Published in the March 2014 issue of Today’s Hospitalist
WHILE HOSPITALISTS CERTAINLY see their share of patients with altered mental status, they may find it challenging to distinguish among different altered states and to decide when and how to treat various etiologies.
In a recent presentation at the management of the hospitalized patient conference held last fall at the University of California, San Francisco (UCSF), S. Andrew Josephson, MD, who heads UCSF’s neurohospitalist program, walked hospitalists through a workup of both common and rare cases of patients with altered mental status. He focused on “what we neurologists think about when we are called by hospitalists who are stuck” and on diagnoses that doctors often miss.
Dr. Josephson also provided an overview of new thinking and treatment options in certain types of dementia, and he stressed the importance of a strategic, broad-based delirium workup. He also urged hospitalists to stop subscribing to the notion that cognitive impairment is a fact of life among the elderly.
“Fifteen-year-olds don’t get confused when they get a UTI,” said Dr. Josephson. “That’s not normal for 85-year-olds either.”
Dr. Josephson presented a case that, at first, seemed to require only a limited workup because the diagnosis ” dementia “was obvious. The patient, an 80-year-old woman, was experiencing short-term memory loss. She was forgetting her keys, asking repetitive questions and no longer taking care of the family finances.
But when the hospitalists attending the conference were asked to vote on what they’d choose next for the workup, only 9% picked Dr. Josephson’s suggested answer: to order head imaging. The other options included ordering TSH, vitamin B-12 and RPR testing; referring for formal neuropsychiatric testing; not ordering testing but starting the patient on donepezil; and ordering screening labs for delirium.
Why order head imaging? “Everybody with a suspected dementia requires a head image,” Dr. Josephson pointed out. “Some end up having bilateral subdural hematomas or a brain tumor, so you want a non-contrast head CT or MRI in any patient with cognitive decline.”
While memory loss is the first sign of Alzheimer’s disease, patients with other forms of dementia “such as frontotemporal dementia “initially exhibit different symptoms. In frontotemporal dementia, psychiatric changes often include apathy, behavioral problems and anxiety.
“Many patients mistakenly are described as having a primary psychiatric disorder,” said Dr. Josephson. “At age 65 or younger, frontotemporal dementia is just as common as Alzheimer’s.”
Hospitalists also encounter vascular dementia, which typically starts with impaired executive function, and dementia with Lewy bodies, where patients have visual hallucinations and/or Parkinsonism symptoms. For reasons not fully understood, a dual diagnosis of Alzheimer’s and vascular dementia is “strangely common,” Dr. Josephson said. He suggested that patients being treated for Alzheimer’s might benefit from aggressively modifying their vascular risk factors.
First-line treatments for Alzheimer’s are still cholinesterase inhibitors and memantine, but recent studies have weighed in on long-term treatment. A study in the March 8, 2012, New England Journal of Medicine found that in patients with advanced disease, stopping donepezil had significant downsides and that patients’ disease worsened compared to those who continued taking the drug.
But evidence backing the use of memantine is more ambiguous. One recent study found benefits when the drug was used in combination with a cholinesterase inhibitor; another did not.
“It’s a little bit of a dealer’s choice,” Dr. Josephson said. “If patients are declining on a maximum dose of a cholinesterase inhibitor, I add memantine, although the evidence remains mixed.” Both drugs, he cautioned, “are modestly effective at best.”
As for therapies to help Alzheimer’s patients with behavioral problems, studies keep coming up empty. “Antipsychotics probably don’t work in large trials nor do cholinesterase inhibitors,” he noted. “We’re in the dark here as to what to do.” He uses antipsychotics at very low doses, and only if patients pose a risk of harm to themselves or others.
While delirium is common in inpatients, Dr. Josephson suggested that both the workup that physicians use to find the cause and treatments vary more than they should.
In thinking about delirium, “it is important to remember that the condition is marked by the relative acute onset of a cognitive change over hours or days,” Dr. Josephson pointed out. Importantly, clinicians shouldn’t miss a delirium diagnosis when patients have a more hypoactive presentation.
“We all recognize the hyperactive subtype, classically someone experiencing delirium tremens and hallucinating,” he said. “But we miss the little old lady with a UTI who isn’t talking much and is sitting quietly in the corner, but still may be quite delirious.” Patients with this type of delirium may have worse outcomes.
