A 64-year-old woman, with no prior medical history, complained of sudden onset of severe vertigo and vomiting, without headache. Her initial blood pressure in the emergency department (ED) was 170/90 (at about 4:00 PM). Physical examination was limited because the patient's vertigo dramatically worsened when she opened her eyes. Neurological exam was grossly normal—the patient could follow commands and there was no focal weakness. A provisional diagnosis of labyrinthitis was made, and prochlorperazine (Compazine) 10 mg IV was administered. An hour later, however, the patient noted little improvement, and so lorazepam 2 mg IV was given. An hour after that, the patient was signed out to a second attending, who administered atropine 0.5 mg IV. Four hours later, she was signed out to a third attending. At that point (10:00 PM), the physician decided to hold her overnight in the ED and admit her if she was not better in the morning. At 11:00 PM, the patient complained of a headache and was given acetaminophen 650 mg by mouth. An hour later, the patient sustained a cardiopulmonary arrest and could not be resuscitated.

Autopsy revealed that the patient had died of a cerebellar hemorrhage. A subsequent review of the case indicated that the death was potentially avoidable, had life-saving neurosurgery been performed within the first few hours of her ED presentation.

by Joseph M. Furman, MD, PhD

Vertigo is an illusory sensation of motion, of self or surroundings, and can be rotational (ie, a spinning sensation) or translational (ie, a sense of floating upward). The causes of vertigo are many; the common denominator is an abnormality in the vestibular system, which comprises the inner ear labyrinth and the central nervous system (CNS) structures that process signals from the labyrinth. Vertigo, unlike nonspecific "dizziness," is never constant but rather occurs as either a single episode or recurrent episodes. Vertigo can occur spontaneously (eg, as the result of vestibular neuritis or Ménière's disease), or can be provoked, as in benign paroxysmal positional vertigo.(1) When severe, vertigo is often accompanied by nausea and vomiting, as it was in this case.

When patients experience vertigo for the first time, especially if associated with nausea and vomiting, they usually seek medical attention immediately, typically in an emergency department. The causes of vertigo encountered in the ED setting are given in the Table. Most patients presenting to an ED with vertigo can be managed conservatively with vestibular suppressants and antiemetics and then sent home. A small percentage of patients with vertigo will require admission to the hospital to treat dehydration with intravenous fluids or for further management, such as for a cerebellar infarction.(2) Rarely, as in this case, vertigo represents the presenting symptom of a life-threatening, treatable condition, such as a cerebellar hemorrhage.(3)

What features of this case suggest that the CNS, not the labyrinth, was the location of the abnormality? Unfortunately, definitively localizing vertigo to the central (brain) or peripheral (inner ear) portion of the vestibular system is difficult and sometimes impossible since vertigo itself and most associated signs and symptoms (eg, nausea, vomiting, and diaphoresis) can be the same in both central and peripheral vertigo. However, some signs and symptoms associated with vertigo can point more specifically to the inner ear or to the brain. For example, hearing loss and tinnitus suggest the inner ear, whereas cranial nerve or cerebellar findings suggest a brain abnormality. This patient did not present with any signs or symptoms that definitively implicated the brain as the location for her vertigo. However, several symptoms and signs, though not pathognomonic for a central cause, are so highly suggestive of a CNS abnormality that brain imaging should be obtained promptly when they accompany vertigo. One such symptom is headache, even though it is an inconsistent symptom of central vertigo and can sometimes be seen in peripheral syndromes as well. In a patient with vertigo, an inability to ambulate should also alert the treating physician that a central process is highly likely, thus prompting immediate brain imaging.(4) We have no indication from the case record that the patient's gait was evaluated (either initially or following treatment for vertigo and nausea). Vertical nystagmus and horizontal direction-changing nystagmus are definite central signs; unidirectional horizontal nystagmus can be a peripheral or a central sign.

Given the stakes involved in missing a central lesion, many evaluation centers now almost routinely perform brain imaging for patients presenting with new-onset acute vertigo, even in the absence of either definitive or suggestive CNS signs. Although such imaging of patients with vertigo will rarely uncover a structural abnormality, brain imaging is non-invasive and is the only means of ruling out a life-threatening disorder. Although a non-contrast computed tomography (CT) scan of the brain is adequate for ruling out a life-threatening hemorrhage, magnetic resonance imaging (MRI) is equally as reliable for this purpose.(5) Additionally, MRI can uncover recent ischemic infarctions and, with magnetic resonance angiography, arterial dissections.

