haloperidol is a high-risk medication with regard to cardiac outcomes.

Outcomes: Arrhythmias and Death

I like that the authors focused their meta-analysis on the outcomes that actually matter to patients: adverse cardiac events, arrhythmias and death. They didn’t get bogged down in the related question as to whether haloperidol has meaningful effects on the QT interval, an assertion that incidentally has also been challenged many times in the recent past, including in a systematic review by our group which demonstrated no QT prolongation compared to placebo at doses less than 20 mg.

EKG Screening Did Not Reduce Events

Among the many interesting findings here is the fact that there was no difference in adverse cardiac events between trials that conducted electrocardiogram screening and those that didn’t. In other words, the findings here suggested EKG screening did not reduce adverse cardiac events with haloperidol.

This is particularly notable because the current FDA guidelines recommend telemetry, that is, continuous cardiac monitoring, for any patient receiving any dose of IV haloperidol, which theoretically represents a huge tax on resources in the hospital setting, although many studies have demonstrated that this recommendation is often not followed. Almost since the FDA warning was issued, studies have called into question the telemetry recommendation, but the FDA language has never been changed.

A recent ICU study in JAMA again suggested that there was no benefit to monitoring, and today’s study really hammers this point home, drawing on a large pool of evidence. For those institutions that do require telemetry with IV haloperidol, this meta-analysis should be a wake-up call that this costly practice has a very poor evidence base.

The authors here point out that the Canadian Cardiovascular Society has moved away from recommending even baseline EKGs on patients receiving QTc-prolonging medication, instead favoring a thoughtful risk-stratifying approach.

Monitoring Restrictions Implications

Why does the monitoring question matter so much? Is there a real downside to the increased caution that has been used with IV haloperidol over the past 20 years? I would argue that there is.

Restrictions steer clinicians toward less appropriate alternatives

First, restrictions prevent clinicians from considering IV haloperidol as a viable option because it’s just too much of a hassle. Medication safety alerts in the electronic health record are ubiquitous for IV haloperidol and almost certainly cause many ordering clinicians to reconsider their choice. In addition to these flags contributing to physician burnout, they may steer prescribers towards other antipsychotics such as olanzapine or quetiapine, which are frequently thought of as safer from a cardiac perspective despite no head-to-head evidence suggesting any difference in risk for ventricular arrhythmias or other adverse cardiac outcomes.

Prescribers in this situation may believe they are choosing a safer medication from a cardiac standpoint and may ignore the real differences in side effect profiles between haloperidol and these other agents, such as the significantly higher degree of anticholinergic and alpha-blocking effects of most atypical antipsychotics, which may have more practical negative implications for their delirious patients.

Hospital policies waste resources and limit effective treatment:

Second, patients receiving IV haloperidol may require telemetry if hospitals have chosen to follow the current FDA guidelines, taking up a precious and potentially unnecessary resource and possibly even requiring ICU admission. Even more concerningly, some hospitals have completely banned or severely restricted the use of IV haloperidol, effectively removing the best studied and most effective management option for one of the most common and expensive medical conditions.

Used correctly, haloperidol has the potential to effectively manage psychiatric symptoms, increase patient and staff safety in the setting of hyperactive delirium, and reduce the risk of post-delirium PTSD or post-ICU syndrome. As the authors of this meta-analysis point out, haloperidol management has even been shown in one study to reduce mortality over placebo at one year. Implementing policies that make it more difficult to prescribe and use haloperidol without a clear evidence base to support doing so has the potential to harm patients and worsen outcomes.

Conclusions and Future Directions

At the end of their article, the researchers here identified that a key priority is the development of a validated clinical prediction tool for patients who require QTc-prolonging medications to help clinicians determine which factors best predict adverse cardiac outcomes. This is really where we need to head as a field.

It doesn’t make any sense for systems to ban potentially useful medications based on low-quality evidence or to require intense monitoring for a very low likelihood outcome. Furthermore, heuristic shortcuts like instituting a specific QTc cut-off at which it would be safe to use IV haloperidol are overly reductionist and fail to account for the nuance of the clinical decision.

Instead, clinicians should get in the habit of thinking holistically about risk and conducting careful risk-benefit, or more likely risk-risk analyses, to determine which patients are likely to benefit from IV haloperidol without incurring significant risk.

Abstract

Major adverse cardiac events with haloperidol: A meta-analysis

Michael Cristian Garcia, Mason Anderson, Michelle Li, Mick Zewdu, Tyler Schneider, Jessyca Matos-Silva, Genevieve Ramnarine, Kassie Rong, Lawrence Mbuagbaw & Anne Holbrook

Background

Haloperidol is a commonly used antipsychotic drug and a frequent source of medication safety alerts because of its listing as a “known risk” QT interval-prolonging medication (QTPmed). We aimed to summarize the high-quality literature on the frequency and nature of proarrhythmic major adverse cardiac events (MACE) associated with haloperidol.

Methods

We searched Medline, Embase, International Pharmaceutical Abstracts, and Cochrane Central for randomized controlled trials (RCTs) involving patients 18 years or older comparing haloperidol to placebo. The FDA-adapted MACE composite included death, non-fatal cardiac arrest, ventricular tachyarrhythmia including torsades de pointes, and seizure or syncope. Random-effects meta-analyses were performed with a treatment-arm continuity correction for single and double zero event studies.

Results

84 RCTs (n = 12180, 46% female), 23.8% of trials reported mean or median ages of their participants to be older than 65 years with 37 (44.0%) involving participants with psychiatric diagnoses, and 50 (59.5%) including electrocardiograms. Median follow-up duration was 28.0 days (interquartile range [IQR]=51.0). There were 1144 events, of which 97.8% were deaths, with 22 ventricular arrhythmias and 3 seizures or syncope. There was no difference in MACE with exposure to haloperidol compared to placebo (risk ratio [RR] 0.93, 95% CI: 0.80–1.08; I² = 0%). IV haloperidol was not associated with increased risk of mortality (n = 5873, RR: 0.88, 95%CI:0.72–1.08).

Conclusions

We did not find that haloperidol was arrhythmogenic or increased mortality in these largely short-duration trials. Further research to clarify actual clinical outcomes related to QTPmeds is important to inform safe prescribing practices.

Reference

Garcia, M.; Anderson, M.; Li, M.; Zewdu, M.; Schneider, T.; Matos-Silva, J.; Ramnarine, G.; Rong, K.; Mbuagbaw, L. & Holbrook, A. Major adverse cardiac events with haloperidol: A meta-analysis. (2025). PLOS ONE, 20(6), e0326804.