This week in Morning Report, we discussed a few challenging cases of headache. This is a very common presenting concern (comprising up to ~15% of visits to family doctors and 2-4% of visits to the ED) and actually poses a very underappreciated burden of disease – in fact, the WHO estimates that migraine on its own (just one of many headache types!) represents the 6th highest cause of worldwide years lost to disability.
In addition to representing a vast burden of disease, headaches also span the spectrum from benign to devastating, and can often be very challenging to diagnose. As internists, we may actually not encounter headache presentations outside of meningitis/encephalitis because so many of these patients are first assessed by our FM and ER colleagues, but you will certainly encounter these in ambulatory medicine (not to mention your admitted patients on the ward). All the more reason for developing a solid approach to headache! Let’s recap our key learning from this Morning Report.

1. First, have a mental schema for classification of headaches
The way to break down the hundreds of different causes for headache is to have a solid mental classification or scaffolding upon which you can build differential diagnoses. We discussed dividing into Primary vs. Secondary Headache Disorders, which is consistent with the International Classification of Headache Disorders (ICHD-3).
Primary Headaches are easy to remember – there are just a few major categories. Secondary Headaches are a much larger list, but can be broken down into a few big categories — these are all summarized for you in a flowchart below, along with a few pearls and clang associations to remember what goes in each category! Click the image to enlarge.
2. Be aware of Red Flags and ‘can’t miss’ headache diagnoses
We talked about headache red flags, which you should be familiar with from medical school. These include:
- Older age (> 50)
- Systemic symptoms such as fever
- Neurological deficits
- Abrupt onset (NB: Thunderclap headaches are often misunderstood – the common definition is a sudden onset headache that reaches peak intensity within 1 minute)
- Onset with exertion or precipitated with sneezing, coughing, exercise, Valsalva
- History of immunocompromised state (e.g., HIV, immunosuppressants)
- History of cancer
- Pregnancy
- Progression
In an Emergency Room or acute care consultative setting, it is important to recognize a few high-yield associations that can help you recognize etiology of headaches and prompt a directed workup. These are nicely summarized in this UpToDate table, below (Click to make larger):
3. A Few Pearls on Subarachnoid Hemorrhages and ‘Traumatic LPs’
We spent some time talking about Subarachnoid Hemorrhage — an entity usually managed in the ED, but important for all internists to recognize because of variability in presentation and a few known challenges in diagnosis. Specifically, let’s recap a few pearls:
- Be aware of the Ottawa Subarachnoid Hemorrhage Rule. The sensitivity of this rule is 100% (i.e., it is a RULE OUT rule), but the specificity is ~15%.
- Know that there is a misconception that patients with SAH always look sick – in fact, up to 40% may have no neurological deficits, which can lead to under-recognition and underdiagnosis
- The general guidance is that you can only rule out a SAH via CT head if the patient presents within 6 hours of headache onset. This is because after 6 hours, RBC breakdown leads to underdetection of bleeding. There is a lot of debate around this topic, particularly given increased resolution of CT scans in recent years – but as a rule of thumb, after 6 hours, consider an LP to assess for RBCs and xanthrochromia to rule out SAH. There is a fantastic recap on this Emergency Medicine blog that I strongly recommend.
- Finally, what do you do when you perform an LP and see RBCs — is this due to true bleeding, or because you punctured a little venule along the way (i.e., a traumatic tap)? This is a challenging situation to encounter, and is discussed extensively here.
- Read this NEJM Journal Watch – a prospective multicenter study of 1739 patients found that RBC count < 2000 x 10^6 and absence of xanthochromia were sufficient to exclude 100% of SAH.
I hope you enjoyed this Morning Report! Thanks for your participation — as usual, please direct all questions and comments to the CMR!