Utility of head CT in the evaluation of vertigo/dizziness ...

Emerg Radiol (2013) 20:45?49 DOI 10.1007/s10140-012-1071-y

ORIGINAL ARTICLE

Utility of head CT in the evaluation of vertigo/dizziness in the emergency department

Courtney Lawhn-Heath & Christopher Buckle & Gregory Christoforidis & Christopher Straus

Received: 30 May 2012 / Accepted: 10 August 2012 / Published online: 2 September 2012 # Am Soc Emergency Radiol 2012

Abstract Acute dizziness (including vertigo) is a common reason to visit the emergency room, and imaging with head CT is often performed initially to exclude a central cause. In this study, consecutive patients presenting with dizziness and undergoing head CT were retrospectively reviewed to determine diagnostic yield. Four hundred forty-eight consecutive head CTs in a representative sample of dizzy emergency room (ER) patients, including patients with other neurological symptoms, were reviewed to identify an acute or subacute cause for acute dizziness along with the frequency and modalities used in follow-up imaging. The diagnostic yield for head CT ordered in the ER for acute dizziness is low (2.2 %; 1.6 % for emergent findings), but MRI changes the diagnosis up to 16 % of the time, acutely in 8 % of cases. Consistent with the American College of Radiology appropriateness criteria and the literature, this study suggests a low diagnostic yield for CT in the evaluation of acute dizziness but an important role for MRI in appropriately selected cases.

Keywords Dizziness . Vertigo . CT . Utilization . Emergency department

Summary statement Head CT was found to have a low diagnostic yield for patients presenting to the emergency room with dizziness.

C. Lawhn-Heath (*) University of Chicago Pritzker School of Medicine, 924 East 57th Street #104, Chicago, IL 60637, USA e-mail: courtneylh@uchicago.edu

C. Buckle : G. Christoforidis : C. Straus

University of Chicago Medical Center, Chicago IL, USA

Introduction

Dizziness accounts for around 7.5 million visits in ambulatory care settings annually and is one of the most common presenting complaints in the emergency department (ED) [1]. Benign dizziness is sometimes life-threatening and can be difficult to evaluate; as a result, dizzy patients in the ED consume more healthcare resources, including imaging, than non-dizzy patients do [2].

The most common type of dizziness is vertigo, which represents around 54 % of dizziness complaints [3]. Unlike other forms of dizziness, vertigo is characterized by the illusory sensation of spinning, tilting, or translational motion. Etiologies can be divided into peripheral (end organ vestibular) and central (CNS) causes, with peripheral vertigo being the most common and generally benign [4]. Central vertigo is less common but can be life-threatening when associated with a posterior fossa bleed or stroke. Missed cerebellar infarctions have an estimated 40 % mortality rate, while missed diagnoses of peripheral etiologies can lead to unresolved symptoms and unnecessary tests [2]. Therefore, distinguishing between central and peripheral etiologies of acute vertigo in the ED is crucial.

While there is evidence that clinical information such as the history and physical examination can help differentiate central from peripheral vertigo, many clinicians use neuroimaging in cases where the clinical findings are equivocal or uncertain. This is particularly the case in a busy emergency room where histories/physicals can be difficult to obtain/ perform and time is limited. Head CT is frequently used as a first-line imaging modality to exclude posterior fossa hemorrhage or a large mass as a cause for vertigo, although there is a substantial amount of variation between EDs in their imaging practices [5]. However, CT has a low diagnostic yield for isolated vertigo. In one study, 100 % of head CT

46

Emerg Radiol (2013) 20:45?49

scans performed for isolated vertigo were negative [6], although vertigo often presents with other nonspecific neurologic symptoms (e.g., weakness, headache) and can be difficult to distinguish from other forms of dizziness (dysequilibrium, presyncope), especially in noncooperative patients, which are common in the ER. CT is also known to have poor sensitivity for both posterior fossa pathology and for ischemic strokes (26 % compared with 83 % with MRI), a common cause of central vertigo [4, 7, 8]. The superior sensitivity of MRI and its use of non-ionizing radiation suggest that it may be a more appropriate imaging modality for vertigo [1, 4, 9, 10, 11]. For example, appropriateness criteria from the American College of Radiology rate head CT for vertigo at a 5 on a scale of 1?9, indicating "may be appropriate," while MRI is ranked at a 7, or "usually appropriate" [12]. However, MRI is expensive and not readily available in many EDs. Understanding how CT compares to MRI in the detection of pathology associated with dizziness would be helpful in determining the optimal imaging algorithm for the dizzy ER patient.

