April Fools!

So much going on in this episode, it’s out of control. It’s a 3-parter, all in 1 episode. To celebrate April Fools Day, I’ve selected 3 topics that may fool you. So don’t be. And check out the episode.

Part 1: Killer headache.

Question: If the head CT is negative for subarachnoid hemorrhage, do I need the LP?

The answer is…maybe?

Think of it this way, when should you expect an aneurysmal subarachnoid hemorrhage? Like Dr.

A “normal” non-contrast head CT. Subarachnoid hemorrhage anybody?

Khankhanian said way back in episode 8, aneurysmal ruptures are like an explosion and create a thunderclap headache.  Often there is nausea and vomiting, and sometimes cranial nerve palsies are observed, and in severe cases there is loss of consciousness or even coma…if the patient survives.

To answer the question, technically yes there can still be subarachnoid blood in a normal head CT. I think 2 of the biggest contributors to a “normal” appearing HCT in aneurysmal SAH are the generation of CT scanner and the timing of the scan. The newer, third generation CT scanners available at most hospitals in the US nowadays, are really not likely to miss an intracranial SAH.  The second contributor to a “normal” HCT is the timing of hemorrhage. If you have an intracranial aneurysmal SAH, and you get the scan within a few hours of onset, there is ALMOST no way you should miss the blood. The sensitivity of a head CT for identifying subarachnoid blood is 99.999% if it is obtained within the first 6 hours of symptom onset. Ok, maybe not 99.999%, but its pretty high.

Part 2: I just can’t fight this feeling any longer.

Question: If a patient sounds like he or she has a peripheral neuropathy, can the EMG/NCS be normal?

Like a head CT in subarachnoid hemorrhage, the timing of the study can dramatically impact the results. A nerve conduction study can be totally normal within the first 4-6 days of nerve injury, even if the nerve is transected.

But now let’s say the patient has a sensory disturbance that’s lasted longer than a week, and the nerve conduction study is STILL normal. Well, now we have to consider the limitations of your testing. If all the patient has complained of is pain or temperature disturbance, these signals are carried by small myelinated (A-delta, or B) fibers and unmyelinated (C) fibers—and these fibers are not measured during a routine nerve conduction study. You can only pick up the large myelinated somatic cutaneous afferent fibers using nerve conduction study. So what do you do?

In general, there’s a lot of overlap between these fibers that carry pain and temperature information and autonomic afferents and efferents. The reason this is clinically relevant, is that it would be extremely unusual to have a pure neuropathy that only affects the C fibers which convey pain and temperature information. More often, the patient will have a ‘small fiber’ neuropathy affecting both the nociceptive C fibers and the autonomic C fibers. So testing a patient for autonomic dysfunction may be telling here.

If your suspicion of a small fiber neuropathy is confirmed, then you can proceed with working up based on that differential diagnosis.

Differential diagnosis of a small fiber neuropathy from Preston and Shapiro.

Part 3: Twist and shout.

Question: If the EEG is normal, can the patient still have epilepsy?

More specifically, if the EEG was normal on a patient actively experiencing a clinical event—meaning a behavior that clinically resembles a seizure—does that exclude the diagnosis of epilepsy? [At this point in the episode I talk about a personal experience I had with a patient with seizures and a “normal” EEG.]

There is a simple law of mathematics, that if X=Y then X always equals Y (the law of equality). If a clinical event means no seizure on EEG, then we make a leap of faith that the same clinical event means no seizure when off EEG. But is that a fair assumption?

Obviously not, or I wouldn’t have asked you. First of all, this patient could have had an electrographic seizure at the time of a prior clinical event, and we just didn’t have her monitored on EEG. Our decision-making is limited by the available data we have, and we are pretty confident that X will always equal Y. A non-epileptic event has always been and will always be non-epileptic. And you can reassure yourself that is the case because many times non-epileptic events will fail to improve with anti-epileptic drugs, and they will often improve with psychiatric intervention. And while the EEG may be limited in NOT showing you something clinically significant, there are A TON of artifacts inherent to EEGs that may fool you into thinking the patient has abnormal electrographic activity. See here:

Common EEG artifacts.

Second of all, there is a huge overlap between psychogenic nonepileptic events, PNEE (or PNES), and epilepsy. As many as 30% of patients with epilepsy have non-epileptic events. =

Third of all, the scalp EEG can be totally normal in a patient who is experiencing an electrographic seizure. Technically, you need 6 square centimeters of abnormally firing cortical tissue before you can detect a seizure on the routine EEG. Also, the standard scalp EEG set up dedicates a large number of superficial electrodes to the temporal lobes—which are thought to be most epileptogenic—and very few electrodes to the frontal lobes. So you can miss some frontal lobe seizures on a routine scalp recording, even during a clinical event, unless you perform a dedicated study. Who knew?

I think the take away message here is to recognize the limitations of your diagnostic techniques. Trust your exam and your instinct when evaluating these more unusual patients. Just because your test may be “normal”, don’t be fooled by the patient’s legitimate symptoms.


[Jim Siegler]

Featured image from https://www.flickr.com/photos/atlih/2073685927/ under a Creative Commons license.

You may also like...

Leave a Reply

Your email address will not be published. Required fields are marked *