Skip to main content
Patient Care

Grand Rounds: Headache, Part I—Differential Diagnosis and Assessment

Grand Rounds is a monthly blog series developed by EMS World and FlightBridgeED that will feature top EMS medical directors exploring the intricacies of critical care in EMS practice. In this installment FlightBridgeED Chief Medical Director Jeffrey Jarvis, MD, begins a two-part review of headaches.

My colleagues Bryce Johnson and Dr. Remle Crowe and I published a recent study describing the EMS assessment and treatment of benign headaches.1 The TL/DR version: Of more than 7 million EMS calls in our data set, 1% were for benign headaches. EMS frequently did not assess a patient’s pain and addressed it even more rarely. When it did, personnel most often inappropriately used opioids.

So, how should we treat benign headaches? That’s for the second part of this series, where we’ll take a deep dive into the evidence-based management of headaches. In this article we’ll cover the differential diagnosis of headaches, how those different diagnoses present and how to tell them apart, which ones are “can’t miss” diagnoses, and what the “red flag” signs and symptoms are.

Case Study

Let’s start with a case. EMS is called for a 35-year-old female with a headache. You find her in her bedroom, curled up in bed. The room is dark, and she is wearing sunglasses. She tells you this headache has been slowly getting worse over the past three days but has gotten intolerable over the past 3–4 hours. It is localized on the right side of her head, most intensely behind her right eye.

It is an intense, throbbing pain. Nothing seemed to have provoked it, and she can’t remember doing anything special when it started. Bright lights and loud noises make it worse. She has had similar headaches in the past, but this is the most intense she’s experienced. She’s gone to the ED before but never not been able to drive herself. Today she has been nauseated and vomiting. The patient felt like she needed to call EMS because she has been throwing up for the past hour and doesn’t feel she can drive.

She denies any recent trauma, fever, or radiation of the pain into her neck. She confirms this headache came on gradually over several hours to days without a sudden onset at maximal intensity. She has not had any recent chiropractic manipulation or surgical procedures. She didn’t experience any abnormal smells or sights before this headache came on. Her prior headaches have usually come roughly once every other month but do not seem to occur in batches. She has no visual loss. She denies myalgias, arthralgias, or recent weight loss.

She has a history of hypertension for which she takes lisinopril but otherwise is healthy. She takes oral contraceptives but no other medications or supplements and has no allergies.

Her vital signs are BP 180/100, HR 82 regular, RR 18 and nonlabored, temp 98.2°F, SpO2 99% on room air, and EtCO2 30. Your exam reveals a well-nourished woman with a normal body habitus who appears her stated age and looks uncomfortable. She is alert, oriented, and speaking in complete sentences without slurring. Her neurologic exam is reassuring and without any focal deficits, including bilaterally reactive pupils with clear corneas. Her balance and coordination are intact. She has full range of motion of the neck without discomfort. Her lungs are clear to auscultation bilaterally, and she has a nontender abdomen.

Worst First

I realize this is a precise case presentation. This description is what I would dictate for my emergency department note on this patient. It’s an example of how a history of present illness (HPI) should be used. It isn’t a chronologic or verbatim description of what the patient tells you; it should be your well-thought-out presentation of the patient’s condition with information that shows you considered an appropriate differential diagnosis. If we in EMS want to insist (correctly) that we are medical professionals, we should make sure we communicate in the same way as our physician colleagues. With that in mind, let’s take this presentation apart and talk about differential diagnoses.

As with the rest of emergency medicine, we need to think worst first. When it comes to headaches, there are some “can’t-miss” diagnoses to consider.2 These include traumatic head bleed, subarachnoid hemorrhage, bacterial meningitis, acute glaucoma, temporal arteritis, cavernous sinus thrombosis, and brain tumor. Once those are considered and felt to be unlikely, you can expand your differential to include the more common and benign causes of a headache, such as migraine, tension, cluster, idiopathic intracranial hypertension, post-LP headache, and trigeminal neuralgia. I recognize most of these diagnoses aren’t covered in usual paramedic school curricula, but just consider this some bonus material that puts the continuing in continuing education. Fortunately, differentiating the actual cause of a benign headache isn’t all that important, but we’ll briefly go over them for completeness.