The purpose of this study was to determine the diagnostic yield of head CT and the value of follow-up imaging, especially MRI, in the evaluation of a realistic population of dizzy ER patients. Specifically, we included patients with either dizziness or vertigo and included patients with secondary neurological signs and with minor trauma. It was hypothesized that the diagnostic yield of CT would remain low but non-zero and that MRI would be more accurate than CT but only change the diagnosis a small percent of the time.

Methods

Setting and subject selection

This institutional review board-approved study was performed at the University of Chicago Medical Center (UCMC). Patients from the UCMC emergency department, a level 2 trauma center, who received a head CT for the complaint of vertigo or dizziness between January 2003 and December 2010 were included. They were identified retrospectively from ICD 9 codes using Illuminate (Softek Inc., Kansas City, KS, USA), a medical imaging search engine. In the case of a subject having multiple visits to the ED for dizziness during the study period, only the single most severe case (as determined by imaging findings, disposition, and symptoms) was included in the study. Patients identified this way were excluded from the study if the provider notes for the accompanying encounter made no mention of the sensation of spinning or translational motion (in the case of vertigo)

or descriptors equivalent to lightheadedness, disequilibrium, or presyncope (in the case of dizziness).

Imaging data collection and analysis

Results of the initial head CT scan and any follow-up imaging (MRI, MRA, etc.) were collected from the clinical reports of the academic neuroradiologists who read or supervised the reading of the studies. Each finding was coded into one of 11 categories, including acute (acute posterior fossa hemorrhage, acute/subacute posterior fossa ischemia, acute hemorrhage/ischemia not in the posterior fossa, neoplasm with mass effect/midline shift, new or worsening hydrocephalus), sub-acute (posterior fossa neoplasm, supratentorial neoplam), non-acute/unrelated (known findings unchanged, sinus disease, subdural effusions, etc.), and no significant finding. In patients with trauma, only intracranial abnormalities were coded. Follow-up imaging was coded by modality and findings according to the acuity (acute, subacute) and the effect on diagnosis relative to the initial CT (CT false negative, CT false positive, confirmed diagnosis, no change in diagnosis).

Clinical data collection

Limited clinical data was collected using the Illuminate search engine including age, gender, time of exam, and ICD-9 code (dizziness and giddiness or vertigo).

Data analysis

The diagnostic yield of head CT for vertigo was defined as the percentage of acute and/or sub-acute findings excluding false positives identified by subsequent imaging. The frequency of follow-up imaging by modality (e.g., MRI, CT, etc.) and the frequency with which it changed the diagnosis were determined. Follow-up imaging was considered to have changed the diagnosis if it (1) revealed a new acute or subacute finding (e.g., a new acute infarct) not identified prospectively on CT (false negative) or (2) demonstrated that an acute/subacute finding identified prospectively by CT was not real (e.g., no thrombosis seen when CT identifies a hyperdense transverse sinus) (false positive) or (3) confirmed a diagnosis when there was uncertainty on the CT.1 Test characteristics (sensitivity, specificity, positive, and negative predictive value) for CT were determined using follow-up imaging as the gold standard.

1 Intracranial stenosis revealed on follow-up imaging was classified as a new acute finding if it was associated with TIA-like symptoms; otherwise, it was considered sub-acute.

Emerg Radiol (2013) 20:45?49

47

Table 1 Results of positive initial head CTs by acuity of finding

Finding

Number Frequency (%)

Acute (total)

7

1.6

Bleed

2

0.5

Neoplasm with midline shift/herniation

3

0.6

New or worsening hydrocephalus

2

0.5

Subacute (total)

3

0.6

Neoplasm

3

0.6

Total

10

2.2

of the time, acutely 8 % of the time. The majority of the acute changes in diagnosis were false negatives, typically ischemic strokes not identified on initial head CT.

