Conference Review – Primary and Secondary Headaches

Thanks again to Dr. Eisinger for his Conference reviews!

Small group summary Wed 9/21/16 – Primary Headache

Case 1:

18 yo F presents with episodic headaches that occur 3-4 times per month. The headaches started 5 years ago, but seem to have increased in severity. They are associated with nausea and she sees waves of light about 30 minutes before the pain begins. Light and noise make it worse. It is usually on the left side and throbs, lasting several hours. She can usually get it to go away with ibuprofen and rest. Her mom and sister have had similar headaches before. On exam she is uncomfortable and sits with the blanket over her face, but has normal VS, normal physical and neurologic exam.

  • What historical and physical exam features are important to obtain for patients presenting with cephalgia?

We are looking for evidence of emergency causes of HA which include subarachnoid hemorrhage or other intracranial bleeding, meningitis, intracranial abscess, obstructive mass lesions, giant cell arteritis, glaucoma, etc. Most important historical features from ED standpoint include: prior HA Hx and similarity of current episode to prior, onset (ie gradual vs peak intensity at onset or “thunderclap”), fevers, localizing neuro Sx, Hx of IVDA, Hx of cancer, Hx of trauma. Other important features include signs of ICP (worse with bending over or Valsalva, n/v, blurry vision), HA which awakens from sleep or is severe on awakeningin the morning.

As for exam, a thorough neurologic exam is crucial with special attention to any localizing deficits, the Pt’s overall appearance, the vital signs, stigmata of IVDA, and meningismus (Kernig & Brudzinski signs). Always walk these pts to look for cerebellar issues. If possible look for papilledema.

  • What diagnostic evaluation for headache should be pursued in this case presentation?

Given fairly typical migraine Sx with no alarm features, duration of 5 years, and nml exam, emergency causes of HA are extremely unlikely and this patient can be treated with no additional workup. As always, a urine hCG would not be a bad idea, especially if you anticipate the potential for going farther down your migraine cocktail treatment algorithm (eg Depakote, etc).

  • Discuss treatment options for this patient and their potential side effects.

Most people have an algorithm that they follow and this varies significantly provider to provider. Start by asking the Pt what has helped them in the past and if it is reasonable, go with that (assuming it’s not some drug that starts with a D that they can’t quite remember the name of…). In the absence of prior successful regimens, think about causes of HAs that may have specific treatments such as high flow O2 for cluster HAs, carbamazepine or phenytoin for trigeminal neuralgia, etc. Also ask about contraindications to any of your first line therapies, such as CKD or HF which preclude NSAIDs, or a history of severe EPS reactions or long QTc which would relatively contraindicate antidopaminergic agents. Barring any of these specific indications, I usually progress as follows:

  1. If the HA is mild, offer oral regimen including an NSAID and antiemetic. Avoids them needing an IV, typically not sedating, often effective (especially if they haven’t tried anything else), and may get them out of the department sooner.
  2. Give a cocktail of ketorolac, diphenhydramine, prochlorperazine, and a saline bolus. Metoclopramide can be used instead of prochlorperazine. There is pretty good data that diphenhydramine itself does not give any relief for HAs and is not effective in preventing the akisthesia associated with metoclopramide, although several studies does show it does help mitigate this with prochlorperazine. It does also help to put people to sleep which often provides relief in my experience.
  3. Haldoperidol: watch for long QTc or EPS type reactions
  4. Magnesium: can cause some cardiac arrhythmias or hypotension but pretty rare. Start with 2g over 30 minutes.
  5. Steroids: most studies look at decadron. I usually use a dose of 10mg iv. There is decent data that this also helps prevent HA recurrence over the next several days. Avoid using in brittle diabetics. Fairly low risk of AVN and other serious SE’s with a single dose.
  6. Valproic acid: avoid in liver failure. Can cause pancreatitis, SJS, and blood dyscrasias. Definitely not for pregnant females.

Be aware that many of the above medications are sedating and make sure your patient is ok to drive when you discharge them.  

There are a million different medications that have been studied for HA and I’m not aware of any guideline directed algorithm for approaching them. Here is a summary of recommendations from the most recent American Headache Association review:  

Should offer: metoclopramide, prochlorperazine, sumatriptan; dexamethasone (for recurrence prevention)

May offer: acetaminophen, ASA, chlorpromazine, dexketoprofen, diclofenac, dipyrone, droperidol, haloperidol, ketorolac,

May avoid: diphenhydramine, hydromorphone, lidocaine, morphine, octreotide, valproate

No recommendation: dexamethasone (for acute Tx), dihydroergotamine, ergotamine, ketamine, lysine, magnesium, meperidine, nalbuphine, promethazine, propofol, tramadol, triamcinolone, trimethobenzamide,

  • How would you counsel this patient upon disposition?

