There are patients on long-term antipsychotics for no reason other than that they were once hospitalized, a troubling conclusion that can be drawn from recent studies of discharges from ICUs and hospital wards.
In an analysis of more than 18,000 veterans hospitalized for heart failure, published by the Journal of General Internal Medicine (JGIM) on Jan. 3, 10.7% were newly prescribed antipsychotics, and of those patients, 21.5% had the drugs continued in skilled nursing facilities (SNFs) after discharge.
An earlier study of 60 patients newly prescribed psychoactive medications while being treated in an ICU for various causes found that 30% had those drugs continued at hospital discharge, according to results published in the December 2020 Journal of Pharmacy Practice. Some of the prescriptions extended as long as six months.
In the latter study, a pharmacist talked to the treating teams about the use of the drugs. “I was surprised by how frequently physicians didn't know what the medications were originally started for, or even being used for,” said lead author Nicole Scherrer, PharmD, a board-certified critical care pharmacist. “It was somewhat alarming.”
The situation is particularly concerning given the likelihood that these prescriptions were not indicated from the start. “It's something I harp on with my pharmacists and colleagues: Every medication on a patient's profile should have an indication, and in the vast majority of situations, antipsychotics lack any evidence-based indication in the ICU,” said Gabriel Fontaine, PharmD, MBA, a board-certified pharmacotherapy specialist who is a clinical pharmacist, pharmacy manager, and associate professor at Intermountain Healthcare in Salt Lake City.
These and other experts offered their perspectives on the causes of problematic prescriptions, the risks, and what hospitalists and hospitals can do to improve the situation.
What the data say
Research does not offer much support for using inpatient antipsychotics, Dr. Fontaine noted. “We've now had a least a handful of pretty well-conducted randomized trials [RCTs] looking at antipsychotics versus placebo or other medications to treat delirium in the acute phase or even to prevent delirium,” he said. “And nearly every single one of the RCTs, especially the large ones, have shown no benefit of antipsychotics.”
Two systematic reviews published by Annals of Internal Medicine in October 2019 agreed that haloperidol or second-generation antipsychotics did not appear to have any effect on sedation status, delirium duration, or length of stay in hospitalized patients, although one found limited evidence that atypical antipsychotics may lower the incidence of delirium in postoperative patients.
The reviews also found cardiac effects even with short-term use of the drugs, and there can be other risks, according to the experts. “We know that they interact with a lot of medications,” said Dr. Fontaine.
The drugs also have metabolic effects and can cause glycemic abnormalities or dyslipidemia, added Melissa Riester, lead author of the JGIM study and a postdoctoral research associate in health services, policy, and practice at Brown University in Providence, R.I. “The FDA has even put out a black-box warning for these medications about an increased risk of death among older adults with dementia,” she said.
The use of typical antipsychotics was associated with a 60% increased risk of death among ICU patients, a recent study found. Atypical antipsychotics did not show the same overall association, but among patients who were ages 65 years and older, the drugs were associated with significantly increased risk of death or cardiac arrest, according to the results, published in the March 2020 Journal of the American Geriatrics Society.
However, the lead author of that study understands why clinicians may prescribe antipsychotics despite the evidence. “The reason they get used as often as they do is because we don't have a lot of tools in our armamentarium to calm a delirious patient down,” said Shoshana Herzig, MD, MPH, FACP, director of hospital medicine research and an associate professor of medicine at Harvard Medical School in Boston.
“Pharmacologic therapies, broadly, are ineffective at preventing or treating delirium,” added Dr. Fontaine. Effective interventions include friends or family present at the bedside, calming music, daylight during the day, and dark and quiet at night, but they may not fulfill clinicians' urge to proactively help patients. “We see patients are acutely agitated … and so we generally want to do something,” he said.
Some of the effective interventions are also hard to come by, particularly lately. “The COVID pandemic has made this situation worse,” Dr. Fontaine said. “Visitor policies, especially early in the pandemic, completely barred friends and family from visiting their loved ones in the ICU, and these patients were now receiving care in an unfamiliar environment by caregivers in suits head to toe, looking completely foreign and alien.”
Hospital staffing issues add to the challenge. “The only thing you can do for a patient who's delirious and acting out is have someone sit there with them and keep them safe, but hospitals often don't have the resources to supply a body to sit there,” said Dr. Herzig. “We're only supposed to be using antipsychotics if the patient is a threat to themselves or others, but I do think that oftentimes, they get used as a Band-Aid in lieu of adequate personnel.”
Antipsychotics should be a very short-term bandage, according to Dr. Fontaine. “In the setting of severe agitation, there may be a role for a single-dose, as needed, immediate-acting antipsychotic such as haloperidol,” he said. “However, I generally don't see antipsychotics used in that fashion.”