The test characteristics of head CT in the evaluation of acute vertigo were obtained using follow-up imaging where available as the gold standard. Head CT had high negative predictive value (88 %) and low sensitivity (40 %) for detecting etiologies of acute dizziness.

Discussion

Results

Clinical/demographic data Average age was 59?21. 54 % of patients were male. Of the patients, 31 % had vertigo by ICD-9 code, and 69 % had dizziness.

Imaging During the study period of 3/1/2005 to 6/17/2011, 448 initial head CTs were performed for all forms of dizziness in the ED. The diagnostic yield of initial head CT was 2.2 % for acute and subacute findings (Table 1). Neoplasm with midline shift or herniation was the most common acute finding. Of the two acute bleeds identified, one was a hemorrhagic stroke and one was a supratentorial subdural hematoma--none were in the posterior fossa. There were no findings on CT that were identified as suspicious for ischemic strokes of the posterior fossa. Masses were the most common subacute finding (see Table 1).

Of the 448 patients who received an initial head CT, 104 received follow-up imaging. Of the initial head CTs, five had positive findings, and the remainder had negative, incidental or unchanged findings. Some patients received multiple follow-up imaging modalities, so a total of 136 followup studies were performed on 104 patients, with MRI being the most common (64 %) (Table 2). Thirteen percent (n017) of all follow-up studies returned a finding that changed (n016) or confirmed (n01) the diagnosis from the initial CT study. MRI accounted for 82 % (14/17) of the follow-up exams that changed the diagnosis, and MRI changed the diagnosis 16 %

Consistent with our hypothesis, head CT was found to have a very low diagnostic yield for acute dizziness in the ED. This is consistent with literature. Wasay and Bakshi found a very low diagnostic yield of 0 % when head CT was used to evaluate 344 ER patients with vertigo. This larger study, with similar patient demographics (slight male predominance and older age) expands the scope of the literature including patients with both vertigo and dizziness and including patients with secondary symptoms such as focal neurological signs, trauma, and headache. In this heterogenous but more representative patient population, the diagnostic yield for vertigo and dizziness remains extremely low. In particular, there were no posterior fossa bleeds in our series. The cases of hemorrhage--punctate in the left thalamus and a small supratentorial subdural--are questionably related to vertigo in the first place. The vast majority of other acute findings in this series, including hydrocephalus/ progression in ventricular size and brain neoplasms with new or progressing mass effect/herniation, were identifiable by history (known tumor, shunt). In all cases of acute findings on initial CT, patients had symptoms other than vertigo/ dizziness, although the symptoms ranged in specificity from generalized headache to focal neurological signs. The extent to which these "acute" findings represent a central cause for the patients' dizziness remains an open question, especially given that dizziness/vertigo is multifactorial, common, and usually peripheral. Hence, this study confirms the low

Table 2 Follow-up imaging by modality

Modality

Total number performed (n/% of total follow-ups)

Studies that changed initial diagnosis (n/%)

MRI

87/64.0 %

14/16 %

MRA CT CTA Other Total

20/14.7 % 18/13.2 % 10/7.4 % 1/0.7 % 136/100 %

1/5 % 0/0 % 2/10 % 0/0 % 17/13 %

False negative

Acute: ischemic stroke(6), aneurysm (1); sub-acute: characterize/confirm mass (3), identify new mass (3)

Vertebrobasilar stenosis (1)

Vertebrobasilar stenosis (1)

False positive Exclude hemorrhage (1)

Exclude sinus thrombosis (1)

48

Emerg Radiol (2013) 20:45?49

diagnostic yield for vertigo/dizziness in a broader but more representative ER patient population while preliminary analysis of the clinical data suggests clinical features may be helpful in identifying patients that would benefit from further imaging.