Traumatic intracranial hemorrhage—Most of these should be pretty easy to pick up because of the trauma. Suspect this in patients with headache and abnormal neuro exam following trauma. The more challenging ones can be the patient with mild trauma on anticoagulants or the young patient with agitation following trauma that we might be tempted to write off as “just drunk.” In the ED we have a saying that agitated drunk trauma patients can go “sleep it off” after their negative CT scan.

Subarachnoid hemorrhage—Most subarachnoid bleeds come from berry aneurysms of the anterior communicating artery, one of the arteries that make up the circle of Willis (“the spider in the brain”).3 Subarachnoid bleeds make up about 10% of all hemorrhagic strokes and are the most common cause of sudden death with stroke. The classic presentation is a “thunderclap” headache described as the worst they’ve ever experienced.

I don’t find the concept of “worst headache” to be all that helpful. Most patients who call EMS for their headache are likely to be worried about not being taken seriously by the medic (or ER doc) and feel almost compelled to say it’s the worst of their life. A much more useful question is, “How quickly did this headache come on and reach its maximal intensity?” A headache that came on suddenly and reached its most severe within minutes, like a thunderclap, is very worrisome for a subarachnoid hemorrhage.4

When I hear this, I think subarachnoid hemorrhage until proven otherwise. Other alarming features are a headache brought on during exertion, including sex. The physical exam may be completely normal but rapidly deteriorate as the size of the bleed expands. I’ve had patients go from completely awake with a normal exam to unconscious and intubated within an hour.

Some subarachnoid bleeds can be what are called sentinel bleeds. These are thunderclap headaches caused by a ruptured aneurysm that quickly coagulates with little bleeding. Patients will completely recover. These headaches are important warning signs that a more devastating bleed may occur in the future. They are our opportunity to diagnose the aneurysm and get it treated before the “big bleed” can occur.

Our role in these headaches is first to suspect it and help the ED physician find it, and second to treat the pain. The response to analgesia is of no diagnostic help, and we shouldn’t be led astray by resolution of the headache.

Subarachnoid bleeds are typically diagnosed with CT scans. The classic teaching is that a normal CT scan needs to be followed by a lumbar puncture to look for blood in the CSF. Modern generations of CT scanners, though, have sufficient resolution to effectively rule out subarachnoid bleeds within the first six hours of onset. These patients with clear scans may not need an LP.5 This is one reason why it is helpful to get patients with thunderclap headaches evaluated rapidly. The severity of subarachnoid hemorrhages is graded using the Hunt and Hess scale.2 A 0 on that scale is an asymptomatic, unruptured aneurysm, and 5 is deep coma.

Bacterial meningitis—Meningitis is inflammation of the meninges and can be from bacteria, viruses, or fungi. Fungal meningitis is relatively rare in the U.S., but viral meningitis is prevalent. Many of us have had it and likely didn’t realize it. Bacterial meningitis, on the other hand, is more severe and can be life-threatening. The most common organisms are Neisseria meningiditis, Haemophilus influenzae, and Streptococcus pneumoniae.6 Fortunately these have become pretty rare with the advent of vaccines against these bacteria. I haven’t seen a case of bacterial meningitis in years.

The hallmark findings of meningitis are fever and signs of infection. Other classic findings are nuchal rigidity or neck stiffness. Kernig’s sign is positive when extension of the knee with the hip flexed at 90 degrees causes pain and stiffness in the hamstrings. Brudzinski’s sign is the other one we’re all taught but rapidly forget; it is positive when the patient involuntarily flexes the hips/knees when you passively flex the neck.

Diagnosis is through the evaluation of CSF, where the classic findings are elevated white blood cells and protein with low glucose. Treatment consists of rapid identification and early antibiotics.

Acute glaucoma—Acute angle-closure glaucoma often presents with a headache, but on closer questioning it’s eye pain. It is caused by an occlusion of the canal of Schlemm. The canal of Schlemm drains the aqueous humor of the anterior chamber of the eye. When it gets obstructed the intraocular pressure builds up, leading to symptoms. The classic presentation is unilateral eye pain with vision disruption, cloudy cornea, conjunctivitis, and a fixed midrange pupil.7 The classic HPI is a sudden onset of pain when walking from a bright area into a dark room, like a theater. The diagnosis is confirmed by measuring intraocular pressure (IOP). Treatment is aimed at decreasing IOP with topical medications like beta blockers, opening the canal up by constricting the pupils, and reducing the production of fluid with oral medications like acetazolamide or diuretics.

