Home Clinical The high-yield neuro exam

The high-yield neuro exam

January 2012

Published in the January 2012 issue of Today’s Hospitalist

To get the most out of a screening neurologic exam, hospitalists should think big, according to S. Andrew Josephson, MD, associate professor of clinical neurology at the University of California, San Francisco, and director the center’s neurohospitalist program. By that, he means that hospitalists should devote the limited time they have to elements of a neurologic exam that “test lots of the brain” at one time.

That strategy should give doctors more usable information than working though most of the litany of neurologic tests they were taught in training, each of which concentrates only on a specific brain area and function.

High yield: language testing
One example of thinking big, Dr. Josephson told hospitalists at a session on neurologic exams at UCSF’s management of the hospitalized patient conference last fall, is language testing. He recommends that hospitalists administer simple language tests on every patient because, taken together, the tests reveal a great deal about a large area of tissue in the dominant hemisphere.

Keep in mind, Dr. Josephson said, that language and speech are not the same. Speech is what a person does to produce sounds, and a problem with that, such as slurred speech, is generally a cranial nerve problem. Language, by contrast, is made up of three elements: fluency, comprehension and repetition. All of those reside in the frontal and temporal lobes of the (usually) left hemisphere.

To test fluency, Dr. Josephson explained, just talk to the patient; taking a history is an excellent opportunity. “Are they speaking in fluent phrases, or are they having difficulty getting the words out? Are they speaking in broken phrases, or having a difficult time coming up with words?”

After assessing fluency, Dr. Josephson said he tests comprehension. He suggested using a three-step command that crosses the midline.

“I say: ‘When I say ‘Go,’ take your right thumb, touch your left ear and stick out your tongue.’ If patients can do it well, that’s impressive,” Dr. Josephson noted. Not only do you learn whether patients can comprehend the three steps, which suggests intact receptive language, but you “see their right hand’s motor function and learn a bit about their proprioception. And because you are asking them to cross their body’s midline, you are learning their corpus callosum is intact, allowing the patient’s two hemispheres to talk to each other.”

Next, he said, test repetition. He recommends asking patients to repeat a simple phrase. At UCSF, he said, they use the same sentence for everybody: “Today is a sunny day in San Francisco.”

With information on fluency, comprehension and repetition, said Dr. Josephson, hospitalists can avoid misdiagnosing a person as having schizophrenia or just being “confused” while they really have a Broca’s, Wernicke’s or other type of aphasia.

A person with bad fluency and repetition but good comprehension has Broca’s aphasia, he pointed out, whereas a person with poor comprehension and repetition but good fluency may have Wernicke’s.

Language testing, Dr. Josephson noted, is a key way to quickly screen patients for mental status issues. He also recommends screening every patient with “digits forward.” (See “Diagnosing delirium,” below.)

“That’s all I do” for a screening mental status examination, he said. “Don’t spend time on the mini-mental status examination. Don’t ask patients what floor they are on or what date it is. None of that is particularly high-yield.”

Testing cranial nerves
Although neurologists can spend 45 minutes or more testing cranial nerves, Dr. Josephson said busy hospitalists shouldn’t bother with that level of detail. And many of the tests commonly used for cranial nerves “are worthless,” he said. “Stick out your tongue,” for instance, or “Shrug your shoulders” gives you very little information. That’s because they focus on a single cranial nerve that rarely is injured, making such tests not worth a hospitalist’s time.

Instead, for a patient who is awake and attentive, Dr. Josephson said that the exams that deliver the most information deal with pupils, extraocular movements, facial strength and, surprisingly to some, visual fields. Like language, visual fields utilize “a huge area” of the brain, making it possible to test “a lot of tissue very quickly.”

First, Dr. Josephson explained, lean over the patient, putting your face in front of his or hers (because people tend to focus on faces). Then tell the patient to look at your nose and point to which of your hands is wiggling. Make sure you put your hands halfway between the two of you, and test all quadrants of vision.

If patients can correctly identify the moving hand, their visual field is normal “and you’ve spent only 10 seconds. If they struggle to identify the correct hand in one quadrant, Dr. Josephson said, have patients cover one eye so you can test each eye individually.

If the patient is too confused to do this test, Dr. Josephson recommends a simple “blink-to-threat” test. Wait until patients fixate on your face, then bring one finger on each side suddenly toward them; it should elicit a blink. But be careful not to blow air toward patients’ eyes at the same time you are bringing a finger toward them. That could elicit a corneal response, which could be mistaken for a normal visual field.