My counseling typically includes a blurb about how there does not appear to be an emergency cause of HA at this time but that does not mean that the workup can stop here if they have not previously been evaluated (ie disclaimer in case they wind up having a brain tumor). I would then make sure they understand the importance of following up with a PCP or neurologist (depending on how much investigation/outpatient mgmt has been tried so far) to discuss starting a prophylactic regimen and possibly an abortive like a triptan. I then counsel them on OTC meds to use if the HA returns (ie Tylenol, NSAIDS). Finally, I tell them to keep a HA diary to look for triggers, try to avoid triggers, and bring the diary to their outpatient appt. Common triggers include: analgesic rebound, caffeine, EtOH, other drugs or medications, lack of sleep, stress, menstruation, and dehydration. I was surprised to learn that despite how often patients report it, mild-moderate HTN (< 200/120) has been found to be a very rare cause of HA.

Case 2:

A 36 year old male presents to the emergency room during your night shift. He experienced a severe headache that woke him up shortly after falling asleep. He reports similar headaches 3 other nights this week that resolved after an hour without treatment and he is concerned because they are continuing and. He reports sharp, severe frontal/periorbital headache on the right side. He also reports rhinorrhea and are very uncomfortable increased sweating during the period that he has pain. The patient has no past medical history and is on no medications but uses tobacco and typically drinks one glass of red wine every evening. On exam, patient is well developed, well nourished, AOx3. He appears slightly uncomfortable. VS are normal. His head exam is atraumatic and unremarkable except for scleral injection and tearing of the R eye, neck is nontender, and heart and lungs exams are normal. No cranial nerve deficits were noted and the rest of the neurologic exam was normal.

  • What diagnostic evaluation of headache be pursued in this case presentation?

This patient presents with the classic symptoms of cluster HA. Once again, assuming none of the red flag symptoms discussed above, he can be treated empirically without further workup in the ED. However, many of these patients will present with some autonomic Sx such as Horner’s on the affected side and in the absence of prior workup or known diagnosis of cluster HA disorder, probably warrants some imaging. Specifically, this can be seen in carotid artery dissections so should be considered especially in a pt with recent trauma or cervical spine manipulation (ie chiropractic)

  • Describe options for treatment and for prophylaxis of this condition.

There is strong evidence for the use of high flow O2 (12-15L/m) in these patients and by itself will abort the HA in about 90% of patients. If this is ineffective, the next best option is sumatriptan which is effective in about 90% of cases, though it should be noted that about half of patients with have unpleasant SE’s like chest pain, n/v, flushing, itching, etc. Other HA treatments discussed above could be tried but are less well-supported. It should also be noted the natural Hx of cluster HA is a short duration of attack (between 15 min and 3hrs) and it will usually get better with time and then recur. Thus prophylaxis to prevent recurrence or break the cluster is important medications frequently used for this purpose include verapamil, short courses of steroids, methysergide, and topiramax.

  • Discuss the role of regional anesthesia for treatment of this condition.

There is growing evidence that regional nerve blocks can be effective in the acute and transitioning (subacute) mgmt of various HA disorders. Some authors advocate for palpation of all the pericranial nerves (eg greater occipital, lesser occipital, auriculotemporal, mental, supraorbital, suprotrochlear, infraorbital, etc) for areas of tenderness and performing a block on the corresponding nerve. Most commonly blocked in the case of cluster HA are the greater and lesser occipital nerves which has been shown to produce a pain free period of 3-70 days in 31-85% of individuals. This may be an effective modality for breaking a cluster. Some authors advocate injection of a mixture of lidocaine and corticosteroid, though the latter should be avoid in injecting facial structures due to possible negative cosmetic effects on the skin. The landmarks for injection of the greater and lesser occipital nerves are shown below.

Additional question: Describe options for prophylaxis for the above conditions.

Although prescribing prophylactic medications for HA is outside the scope of emergency medicine, it is good to be aware of the medications which patients may be on for this purpose. A number of medication classes are frequently used for HA prophylaxis with varying results in the literature. These include beta blockers (propranolol), non-dihydropyridine Ca channel blockers (verapamil), tricyclic antidepressants (amitriptyline), SSRIs (sertraline), SNRIs (duloxetine), antiepileptic drugs (valproate, carbamazepine, topiramate).

Secondary Headaches

Word doc with extra picutures – sm-gp-summary-secondary-headache

Case 1:

18 yo F with history of morbid obesity presents with daily headaches that started 2 months ago and have become constant for the past one week. She has never had this type of headache before and that the pain is diffuse and bilateral. They are associated with mild nausea and occasional blurred vision on both sides. She notices brief ibuprofen but the headache never goes away completely. She feels relief when she is upright and feels that it worsens when laying down to sleep at night. On exam she appears comfortable. She has normal VS and normal standard physical and neurologic exam.


