Your first patient of the shift is a 19-year-old male with “the worst headache ever.” His mother states she’s quite worried because he “never gets headaches.” He had some slight nuchal rigidity, but is afebrile. And as such, you arranged a non-contrast CT Head, which was found to have a normal without signs of intracranial hemorrhage. Now, 12 hours after onset of his headache, and after discussing the risk and benefits of a lumbar puncture to this patient, he and his mother ask, “So what happens if my spinal tap is not normal?”
You pause, and reflect on the evidence…
Using Pubmed, three separate searches were performed:
(2) “lumbar puncture”
(3) “subarachnoid hemorrhage”
From the results of this literature search you find:
One study showed that out of 660 samples to be analyzed, 28 patients were found to have CSF with xanthochromia, 5 of these 28 had a subarachnoid hemorrhage  . Three other studies found xanthochromia to have a PPV of 6.8%, 8.3%, and 8.7% respectively, with a 100% NPV [2,3,4]. A fifth study looked at 15 patients with positive findings for xanthochromia found on CSF spectrophotometry who also received a CT angiogram of their brain. Nine of these fifteen CTA’s were found to be normal, and one CTA had a coincidental unruptured aneurysm. Thus, even spectrophotometry only had a PPV of 33% .
With that said, many labs do not use spectrophotometry to measure xanthochromia- including my own. This is important, as there is not an insignificant amount of intra-observer disagreement amongst individual samples, with one study finding disagreement among 19.2% of samples .
So how long does it take for us to detect xanthochromia?
The release of oxyhemoglobulin from lysed red blood cells in CSF can be detected within 4-10 hours of a bleed on spectrophotometry, while the conversion of oxyhemoglobulin to bilirubin has been reported to be 9-10 hours. The conversion to bilirubin gives xanthochromia its yellow color. It has been thought that one must wait 12 hours for xanthochromia to develop based off one study of 111 patients in which an LP was not performed for at least 12 hours who all had xanthochromic CSF . An additional study showed xanthochromia persisted in 32/32 patients at 2 weeks from onset, 20/22 patients at 3 weeks, and 10/14 patients at 4 weeks .
You walk away for a minute to review the evidence. You then explain to the patient that the LP is a screening tool. If it is normal, we push the probability of a diagnosis that has significant morbidity and mortality to extremely close to zero. An abnormal LP increases the probability of the diagnosis, but only to about 8% – but this is certainly high enough to warrant further investigation given the potential diagnosis.
Resolution of the Case
Fortunately, after metoclopramide, they feel well, and their LP revealed no WBC’s, no xanthochromia, their CSF lactic acid level was normal, and they are happy to be safely discharged.
You explain to the patient & their mother that an abnormal LP increases their risk of having a life threatening process. You stress that an abnormal LP does not guarantee that they have a subarachnoid hemorrhage, but rather, that you will have to get a specialist involved to discuss what the appropriate next step would be- and that the patient may be required to stay in the hospital at that point. Fortunately, after metoclopramide, they feel well, and their LP revealed no WBC’s, no xanthochromia, their CSF lactic acid level was normal, and they are happy to be safely discharged.
1. [ref id=”25766384″ link=true]
2. [ref id=”25439758″ link=true]
3. [ref id=”23619131″ link=true]
4. [ref id=”24839296″ link=true]
5. [ref id=”24087797″ link=true]
6. [ref id=”24448178″ link=true]
7. Williams A. Xanthrochromia in the cerebrospinal fluid. Practial Neurology. 2004; 4: 174-175.
8. [ref id=”PMC1031927″ link=true]