In screening for delirium, the standard questions ” What year is it? Where are you? Who is the president? “aren’t as effective, in Dr. Josephson’s view, as the simple “digits-forward” test. Say three to five random digits, one second apart, then ask the patient to repeat them. The objective is to see how many numbers the patient can remember. (Normal individuals can repeat seven on average.)
“If someone does worse than five, they have a deficit of attention,” he said, “and the most likely cause is delirium.” Digits-forward “is far more useful than a mini-mental state examination, which takes a lot of time but is predicated on the person having normal attention.”
Delirium: “a stress test for the brain”
Dr. Josephson noted that he thinks of delirium as “essentially a stress test for the brain in the hospital that patients have failed. You may treat the UTI or stop the anticholinergic that’s causing it, but you also have to refer patients to an outpatient physician who can figure out the underlying etiology predisposing them to delirium.”
Often the culprit is an unrecognized neurodegenerative disease, which can mimic delirium. At the top of that list is dementia with Lewy bodies. “This is an extraordinarily common diagnosis, especially in the hospital,” said Dr. Josephson. “If you’re not making this diagnosis in patients with neurodegenerative disorders, you’re probably missing a lot.”
And missing the diagnosis leads to inappropriate treatment. Patients with dementia with Lewy bodies are “exquisitely sensitive to antipsychotics,” he pointed out, and may go into a rigid, confused state if given the drugs.
But because these patients have “an extreme deficit of acetylcholine,” they respond to cholinesterase inhibitors. Dr. Josephson treats them with high doses of medications such as donepezil: up to 20 mg daily. “It not only helps their dementia, but also their behavioral problems,” he noted. “This is the form of dementia that we are actually pretty successful in treating for a period of time.”
A deadly diagnosis to miss
Dr. Josephson then considered this case: a 50-year-old woman who exhibited paranoia over a week and aggressive behavior. But her general exam, delirium screen and CT were all normal.
What test should you order? A lumbar puncture. “If the CT and initial workup are negative but the patient is delirious, I have a very low threshold for obtaining a lumbar puncture,” Dr. Josephson said.
This particular case illustrates the reason why: The lumbar puncture found a mild pleocytosis “18 white blood cells and a negative gram stain. Empiric treatment with acyclovir was started. An MRI ordered the next day found bilateral abnormalities in the frontal and temporal lobes. By day five, the HSV-1 PCR in the cerebrospinal fluid was positive.
“HSV-1 meningoencephalitis is a deadly diagnosis to miss,” Dr. Josephson stressed. “If you’re not starting a number of confused patients on acyclovir and then being wrong when the PCR turns out to be negative, you’re not doing it enough.”
In Dr. Josephson’s view, lumbar puncture is underutilized in working up altered mental status. At his center, lumbar punctures are performed quickly if, after imaging, the patient has a suspected CNS infection or unexplained altered mental status.
“In addition to diagnosing infection, it also gives additional useful information for inflammatory conditions like CNS vasculitis, CNS lymphoma and multiple sclerosis,” he said.
Dr. Josephson posed another case of altered mental status that called for an aggressive workup. This one involved a 45-year-old patient with an unremarkable medical history, other than gastric bypass surgery a few months earlier.
The patient presented with confusion, ataxia and eye-movement problems “and was diagnosed with Wernicke’s encephalopathy from thiamine deficiency, another diagnosis that, in Dr. Josephson’s experience, is often missed. Classically found in alcoholics, the deficiency has been largely eradicated because so many foods are fortified with thiamine. Now, the condition is more associated with malabsorption than with deficiency, so it’s a risk for gastric bypass patients.
“That’s important to remember because you want to treat them with intravenous thiamine, not oral thiamine,” he said. Hospitalists who suspect the disorder, which rarely presents with the “classic triad” of confusion, ataxia and eye-movement abnormalities, should start patients on at least 100 mg of IV thiamine a day.
“It’s cheap, it’s harmless and it can save somebody,” he said. Untreated, the condition often progresses to Korsakoff syndrome, an irreversible dementia state.
Dr. Josephson cited another case in which an 86-year-old patient’s general exam and delirium screen were normal, even though she had a prior history of stroke. But on neurological exam, the patient was somnolent and unarousable to voice, and she exhibited fine nystagmus in all directions. Should the hospitalist order an MRI, lumbar puncture, blood cultures, urinary porphyrins or EEG?