  • What additional physical exam component is important to obtain in this patient presentation?


In addition to a thorough neurologic exam, the most physical exam step is to perform fundoscopy to assess for papilledema. The patient in this scenario is presenting with classic signs and symptoms of increased intracranial pressure – HA which is worse when laying down/bending over, coughing or Valsalva, blurry vision, confusion, gait abnormality, nausea and vomiting. The least invasive way to confirm elevated intracranial pressure is identify papilledema on fundoscopy. The optic nerve sheath directly communicates with the subarachnoid space and as such transmits pressure changes in the brain to the back of the eye where it can be visualized as a blurring of the normally sharp borders of the optic disc. While extremely important, this physical exam step is often very difficult to perform using nondilated direct fundoscopy. As such there are several options available: 1) dilate the eye using drops such as tropicamide or atropine (watch for contraindications such as a history of glaucoma); 2) use a pan-optic ophthalmoscope; 3) consult ophthalmology (preferred at our institution); 4) use ocular ultrasound. A recent metaanalysis published in the Annals of EM found a SN of 96% and SP of 93% for the use of ultrasound in detecting elevated ICP. The procedure is done by filling the orbit with a copious amount of ultrasound gel, examining the eye in axial section, identifying the optic nerve sheath, and measuring its diameter at 3mm posterior to its junction with the globe. A value of >5mm is considered abnormal in adults, 4.5 mm in children, and 4mm in infants.




  • What diagnostic evaluation for headache should be pursued?


As mentioned, the above patient is presenting with signs and symptoms of elevated intracranial pressure. Although this patient’s presentation is most consistent with idiopathic intracranial hypertension (IIH, pseudotumor cerebri), structural causes of elevated ICP must be ruled out in a patient without a prior established Dx of IIH. At OSU, we have a defined algorithm for the workup of these patients, both with a prior Dx of IIH and without. For patients without a prior Dx, the first step is ophthalmology evaluation for assessment of papilledema. For most patients, the first step will be a noncontrasted CT of the head to look for blood or obvious mass lesions. This is followed by an MRI brian if one has not been performed in the past 6 months, with MRA and MRV as well if this has never been done before. If papilledema is graded as mild to moderate and brain imaging reveals findings consistent with IIH, the Pt’s HA can be managed symptomatically and they may be referred for outpatient ophthalmology follow—up and LP on an outpatient basis. If the papilledema is severe, then LP is performed in the ED, neurosurgery is consulted, and the patient is admitted to ophthalmology. In pts with a known Dx of IIH WITHOUT a shunt, those with mild-moderate papilledema can be managed symptomatically and discharged with optho and/or neurosurgery f/u in 3 days and LP on outpatient basis. Those with severe papilledema should have MRI brain (if not done within 6mos), LP in the ED, and should be admitted to ophthalmology with neurosurgery consult. For those with an established Dx of IIH and a shunt in place, those with low pressure HA symptoms OR high pressure Sx and mild-moderate papilledema should have an XR shunt series performed and may be managed symptomatically in the CDU or on an outpatient basis. If high pressure Sx and severe papilledema are present, shunt series (if not done in past month) and MRI brain (if not done in past 6mos) should be done, LP should be performed (if Pt does not have a lumbo-peritoneal shunt), and neurosurgery should be consulted for shunt adjustment.


Here is the link to the OSUMC clinical practice guideline:


  • Which consultants may be helpful in evaluation and treatment of this patient’s complaint?


As discussed above, ophthalmology and neurosurgery consultations are often necessary in the management of these patients.



  • Discuss treatment options for this patient.


Though the pathophysiology of IIH is poorly understood, there is a strong association with obesity. As such, weight loss either through lifestyle modifications, pharmacologic management, or bariatric surgery is the cornerstone of treating the underlying condition. These patients should have an established relationship with a PCP for assistance in managing weight and consideration of bariatric surgery referral and should be referred to a nutritionist as well. Beyond weight loss, management is focused on lowering intracranial pressure. In the acute setting, this can be accomplished through high volume LP. Other ED treatments include standard iv headache cocktails which may be effective for some patients. Narcotic medications should be avoided due to lack of demonstrated benefit, high rate of rebound HAs, and potential for dependency, increased ED visits, etc. For maintenance therapy, many patients will be placed on acetazolamide, a carbonic anhydrase inhibitor diuretic which has equivocal evidence for benefit and poor tolerance in many patients. Other medications such as furosemide have been used as well but have even less evidence to support them. Surgical procedures such as optic nerve fenestration may provide some benefit. CSF diversion through ventriculoperitoneal, lumboperitoneal, ventriculoatrial, etc shunts is also commonly performed, though the evidence for benefit is modest at best.