Temporal arteritis—Temporal arteritis, or giant cell arteritis, is a vasculitis or inflammation of vessels.8 It most commonly affects the aorta and temporal arteries. If left untreated, it can lead to permanent blindness. It most commonly affects older women and presents with gradual onset of unilateral throbbing/aching headache, often localized to the temporal region. It is often associated with jaw claudication or pain with chewing. There may also be tenderness over the temporal artery. It can be associated with other systemic conditions such as polymyalgia rheumatica, which is an aching stiffness of the proximal joints that is usually worse in the morning and gets better throughout the day. I’ll suspect this diagnosis based on the right clinical picture and evidence of systemic inflammation on labs like an elevated CRP and sedimentation rate. While the diagnosis is confirmed with a temporal artery biopsy, the key treatment is the early use of steroids to prevent eye loss.

Cavernous sinus thrombosis—Cavernous sinus thrombosis is essentially a DVT of the brain. It is a blood clot in the venous system of the brain, as opposed to the arterial system we usually associate with a stroke. We should suspect this in patients with a history of DVT/PE or risk factors for clots like oral contraceptive use and smoking. The clinical picture is of a stroke, but ocular symptoms may be the most obvious. The diagnosis is made through advanced imaging, and treatment is roughly the same as it is for DVT, i.e., anticoagulation to prevent clot propagation.

Brain tumor—Tumors of the brain have a bimodal age distribution: primary tumors typically affect younger patients (under 50). In comparison, secondary tumors affect older patients and have spread from metastatic cancer, with breast and lung being the most common. Onset is usually gradual, as would be expected from a tumor that takes time to grow. We should suspect this in patients with known cancer and those with altered mental status or focal neurologic deficits. For most headaches, though, Arnold Schwarzenegger was probably right: “It’s not a tumor.”9

More Common Causes

So those are the bad causes of headaches. Fortunately they’re relatively uncommon. Let’s talk about some of the causes of benign headaches, the ones we’re most likely to see.

Migraine—Many people will use the term migraine to mean bad headache, even if it isn’t an actual migraine. For the most part, this is a distinction without a difference and certainly isn’t worth arguing with patients.

There are specific diagnostic criteria for two types of migraines: classic and common.10 Classic migraines are associated with an aura, whereas common migraines have no aura. The textbook presentation of migraine is a gradual onset over 4–72 hours of unilateral, pulsating, or throbbing headache. Remember, though, presentations can and do vary. They are often associated with nausea and are made worse by light (photophobia) and loud noises (phonophobia).

Your neurologic exam should be completely normal. If you find any neuro deficits, you must assume it is something other than a migraine. However, there are diagnostic dilemmas called complicated migraines.11 These present like strokes and can have hemiplegia or visual loss. These are sometimes impossible to differentiate from acute ischemic stroke and end up being treated with thrombolytics. The CT scan on a patient with migraine will be normal, but that doesn’t help much because CT scans are expected to be normal in early ischemic strokes, too.

Tension headache—Tension headaches are the most common cause of recurrent headaches. Most of us have had these. Tension headaches feel like a tight band is constricting the head like a vise. There may be palpably tight neck muscles. Most patients with healthy coping skills do not ever need to call EMS for tension headaches. They rarely interfere with normal daily activities and are typically well treated with over-the-counter NSAIDs or acetaminophen.

Cluster headache—Cluster headaches are unique in a couple of ways. First, they are the only headache type that is more common in men than women. They are most common in young adults, are associated with smoking and alcohol use, and by definition occur in clusters. Patients may go months without any headaches and then have a cluster of many during a single day. The pain seems to be unilateral and in the distribution of the trigeminal nerve. They may be associated with Horner syndrome: ptosis, miosis, and anhidrosis. The other unique feature to cluster headaches is they can effectively be treated with oxygen.2

Trigeminal neuralgia—Speaking of the trigeminal nerve, trigeminal neuralgia, or tic douloureux, is caused by an inflammation of the trigeminal nerve. It presents as a sudden, sharp, intense, lancinating pain of one side of the face. It lasts just a few seconds at a time but recurs often. These can be incapacitating. Treatment is with antidepressants, antiseizure medications, Neurontin, and ultimately ligation of the nerve.

