What do you do when headache appears clinically attributed to intracranial hypotension but imaging and opening pressure are normal or not obtainable?

Regarding the IHS classification and diagnostic criteria, it is important to remember the following:

1. The initiative came from Glaxo Pharmaceuticals in order to make it possible for the company to pursue the development of sumatriptan for migraine abortion globally. The company also paid for the first edition.
2. The classification with its diagnostic criteria was never meant to be employed in clinical practice; it was developed for use in clinical trial research only.
3. It is not evidence-based. Evidence-based medicine was invented in the early nineties and the first edition of the classification was developed in the mid-eighties.
4. The people who developed it were not necessarily clinically experienced in headache management. For example, I myself, a clinician, did not want to participate in it, because I did not think that we had enough knowledge of headaches in general and migraine, in particular, to classify either in a clinically meaningful way. I foresaw that such a classification with its arbitrary diagnostic criteria would be used beyond its original intention, that is, in clinical practice and other types of research. I thought and still, think this to be detrimental both to patient care and to furthering our understanding of headaches and migraine.

If one is presenting with clearly positional headaches where my clinical suspicion is high for spontaneous intracranial hypotension, I would start with an empiric lumbar blood patch. I may consider more than one empiric lumbar blood patch. If no improvement is seen, then my next step would depend on what type of imaging was done. If prone CT myelogram showed meningeal diverticula without a clear leak, then I would still consider a targeted blood patch at those levels. May need to do more than one level.

In my experience, what is referred to as type 3 leaks (CSF venous fistulas) are the most difficult to diagnose. At my institution, we have done lateral decubitus dynamic CT myelogram. If we still do not have a clear diagnosis, then I would offer a patient options of treating a primary headache phenotype (in most cases its chronic migraine phenotype), or if feasible, a referral to an outside neuroradiology center that may have more expertise. I have in the past referred to Dr. Linda Leithe at Duke.

I like this paper by Carroll et al., PMID 38820488, and many other good headache clinicians. If you think a headache has clinical characteristics of low CSF pressure headache, you should treat with lying flat and any other non-invasive techniques for a few days. If the headache persists, an epidural blood patch should be done and results are often good, although not always complete or permanent. Repeat patches can be helpful. I have often had patients improve very quickly and sometimes completely.

Although I think ICHD-3 has been very helpful for doctors' communication about patients throughout the world, it is not perfect and sometimes you have to treat it clinically.