Case 2:

A 20 year old female college student presents to the emergency room and reports experiencing a severe headache. She was seen 2 days ago for fever, headache, and neck stiffness in another ED and underwent lumbar puncture at that time that she was told was normal. Her fever has resolved, but she reports onset yesterday of a headache that has changed in character from aching to severe, diffuse pain. It is worse when sitting or standing and only tolerable when laying flat in bed.  Physical and neurologic exam were normal as well.


  • What diagnostic evaluation of headache should be pursued in this case presentation?


This patient is presenting with a story concerning for post lumbar puncture headache (PLPH) and does not require any further specific diagnostic workup. The important features of her Hx include recent LP and low pressure HA, defined by worsening with upright position and improvement with laying flat. Although post lumbar puncture meningitis has been documented due to introduction of skin flora into the CSF during the performance of LP, it is very uncommon and in the absence of compelling symptoms such as fevers or profound meningismus, repeat LP to rule it out need not be performed. Other post-LP complications to be aware of include cerebral herniation or development of intracranial hemorrhage from tearing of bridging vessels due to intracranial pressure shifts; again these complications are extraordinarily rare. The pathophysiology of PLPH is low intracranial pressure caused by dural defect allowing for ongoing leakage of CSF at the LP site.


  • What are the risk factors for this type of headache?


Risk factors for PLPH include number of LP attempts, volume of CSF drained, and size and type of needle used. Though classically taught that maintaining supine position for 1hr following LP can decrease the incidence of PLPH, this has not been bourne out in the literature and a recent Cochrane review recommends against routine use of supine positioning following LP, instead advocating for whatever position is most comfortable for the pt. Ways to decrease the incidence of PLPH include using the smallest caliber needle possible (a 25 or 22g needle if possible), as few attempts as is necessary, positioning of the bevel of the needle UP (towards the ceiling when Pt is in lateral decubitus position), taking the smallest volume of CSF necessary for the planned diagnostic testing, and using an atraumatic needle such as the Sprotte needle, rather than the Quincke which is commonly included in LP kits (see photo below). It should be noted however that many clinicians find the LP more difficult using the Sprotte needle.



  • Describe options for treatment of this condition.


There is moderate evidence for the use of iv caffeine in the acute treatment of PLPH. A recent systematic review showed an absolute risk reduction of 0.61 for the use of IV caffeine which is typically dosed at 300-500mg. Oral caffeine is well-absorbed and so discharged patients can be given Rx for 300-500mg either daily or bid for the next several days. The mechanism of action is thought to be related to vasoconstriction of dilated intracranial blood vessels. Based on this presumed mechanism, sumatriptan has also been advocated and has been reported effective in a few case reports but does not have strong prospective data to support its use for this indication. Other measures aimed at replenishing the lost CSF such as iv fluids, DDAVP, and ACTH have not been demonstrated to be effective. Typically headache cocktails may also be tried though they have not been specifically studied.  In PLPH lasting longer than 72hrs and refractory to conservative medical mgmt, consider anesthesia consultation for blood patch in which a quantity ranging from 2-30cc of the patient’s own blood is injected into the epidural space at the level of the LP. The theoretical origins of this practice come from the observation of decreased incidence of PLPH following bloody taps. The injected blood forms a clot which seals off the hole in the dura, allowing for reaccumulation of CSF and resumption of nml ICP. Success rates of beween 70-98% have been reported.


Additional Question: How should patients undergoing lumbar puncture be counseled prior to the procedure?


As with any procedure, the Pt should be counseled on the indications, risks, benefits, and alternatives to the procedure. I typically begin by explaining that the LP is a very common ED procedure which is typically well-tolerated with major complications being rare. I typically explain that the procedure is much scarier to patients than it is dangerous and that, while tolerance of the procedure varies considerably between patients, typically the most uncomfortable part is the injection of local anesthetic, which is similar to when a dental procedure is performed, and that the majority of my patients do not find it significantly more painful than PIV placement. I then go through the common complications before moving on to rare but serious complications. The most common complications would be pain, discomfort from the positioning, post lumbar puncture headache, and unsuccessful procedure requiring repeat attempt under fluoroscopy. This is followed by bleeding and infection for which I recommend PLT >50k and INR <1.6 before performing the procedure, and use of sterile technique (particularly face masks, as the most common organisms implicated in post-LP meningitis have been oral flora). Finally I mention the less common complications such as transient leg weakness, radicular back pain, injury to the spinal cord (only in Pt’s with undiagnosed tethered cord), cerebral herniation (rare – only about 1% even in pts with known elevated ICP), and aortic injury. In the case of suspected meningitis, the alternative to the procedure would be to complete an empiric full course of IV ABX with all the unnecessary risk of ABX side effects as well as the risk of partially treated meningitis with sterilized CSF Cx and inability to isolate a resistant organism if LP is performed at a later time.