His first choice was EEG monitoring, although he’d also consider a lumbar puncture. The EEG found nonconvulsive status epilepticus, another condition that’s often missed and is more common than hospitalists suspect.
He cited his own study in the April 2013 issue of Mayo Clinic Proceedings, which included 1,100 patients admitted to the UCSF medicine and neurology services. The patients subsequently underwent an EEG for unexplained altered mental status or spells.
“What we found surprised us,” said Dr. Josephson. “Seven percent had seizures and 18% had epileptiform discharges, which means they probably just had seizures. I would argue that EEG is, like lumbar puncture, a very underutilized part of the altered mental status workup.”
And if you can’t order an EEG in your hospital? “If you have someone who’s altered and you’ve done everything else in terms of workup, it is not unreasonable to transfer that patient,” he said. “This is a clearly treatable cause of altered mental status.”
Bonnie Darves is a freelance health care writer based in Seattle.
Basilar artery thrombosis
THE 30-YEAR-OLD MAN presented with no previous medical history, but he’d experienced six hours of stupor that progressed to coma. S. Andrew Josephson, MD, who heads the neurohospitalist program at the University of California, San Francisco (UCFS), said that physicians need to keep two things in mind when seeing patients in this type of state: “They’re in a coma for one of two reasons: either their brainstem is injured or their bilateral hemispheres are injured.”
Speaking about altered mental status at a UCSF conference, Dr. Josephson said he focuses his neuro exam on the cranial nerves because they “all live in the brainstem,” he explained. “If the cranial nerves are abnormal, then the localization is brainstem. If the cranial nerves are all normal, it’s bilateral hemispheres.” In up to 95% of cases, he said, “the answer is bilateral hemispheres.”
But this patient had vertical bobbing movements of both eyes, indicating cranial nerve involvement. Unfortunately, the first CT done in the ED was read as normal. A few hours later, however, “the CT scan is repeated with a CT angiogram and now the patient was found to have a hypodensity in the pons and the cerebellum,” Dr. Josephson pointed out.
“The artery missing at the bottom of the CT angiogram is the basilar artery. This patient has a basilar artery thrombosis” “a type of stroke that can present with altered mental status.
“It’s one you do not want to miss,” he added. “It carries essentially 100% mortality.” Basilar artery thrombosis is common from cardioembolic disease in Afib patients and in young people with a vertebral artery dissection.
The good news is that embolectomy can work. With patients in a coma who have cranial nerve abnormalities, “make sure they don’t have a basilar artery thrombosis by imaging the cerebral vessels with an MRA or CTA.”
In patients not in a coma, another indication of posterior circulation dysfunction is asymmetry of cerebellar signs. “When someone does great on finger-nose-finger on one side but poorly on the other, that’s a real red flag,” said Dr. Josephson. “It suggests a lesion in the cerebellum or the brainstem until proven otherwise.”
Delirium and environmental interventions
DESPITE THE WIDE RANGE of delirium etiologies “drugs, infection, out-of-kilter electrolytes” optimal treatment is fairly standard, once you’ve addressed or removed the precipitant.
But a key treatment strategy, said S. Andrew Josephson, MD, who heads the neurohospitalist program at the University of California, San Francisco, is consistently underused: systematic environmental interventions. These include turning off lights at night and not waking patients for vitals; mobilizing patients and making sure they don’t nap during the day; getting rid of physical restraints and Foley catheters; and reorienting patients, with “orienting” personal items like hearing aids and glasses.
“Spend time with your hospital, if you haven’t already, thinking of systems-based approaches, because these interventions make a huge difference,” said Dr. Josephson, speaking to hospitalists at a UCSF conference last fall. “They are much more effective than medical management.”
Antipsychotics should be a last resort and used only when needed to prevent harm to the patient or staff, Dr. Josephson said. If indicated, give antipsychotics in small doses at bedtime.
And always avoid benzodiazepines. “These are really delirium provoking,” he said, “unless the patient has delirium caused by alcohol or benzodiazepine or barbiturate withdrawal.”
At the same time, Dr. Josephson made a strong case against viewing delirium as merely a common side effect of being hospitalized, particularly in older patients. Instead, after you diagnose and treat patients’ delirium, make sure you refer them to a neurologist or geriatrician, “who can figure out if you’ve just unmasked an underlying problem with the brain.”