What led to this patient's demise despite having an easily diagnosed, treatable condition? Primarily, three different physicians failed to seriously consider a central cause of vertigo, as evidenced by what appears to be (i) the omission of a complete neurologic examination by each examiner, and (ii) the failure to obtain brain imaging by the third attending, despite the presence of headache in a patient who had presented more than six hours earlier with vertigo and who had been treated with three different vestibular suppressants. Clinicians must never forget that a peripheral localization for vertigo cannot be confirmed with certainty. Rather, a peripheral localization is a diagnosis of exclusion when there are no symptoms, signs, laboratory findings, or evidence on imaging that point to a central process. Since the cerebellum has numerous connections with central vestibular structures, cerebellar lesions can masquerade as peripheral vestibular disorders. That appears to have been the case here: this patient's presenting symptom complex of vertigo, vomiting, and visual intolerance were, in retrospect, all produced by a cerebellar hemorrhage (the Figure, of a cerebellar infarction, demonstrates the vascular anatomy), probably because of involvement of the vestibulocerebellum—that is, the flocculonodular lobe, which lies at the caudal extent of the cerebellum.

Was a neurologic consultation necessary in this case? Probably not. Had any one of the three attendings considered obtaining a neurologic consultation, they probably would have done so because they were seriously entertaining a central disorder. That concern should have led them to assess the patient's gait and obtain diagnostic brain imaging, with or without a formal neurologic consultation.

An obvious "system failure" occurred in this case, wherein a patient had three different attendings in the span of six hours. In a busy emergency department, it is not reasonable to expect each new attending to repeat a full neurologic examination on each patient. However, when the presenting complaint is vertigo, a focused neurologic examination can successfully uncover a central abnormality, since the brain regions that can cause vertigo are limited to the brain stem and cerebellum. Thus, an abbreviated examination looking for an altered level of consciousness, gaze palsy (6), or a central (eg, vertical or horizontal direction-changing) type of nystagmus, facial numbness or weakness, upper extremity incoordination, or an ability to walk should suffice to rule out a central abnormality. This type of focused examination requires only a few minutes and should be routinely repeated when a patient with vertigo is "handed off" to another provider. Lowering the room lights to reduce vertigo can sometimes make it possible to examine a patient's eye movements. Since a patient's condition may change over the course of hours, this patient should have been examined repeatedly, at least at hourly intervals, to make a diagnosis in a timely fashion, even if multiple attendings had not been involved in the case.

Take-Home Points

• Most patients presenting to an ED with vertigo can be managed conservatively with vestibular suppressants and antiemetics and then sent home. Rarely, however, vertigo represents the presenting symptom of a cerebellar hemorrhage, which is a life-threatening, treatable condition.
• In a patient with vertigo, a head CT or MRI is required if:
  • There is an associated headache;
  • The patient is unable to ambulate; or
  • Physical examination findings suggest a central cause.
• A central process can be ruled out if all of the following are true:
  • There is no altered level of consciousness;
  • There is no gaze palsy;
  • There is no central (ie, vertical or direction-changing horizontal) nystagmus;
  • There is no facial numbness;
  • There is no upper extremity incoordination; and
  • The patient can walk.
• A peripheral (ie, inner ear) localization for vertigo cannot be confirmed with certainty. Rather, a peripheral localization is a diagnosis of exclusion when there are no symptoms, signs, laboratory findings, or evidence on imaging for a central process.

Joseph M. Furman, MD, PhD
Professor of Otolaryngology and Neurology
University of Pittsburgh School of Medicine

1. Furman JM, Cass SP. Vestibular disorders: a case-study approach. 2nd ed. New York, NY: Oxford University Press, Inc; 2003.

2. Huang CY, Yu YL. Small cerebellar strokes may mimic labyrinthine lesions. J Neurol Neurosurg Psychiatry. 1985;48:263-5.
[ go to pubmed ]

3. van der Hoop RG, Vermeulen M, van Gijn J. Cerebellar hemorrhage: diagnosis and treatment. Surg Neurol. 1988;29:6-10.
[ go to pubmed ]

4. Hotson JR, Baloh RW. Acute vestibular syndrome. N Engl J Med. 1998;339:680-5.
[ go to pubmed ]

5. Fiebach JB, Schellinger PD, Gass A, et al. Stroke magnetic resonance imaging is accurate in hyperacute intracerebral hemorrhage: a multicenter study on the validity of stroke imaging. Stroke. 2004;35:502-6.
[ go to pubmed ]

6. Kubo T, Sakata Y, Sakai S, Koizuka I, Matsunaga T, Nogawa T. Clinical observations in the acute phase of cerebellar hemorrhage and infarction. Acta Otolaryngol Suppl. 1988;447:81-7.
[ go to pubmed ]

7. Alvord LS, Herr RD. ENG in the emergency room: subtest results in acutely dizzy patients. J Am Acad Audiol. 1994;5:384-9.
[ go to pubmed ]

Table. Specific Diagnoses of 93 Patients Presenting with Dizziness in an Emergency Room

Reprinted with permission from The Journal of the American Academy of Audiology.(7)

Figure. MRI of a Right Inferior Cerebellar Infarction

The inset illustrates the vascular supply to the inferior cerebellum, which is perfused by the medial and lateral branches of the posterior inferior cerebellar artery (PICA) and the anterior inferior cerebellar artery (AICA).

Reprinted with permission from The New England Journal of Medicine.(4)