Spotting delirium


Delirium is common among older hospitalized patients and can lead to poor outcomes. However, the condition is often missed during rushed morning rounds. Patients may seem simply sleepy or physicians might catch them in a good moment, despite fluctuating periods of delirium.

The 10-item Confusion Assessment Method (CAM) is often recommended for diagnosing delirium, but although a recently validated short version, the 3D-CAM, takes only 3 minutes to administer, even this may be too time-consuming for busy hospital practice.

“One of our challenges as hospitalists is that the methods of diagnosing and validating delirium in research trials don't work very well in daily practice,” said Ethan Cumbler, MD, FACP, medical director of the Acute Care for the Elderly service at the University of Colorado Hospital in Aurora. “In general we don't have 30 minutes to complete a dedicated cognitive assessment, so what practicing hospitalists really need is a rapid assessment tool.”

Recently developed screening tools that can be administered quickly at the bedside have the potential to provide that. A 2-item tool that asks patients to name the months of the year backwards and the day of the week, for example, accurately identified 93% of delirium cases in a study published in the October 2015 Journal of Hospital Medicine (JHM).

More user-friendly screening tools may encourage clinicians to routinely screen for delirium and raise awareness about the signs and symptoms, which include inattention, disorganized thinking, and altered level of consciousness. However, experts caution that every test has both advantages and limitations.

“When there is limited time, a brief assessment makes sense, but realize that you may sacrifice some diagnostic accuracy,” said Jin H. Han, MD, MSc, associate professor of emergency medicine at Vanderbilt University in Nashville, Tenn., who studies delirium. “These newer brief methods are appealing for hospitalists, but they have not been validated in large studies.”

Delirium often overlooked

Despite the prevalence of delirium among older hospitalized adults, it is estimated that more than 50% of cases are missed in the course of routine care. That's largely because the symptoms of hypoactive—as opposed to hyperactive—delirium are not always apparent.

“Delirium can be obvious if a patient is hallucinating, nonsensical, or striking out,” said Dr. Cumbler. “But it can also be subtle. For instance, in hypoactive delirium the patient may simply seem very sleepy.”

That sleepiness could easily seem normal during early morning rounds when the physician may ask a few brief questions requiring “yes” or “no” responses, he said. Symptoms of hypoactive delirium, including sluggishness or reduced motor activity, often don't fit into physicians' mental schema of what delirium looks like, which tends to be based on the hyperactive variety.

Another complicating factor is the fluctuating nature of delirium, said ACP Member Heidi Wierman, MD, medical director of geriatric programs at Maine Medical Center in Portland. The condition tends to wax and wane at different times of day or even from hour to hour.

“If you catch a patient at a moment when they are feeling pretty good, you may not recognize the symptoms,” she said. “That's why it's important that everyone on the health care team knows what to look for in terms of inattention or altered level of alertness.”

Identifying delirium is crucial because it has been associated with a variety of serious adverse events both inside and outside of the hospital.

For example, Dr. Wierman and colleagues recently conducted a retrospective study exploring the relationship between delirium and falls among older patients at Maine Medical Center. They found that falls were associated with evidence of delirium in 73% of the patients studied.

The findings suggest that improving delirium recognition may help reduce the number of falls and lead to better outcomes, such as shorter length of stay, said Dr. Wierman, whose study was published online Jan. 12 in Psychosomatics.

Patients with low scores on a brief delirium assessment also risk increased length of stay, higher in-hospital mortality, and greater need for skilled nursing after discharge, according to another study published in the October 2015 JHM. The study used a quick 1-minute screen that included the months-of-the-year-backwards question and the modified Richmond Agitation and Sedation Scale (mRASS), an observational tool that measures arousal.

“Once someone has delirium, we can't turn it off quickly,” said Fred Rubin, MD, FACP, a geriatrician and chief of medicine at the University of Pittsburgh Medical Center Shadyside Hospital. “If we can quickly identify patients at increased risk, we may be able to intervene to mitigate those risks.”

Pros and cons of brief screens

Experts emphasize that very brief screens should not be used as the basis for diagnosis, but they can help clinicians identify patients at high risk for delirium who may warrant further assessment. The key is zeroing in on the cardinal clinical features of delirium: altered levels of arousal and inattention.

When researchers in the first JHM study analyzed 20 individual items from the 3D-CAM to determine which had the highest sensitivity and specificity for identifying delirium, they found that the months-of-the-year-backwards question, which screens for inattention, by itself identified 83% of delirium cases.

“Inattention really comes out when patients attempt to count the months backward,” said Dr. Cumbler, who uses the 2-question screen every day with elderly patients on his rounds. “Not infrequently, I have been surprised to discover new inattention flagged by this screen that I wasn't going to pick up based on my routine morning rounding.”