Idiopathic intracranial hypertension—Idiopathic intracranial hypertension is the new name for pseudotumor cerebri. It is classically seen in young obese women and is associated with oral contraceptive use. It is caused by either an overproduction of CSF or inadequate drainage. We used to treat these with multiple lumbar punctures, but these have proven to be of short-term benefit only (like less than eight hours). Patients usually respond better to typical headache medications.

Post-LP headache—The final headache I want to mention is the post-LP headache. Not surprisingly, these occur after lumbar punctures. They’re thought to be caused by a persistent leak in the dura mater leading to CSF hypotension. Headaches worsen when standing and are bilateral and throbbing. The standard treatment is a blood patch, where venous blood is taken from an IV and inserted into the dura to form a clot. Fortunately, most of these headaches also respond to more typical treatment.


We just covered a lot of information. There are four key things I’d like you to take away from this.

  1. Treatment isn’t associated with diagnosis. How a patient responds to treatment should not impact the actual diagnosis at all.
  2. We should approach treatment the same regardless of cause.
  3. Most headaches are benign.
  4. The dangerous causes of headaches include subarachnoid hemorrhage, meningitis, acute glaucoma, temporal arteritis, tumors, and cavernous sinus thrombosis.

Part 2 of this series will cover the treatment of headaches.


1. Jarvis JL, Johnson B, Crowe RP. Out-of-hospital assessment and treatment of adults with atraumatic headache. JACEP Open, 2020; 1: 17–23.

2. Russi CS. Headaches. In: Marx JA, ed. Rosen’s Emergency Medicine, 7th ed. Philadelphia: Elsevier, 2010.

3. Abraham MK, Chang WW. Subarachnoid Hemorrhage. Emerg Med Clin North Am, 2016; 34: 901–16.

4. American College of Emergency Physicians Clinical Policies Subcommittee (Writing Committee) on Acute Headache; Godwin SA, Cherkas DS, Panagos PD, et al. Clinical Policy: Critical Issues in the Evaluation and Management of Adult Patients Presenting to the Emergency Department With Acute Headache. Ann Emerg Med, 2019 Oct; 74(4): e41–e74

5. Pouryahya P, Haydon R, Meyer A, et al. Utility of lumbar puncture after a normal brain computed tomography scan in patients presenting to the emergency department with suspected subarachnoid haemorrhage: A new more rational approach. Emerg Med Australas, 2020 Mar 20 [epub ahead of print].

6. Brouwer MC, van de Beek D. Epidemiology of community-acquired bacterial meningitis. Curr Opin Infect Dis, 2018; 31: 78–84.

7. Walker RA, Adhikari S. Eye Emergencies. In: Tintinalli JE, Ma OJ, Yealy DM, et al., eds. Tintinalli’s Emergency Medicine. New York: McGraw-Hill Education, 2020.

8. Naik PD, Mollman M. Headache. In: Stone CK, Humphries RL, eds. Current Diagnosis & Treatment: Emergency Medicine. New York: McGraw-Hill Education, 2017.

9. Kindergarten Cop. Universal Pictures, 1990.

10. MacGregor EA. Migraine. Ann Intern Med, 2017; 166: ITC49–ITC64.

11. Blumenfeld AE, Victorio MC, Berenson FR. Complicated Migraines. Semin Pediatr Neurol, 2016; 23: 18–22.

Jeffrey L. Jarvis, MD, MS, EMT-P, FACEP, FAEMS, is the chief medical director for FlightBridgeED, LLC, and cohost of the FlightBridgeED EMS Lighthouse Project Podcast. He also serves as an EMS medical director for the Williamson County EMS system and Marble Falls Area EMS and is an emergency physician at Baylor Scott & White Hospital in Round Rock, Tex. He is board certified in emergency medicine and EMS. He began his career as a paramedic with Williamson County EMS in 1988 and continues to maintain his paramedic license.


Back to Top