The next highest yield cranial nerve test, he said, involves facial strength. Dr. Josephson asks patients to “Show me your teeth,” and then “Raise up your eyebrows.”

The third cranial nerve test he recommends utilizes extraocular movements and covers cranial nerves III, IV and VI at one time. The best test for extraocular movements, said Dr. Josephson, is to have patients follow your finger with their eyes while you draw a big letter H in the air. For patients too inattentive to follow a finger, Dr. Josephson said to use money “particularly a $20 bill.

“If you get a $20 bill out of your wallet,” he said, “they tend to follow it pretty well. It must be some primitive response preserved in confused patients.”

Motor exam
Finding the right screening tool to test for upper motor neuron-type weakness can be tricky for some physicians, Dr. Josephson explained. But those tests are important because when patients are weak, hospitalists have to figure out whether the weakness is due to an upper motor neuron problem in the brain or spinal cord.

The good news is that screening for upper motor neuron-type weakness takes only a minute and a half. “When upper motor neurons are injured, they fail in a predictable pattern, with distal muscles being weaker than proximal muscles and extensors of the upper extremities being weaker than the flexors,” said Dr. Josephson. “There are no exceptions.”

The three quick-and-easy tests that he recommends are pronator drift; fast finger movements and toe taps; and testing the strength of one muscle in each of the four extremities, focusing on those muscles that are weakest in the upper motor neuron pattern.

If these tests are normal, Dr. Josephson said, “patients aren’t weak from a stroke, a brain tumor or a spinal cord problem. Period.” Beware, however, of a common test done in many emergency rooms: the hand grasp. “It takes a long, long time for a big, strong flexor muscle to become weak,” he pointed out. “That makes this an extremely poor test for weakness.”

But you do want to spend time distinguishing between upper and lower motor neuron weakness, he added, because making that determination will inform what you do next. If weakness is due to upper motor weakness, for instance, the next step is to order imaging of the brain and the spinal cord. If the weakness is localized to the lower motor neurons, the problem is in the peripheral nervous system, so imaging won’t be any help.

Sensory testing: use sparingly
Dr. Josephson’s general recommendation regarding sensory testing “if patients haven’t come in with a specific sensory complaint “is to use it “sparingly” or “skip it altogether. It’s very subjective, and it’s unlikely to give you much information.”

If you do think sensory testing is needed, Dr. Josephson advises testing only the longest nerves. Most of what you’ll encounter is peripheral neuropathy, which typically affects the longest nerves first. First, test a patient’s toes, and if they’re fine, probably stop there, he noted. If you carry around a tuning fork, you can test both vibration and temperature.

Using “light touch” probably is of no use, nor is pinching patients unless they are in a coma and you’re trying to get them to move based on deep pain. When a pinprick would be useful, the easiest tool is the sharp point of a tongue depressor broken in half.

One of the most useful sensory tests is the Romberg, said Dr. Josephson, but it is often done incorrectly and misinterpreted. To be “positive,” he noted, the patient needs to fall, but only after the test is done in this manner:

First, stand up. Second, bring feet completely together. Third, close eyes.
If the patient falls before step 3 or can’t get to step 3, you cannot say there has been a positive Romberg.

Reflex symmetry
Like sensory testing, Dr. Josephson said he thinks testing reflexes is probably not necessary most of the time as part of a hospitalist’s screening exam.

For hospitalists who want to test reflexes, it’s symmetry that counts, not absolute value. “What matters is comparing one side to another on an individual patient,” he said. For example, if you test the right biceps, then immediately test the left biceps, not the other reflexes on the right arm. If you are testing the patellar reflex, don’t watch the leg kick out, but instead watch the quadriceps; that’s the muscle you are testing.

Moreover, said Dr. Josephson, probably the most important reflex to test in older patients is the hardest one to elicit: the ankle reflex. “People over age 70 or 75 really should not have ankle reflexes,” he said. “If they do, they may have a superimposed upper motor neuron process such as cervical spine disease.”

Coordination exam
As with reflexes, what matters with coordination is asymmetry. “Bilateral dysfunction is often benign and drug-related,” said Dr. Josephson. But if somebody does well on a finger-nose-finger test on one side but poorly on the other, that’s usually a clue that they may have a cerebellar lesion.