The mRASS, which takes as little as 15 seconds, can be an effective way to quickly assess level of arousal. The test asks patients to answer an open-ended question while a clinician observes for 10 seconds and scores the answer on a scale of −5 (coma) to +4 (combative).

A study led by Dr. Han, published in the June 25, 2015, Academic Emergency Medicine, found that an abnormal mRASS score had 82% sensitivity and 85% specificity for detecting delirium in patients ages 65 or older admitted to the ED.

“In the emergency department, time is a huge issue because we are evaluating a large number of patients in a short period,” noted Dr. Han. “As a result, these newer brief tools are very appealing.”

The chief advantage of brief screening instruments is that they are easy to integrate into routine clinical care, said ACP Member Edward Marcantonio, MD, section chief for research in the division of general medicine and primary care at Beth Israel Deaconess Medical Center and professor of medicine at Harvard Medical School in Boston.

However, he added that “it's hard to have a really brief instrument that is also extremely accurate” because convenient bedside tools tend to be less precise than more in-depth assessments.

It's also important to note that brief screening tools are often not effective for patients with dementia, noted Dr. Cumbler. An inability to recite the months of the year backwards, for example, would not be a reliable indicator of delirium in a patient with moderate Alzheimer's who could not answer that question even on his or her best day.

“We should recognize that the very brief screen functions pretty well for most patients,” he said. “But it does not make a diagnosis of delirium in patients with underlying dementia. For that we need a more complete assessment.”

To identify delirium in dementia patients, Dr. Cumbler talks to family members about what constitutes normal for a particular patient. For example, one of his patients could not say the months backwards but was able to sing the ABCs. Dr. Cumbler used that baseline to assess the patient's cognitive state postoperatively.

“When she became delirious, she lost the ability to sing the ABCs,” he said. “After a few days, her cognition returned to baseline and she regained the ability to perform that test.”

Establishing patients' baseline cognitive function is especially important in cases of dementia, added Dr. Wierman.

“If someone has dementia, it's often assumed that any cognitive impairment is due to their dementia even if we haven't established a baseline,” she said. “But we often miss delirium because we assume people are at their baseline when they are not.”

Establishing best practices

Even brief, user-friendly screening tools may be difficult to integrate into practice if left to the discretion of individual clinicians, noted Dr. Marcantonio. It's best to establish protocols about which screening methods to use and how to follow up on the results.

“For screening, using the same assessment tool across the board allows you to integrate it into care systematically and to more easily assess for acute change by looking at how the patient does over time,” he said. “Screening in itself is not necessarily a good thing unless you have a plan of what to do with cases you identify in terms of treatment or next steps.”

Maine Medical Center's policy is to screen every patient in the hospital at least once a day using the Short CAM, said Dr. Wierman. Everyone on the team, including nursing assistants, physical therapists, and social workers, is trained in how to spot the signs of delirium and the importance of intervening early.

“I worked for years trying to get nurses to do the Short CAM on our Acute Care for the Elderly unit, but until it became an institutional standard it wasn't always consistent,” she said.

New tools that take advantage of technology, such as smartphones and other handheld devices, may also boost clinicians' use of screening tools.

For example, researchers in China recently tested a new smartphone application aimed at making it easier for nurses to detect delirium in the ICU using the CAM-ICU, a version of CAM for nonverbal patients. About 100 ICU nurses tested the app and reported that it was easier to use than the regular CAM-ICU, according to results reported at the World Chinese Health Informatics Symposium, held in São Paulo, Brazil, last August.

The app is programmed to automatically retrieve patients' baseline data for comparison and print out the results of the screening. Nurses reported that the user-friendly interface was helpful in guiding them through the assessment while allowing them to simultaneously observe patients' conditions.

Another group of researchers is working on a smartphone version of the Edinburgh Delirium Test Box, or DelBox, a handheld, battery-powered device designed for use at the bedside. Using the DelBox, clinicians can administer various tests of attention, such as a word-building task in which 3- to 6-letter words are displayed on the device 1 letter at a time separated by an interval of a few seconds. Patients are then asked to select the correct word.

In the mobile version, patients take the same tests on a smartphone app. Pilot studies suggest that the mobile version performs as well as other screening methods and may be a more practical and convenient option for clinicians, according to the Edinburgh Delirium Research Group, based at the Royal Infirmary of Edinburgh in Scotland, which developed the DelBox.

“Most clinicians carry a smartphone now,” said Dr. Han, “so apps are an exciting advance in delirium assessment.”