Say no to antipsychotics and bed alarms for older patients

Some well-intentioned but misguided interventions may harm geriatric inpatients more than they help.


Some common clinical practices for treating acutely ill older patients may offer no benefit.

Melissa L Mattison MD FACPslashPhoto courtesy of Dr Mattison
Melissa L. Mattison, MD, FACP/Photo courtesy of Dr. Mattison.

Examples of such well-intentioned but misguided interventions include antipsychotics to treat delirium and bed alarms to prevent falls, according to Melissa L. Mattison, MD, FACP, chief of the hospital medicine unit at Massachusetts General Hospital in Boston.

As part of her update in geriatric medicine at Hospital Medicine 2017, Dr. Mattison reviewed the problems with these practices, offered advice on managing medications in the elderly, and made a case for using the Clinical Frailty Scale.

Doing it wrong

She opened with a story familiar to many hospitalists: An 84-year-old patient admitted for pneumonia becomes a little confused, refuses medications, and is better while family visits but agitated and restless again by evening.

To determine if the patient has delirium, Dr. Mattison recommended an ultra-brief two-item bedside test: Ask the patient to list the months of the year in reverse and tell you the current day of the week. According to a study published in the October 2015 Journal of Hospital Medicine, this test has 93% sensitivity and 64% specificity for detecting delirium.

If a patient does have acute delirium, a nurse may suggest ordering an antipsychotic like olanzapine. “Is that the right thing to do?” Dr. Mattison asked.

Not quite, according to a meta-analysis published in the April 2016 Journal of the American Geriatrics Society. Researchers reviewed 19 studies on the use of antipsychotics for prevention and treatment of delirium and found that the drugs had no significant effects on preventing delirium and did not decrease the duration or severity of delirium, mortality, or length of stay in the hospital or ICU.

“What they found was striking … since we use this class of medications regularly to treat patients with delirium,” said Dr. Mattison.

Another study, published in the August 2016 Journal of Hospital Medicine, retrospectively followed about 260 older patients given antipsychotics during a hospitalization with no psychiatric indication. A year later, 41% had been readmitted at least once and 29% had died. “The 41% who were readmitted: Two-thirds of those patients were still taking the antipsychotic they had been started on during their hospitalization,” Dr. Mattison noted. “I sure as heck hope they weren't still delirious up to a year later.”

There were some particular markers of risk: The odds ratio of death was 2.28 for those discharged to a postacute care facility and 3.41 for those with baseline QT prolongation, she noted.

Instead of using antipsychotics to try to treat the symptoms of delirium, identify and treat the cause, Dr. Mattison advised. “Essentially, antipsychotics are chemical restraints,” she said. “They're not the answer for delirium, they're not the answer for dementia, and I would say you should really only use them when you're absolutely pushed to do so for patient safety or comfort.”

Bed alarms, used in many hospitals, are another imprudent approach to older hospitalized patients, said Dr. Mattison. “It is a very common strategy used to prevent falls, and yet, never have bed alarms or chair alarms been shown to be effective at reducing falls or falls with injury,” she said, citing a June 2016 paper in JAMA Internal Medicine. “They contribute to alarm fatigue, and if the patient is ‘with it’ enough, they will be frustrated.”

Furthermore, the alarms are not recommended by the Agency for Healthcare Research and Quality for the purpose of fall prevention in cognitively impaired patients, Dr. Mattison noted. Instead, she recommended attempting to get patients out of bed three times a day with meals and, if possible, also walking in the hallways twice a day. In addition, she said, “I would advocate for going back and talking to hospital leadership, nursing leadership and discussing the use of these alarms and avoid them, especially in cognitively impaired patients.”

How to do better

After explaining the strategies that don't work for older patients, Dr. Mattison offered evidence-based tips that may help.

First, it's useful to know which drugs are candidates for de-prescribing, that is, tapering or discontinuing to minimize the risk of adverse drug events. Docusate is one example. An August 2016 editorial in JAMA Internal Medicine noted significant evidence of its ineffectiveness. “And yet, when they looked at this, 64% of laxatives prescribed were docusate,” she said.

The authors considered the costs, both to patients and to the health care system, associated with docusate. “They actually talk about how disgusting liquid docusate is, evidently. I thankfully have never tasted it, and now that I know it doesn't work, I don't think I ever will,” Dr. Mattison said. “They basically say we're flushing hundreds of millions of dollars per year down the toilet, and we are delaying more effective interventions to relieve constipation.”

An entire class to target for de-prescribing is proton-pump inhibitors (PPIs). “PPIs are associated with increased risk of [Clostridium difficile] infection, especially when used in conjunction with fluoroquinolones,” said Dr. Mattison. She added that a useful Canadian website about deprescribing provides evidence-based algorithms for de-prescribing PPIs, benzodiazepines, antipsychotics, and antihyperglycemic agents.

In contrast, she recommends clinicians increase their use of the Clinical Frailty Scale. It scores patients from 1 to 9 and has been shown to be a valuable tool in assessing and addressing frailty, Dr. Mattison noted. “If you're a 1, you're very fit, you're robust, you're energetic,” she said. “If you're a 9, you're terminally ill.”

In a study published in the June 2016 BMC Geriatrics, researchers found that residents could reliably use the tool to predict patients' functional decline and mortality. Another study, published in the June 2016 Canadian Geriatrics Journal, showed that scores correlated with length of stay on an acute medicine unit.

Researchers divided patients into three groups based on score: 1 to 4 (non-frail, n=21), 5 to 6 (moderately frail, n=38), and 7 to 9 (severely frail, n=64). They found that length of stay was 4.1 days for non-frail patients but jumped to 11.2 and 12.6 days, respectively, for moderately and severely frail patients.

“When we know what our patients' Clinical Frailty Scale score is when they walk in the door or roll in the door, we can help target therapies to the highest-risk patients and certainly consider advance directive discussions,” said Dr. Mattison.

The frailest patients may benefit from proactive palliative care consultations, according to a study published in the November 2016 Journal of the American Geriatrics Society that found reduced use of acute care services with targeted consults of nursing home patients.

“Now, I know most of us don't set foot in nursing homes, but when we get a patient [who is] severely frail … you know he's got limited improvement that's even possible,” said Dr. Mattison. If palliative care and advance directives for future hospitalizations cannot be discussed during the admission, she said, “It may be worth mentioning to your colleague on that transition, either [when] discharging someone or maybe just picking up the phone.”