In the end, Dr. Josephson said, if you have only one minute, probably the most useful test is to watch the patient walk.

“If somebody can walk just fine, you’re in really good shape,” he said. That’s because gait illuminates motor, sensory, cerebellar and extra-pyramidal ability. “I know it can be very difficult” logistically in the hospital, Dr. Josephson added, “but if you can walk someone, it is tremendously useful.”

Deborah Gesensway is a freelance writer who covers U.S. health care from Toronto.

Diagnosing delirium

ASKING PATIENTS to repeat back to you progressively longer strings of numbers is one of the most efficient ways to screen for delirium, said S. Andrew Josephson, MD, director of the neurohospitalist program at the University of California, San Francisco (UCSF).

“I would do ‘digits forward’ on every single patient,” he said, speaking at the UCSF’s management of the hospitalized patient conference this fall. “It’s much more high-yield than asking patients what the date is or where they are, and tremendously more helpful than the mini-mental status examination.”

In the 20 seconds that “digits forward” takes, Dr. Josephson added, “you are going to pick up a number of people who are delirious that you wouldn’t have otherwise thought of “the more hypoactive subtype, the elderly lady with pneumonia who is just not talking that much but is pleasantly confused.” In addition, he said, the test can tell you which patients with known dementia have superimposed delirium.

This is how he recommends doing “digits forward”: Say to the patient, “When I say ‘go,’ I want you to repeat back to me the following numbers: 3, 9, 2.” If patients can repeat them back, try it with four different numbers, then a five-digit string, a six-digit string and so on.

“The average American can do seven digits forward,” Dr. Josephson said. “If you have a digit span forward of less than five “four, three, two or one “you have a deficit attention, and that’s the neuropsychological hallmark of delirium.”

While it doesn’t matter what numerals you use, Dr. Josephson said he recommends always using the same ones, like your spouse’s phone number or the last four digits of your social security number, just to make them easier for you to remember.

Examining patients in a coma

For patients in a coma, doctors need a strategy for testing cranial nerves, said S. Andrew Josephson, MD, director of the neurohospitalist program at the University of California, San Francisco. For these patients, he said, the pupillary exam is “probably the most important” because it can tell you if the cause of the coma is in the brain stem or because the patients’ bilateral hemispheres are dysfunctional.

Because cranial nerves III through XII all live in the brain stem, if the patient’s pupils are uneven, “you know it’s got to be a structural problem rather than a metabolic problem,” said Dr. Josephson.

He offered some caveats about looking at pupils:

    • If the pupil doesn’t react, make sure you did the test correctly. Turn off all the lights in the room and use a big, bright flashlight. If you’re going to say that the pupil doesn’t react, “you have to make sure your light stimulus is adequate,” Dr. Josephson said.
    • Check that the patient isn’t taking a drug like Atrovent. Ipratropium nebulizer is the most common cause of a blown pupil in the hospital.
    • Be aware that a large, dilated pupil likely means that the patient’s brain stem is not irreversibly injured. “That is not necessarily an ominous sign,” Dr. Josephson said. “If you are running a code and you look down and the guy’s got big, dilated, blown pupils bilaterally, do not stop the code. Epi alone usually leads to large pupils.”

In patients with brain death, their pupils are midsized, he pointed out, between 3 mm and 5 mm.Beyond looking at pupils and seeing if the patient “blinks to threat,” testing cranial nerves in a comatose patient requires some different techniques. One that works, Dr. Josephson said, is the doll’s-eye maneuver.

Hold the patient’s eyelids open, turn the head to one side and hold it there. If the brainstem is working, said Dr. Josephson, the eyes will slowly drift back toward the midline. It may take some time, so make sure you wait between 10 and 15 seconds. If the patient has an endotracheal tube, hold it as you turn the head to avoid an accidental extubation.

Another worthwhile test is the corneal reflex, touching the cornea with a piece of cotton to see if both eyes blink. One common mistake is touching the sclera (on the white part of the eye), not the cornea (over the colored part of the eye). Touching the sclera will not elicit the same blink reflex, Dr. Josephson pointed out.

“Remember the definition of coma: Eyes are closed, not awake, not arousable, but aware,” said Dr. Josephson. He reminded hospitalists that “all comas end up one way or the other. After a week or two, the person either progresses to brain death or their eyes open up. Comas don’t last for more than a couple of weeks. The eyes eventually open.”