The problem is that it's a slippery slope, noted Dr. Herzig. “Is the patient really an imminent threat? How is that defined? If we don't have someone to sit here with them, even though they're not in imminent threat right this second, they could at any second do something that does cause a problem for themselves or for others,” she said.
The use of these drugs has been more common than even that concern could justify, however, Dr. Herzig added. “I see less of this now, but in a patient who requires these meds frequently overnight, some providers were starting to order them as a standing order, so every night around 8 p.m., the patient would get this medication,” she said.
Once that medication order gets placed, it tends to stick around. “For a lot of us practicing, the assumption is that they are taken off by someone at some point, whether it's us, whether it's the team on the floor,” said Dr. Scherrer, who is a medical ICU clinical pharmacist at West Virginia University Medicine in Morgantown.
In fact, other clinicians may be very reluctant to stop the drug. “A lot of times providers don't know how to remove these medications. There's fear that the patient might become agitated again or have issues with sleep or delirium,” Dr. Scherrer said.
That can be a concern even for the physicians who start the drugs. “We end up having to use antipsychotics at times to keep a patient calm in order for them to qualify for a postdischarge facility and move their care forward,” said Dr. Herzig. “They need not to have required restraints for a certain period of time or facilities don't want to take the patient.”
Once a disposition is found, discharge can come around quickly before the issue of antipsychotics is ever addressed. “There's not time to re-evaluate the med list or not time to tell the SNF or [long-term acute care facility] how they should get the patient off these medications,” said Dr. Scherrer.
Fixing the problem
One clear way to reduce antipsychotic use is to not allow ongoing prescriptions. “We can make sure antipsychotics, if they're not a home medication, are only prescribed as prn, or there's hard stops in our electronic health record to say they'll only be a single-dose administration,” said Dr. Fontaine.
If more than one dose is allowed, there could be a set discontinuation time, such as 24 hours from the first dose. “So it needs to be mindfully re-evaluated every single morning during rounds,” he said. “We know that interventions like this have been pretty effective in the antimicrobial stewardship world.”
Frequent mindful checks of the medication record were on all the experts' lists of solutions. “If somebody starts an antipsychotic in the ICU, for example, I'm giving a clear handoff with clear documentation when somebody is transitioning from the ICU to a general medicine floor, then from the medicine floor to a skilled nursing facility,” said Dr. Riester. “Each of those transitions would be an important point to assess the appropriateness of any antipsychotic.”
That assessment can be made easier if the initial prescriber clearly states in the record why the antipsychotic is being used. “Right from order entry, they would be getting a label on them that says that they were used for ICU agitation and delirium, so that it is clear and very obvious whenever the patient leaves the ICU what this medication was being used for, and hopefully allows providers to be more comfortable taking them off,” said Dr. Scherrer.
Transitions could also be points for hard stops of prescriptions to be built into systems. “If a patient goes from ICU status to stepdown status, then the medication would be automatically discontinued from the med list based on the change in acuity,” said Dr. Scherrer. However, she noted, “There's pushback from providers with that type of approach.”
Teamwork could be a more popular solution. Dr. Herzig urged hospitalists to push for increasing staffing to reduce antipsychotic use. “Hospitals really need to fund support for providers in the form of people who can come in and sit with patients,” she said. Starting antipsychotics in the hospital is associated with increases in length of stay and readmissions, so there's a financial case to be made for tackling this issue, Dr. Scherrer noted.
More pharmacists could help, too, according to Dr. Fontaine, who noted that every ICU in his health system has a dedicated one. “I think pharmacists as part of the ICU team, and also part of the medical teams and, particularly important, a transitions-of-care team, play a critical role here, too,” he said. He added that California recently passed a law which requires hospital pharmacists to obtain an accurate medication list for each high-risk patient upon admission, a concept that might make sense for the end of a hospitalization, too.
“We've very, very mindful about ensuring correct medications when patients are coming into the hospital … but we don't necessarily do that upon discharge,” he said.
There are hints from the research that clinicians are already to some extent thinking about these medications at discharge. “The atypical antipsychotics carry a higher risk of being continued at discharge from the hospital than the typical antipsychotics,” said Dr. Scherrer. “It seems providers are more comfortable discontinuing haloperidol or don't feel as comfortable sending someone home with a haloperidol prescription.”
Dr. Fontaine hopes to expand that discomfort to all antipsychotics. “These are not benign medications by any means,” he said. “I was reviewing some papers recently showing that the number need to harm to increase one incidence of death in an elderly demented patient is somewhere between 25 and 50, so of every 25 elderly or demented patients leaving the hospital on an antipsychotic that lacks an indication, there's a decent chance that one of them will die because of that medication.”