This study also highlights the value of follow-up imaging, especially MRI. With its greater sensitivity for ischemic stroke, MRI identified several false negatives and false positives on the initial CT, changing the diagnosis 16 % of the time, acutely in 8 % of cases. While the overall negative predictive value of CT was high, this is due to the low prevalence of disease in our population as the sensitivity of CT for stroke is poor.

Vertigo/dizziness is a common and challenging problem faced by the ER physician. CT imaging is costly, and associated with radiation, so should be used only when necessary. This study, supporting the literature suggests an extremely low yield for CT on dizziness patients in the ER. As CT costs about $300 per scan, in our population approximately $134,000 was spent imaging patients when only $2,100 was immediately helpful in patient management. While it could be argued that a negative CT saves money by avoiding hospital admission because of its excellent negative predictive value, a negative CT does not exclude a central cause for vertigo and may provide a false sense of security given its low sensitivity for the identification of ischemic strokes. Consistent with the American College of Radiology appropriateness criteria, this study finds that MRI is more helpful in the evaluation of a central cause for dizziness, and suggests that the less frequent use of a more accurate test may better optimize imaging of the acutely dizzy patient. The challenge is then to use clinical factors to appropriately select which patients to image, and preliminary results from this study suggest that certain clinical factors (history, secondary symptoms) may be helpful.

Future studies (1) will compare the clinical characteristics of the dizziness patient population with acute findings on head CT to those without to determine if clinical rules can be useful in identifying patients who need imaging from those that do not and (2) will work with our ER colleagues to develop a dizziness treatment algorithm that balances accurate identification of central vertigo with the practical constraints of a busy ED in a cost effective manner and (3) will obtain prospective data including imaging frequency before and after the implementation of a imaging algorithm for acute dizziness.

The study has limitations. First, it is retrospective, which limits standardization of data and limits data collection to that obtained during the course of the patients' care. While patients who receive head CT for acute dizziness are a selected and thus biased subset of dizzy patients (even more so for patients that receive follow-up imaging), they are presumably more likely to have imaging findings than the

larger population of acutely dizzy patients, so the findings here are generalizable, although the frequency of findings on MRI may be overestimated. Second, this is a single-center study. Future studies will attempt to extend these findings to other institutions, especially community hospitals.

Third, data collection focused on imaging without detailed clinical data collection (history, physical, past medical history etc). In particular, the ICD-9 code diagnosis of vertigo or dizziness is only a rough proxy for the patients' true symptoms. Also, there is some inherent variability in physicians' individual criteria that cause them to label a patient as having vertigo or dizziness. Clinical factors predictive of positive imaging will be examined in a subsequent study. Lastly, we did not determine the total number of patients seen for dizziness or vertigo overall during the study period and so the frequency of imaging is unknown. However, this does not change the finding that head CT is a low yield exam for the acutely dizzy patient.

Conclusion

We conclude that CT has a very low diagnostic yield for the acutely dizzy patient in the ER. In particular, 0 posterior fossa hemorrhages were identified in our sample of 424 patients. Furthermore, MRI is a more accurate imaging modality than CT for the evaluation of acute vertigo/dizziness. Our findings suggest that clinical factors may help to select patients who are at risk for central vertigo and that imaging directly with MRI may be more efficient in the atrisk group.

Acknowledgments The authors would like to thank Jack Collins, Sarah Orrin, Jay Patel, Danny Cheng, and Will Whetsell for their assistance with data collection. Our work was supported by the Arthur Quern fellowship and the University of Chicago Pritzker School of Medicine Summer Research Program.

References

1. Kerber KA, Brown DL, Lisabeth LD et al (2006) Stroke among patients with dizziness, vertigo, and imbalance in the emergency department: a population-based study. Stroke 37:2484

2. Chase M, Joyce NR, Carney E et al (2011) ED patients with vertigo: can we identify clinical factors associated with acute stroke? Am J Emerg Med 30(4):587?591

3. Kroenke K, Hoffman RM, Einstadter D (2000) How common are various causes of dizziness? A critical review. South Med J 93:160

4. Seemungal BM, Bronstein AM (2008) A practical approach to acute vertigo. Pract Neurol 8:211

5. Kim AS, Sidney S, Klingman JG et al (2011) Practice variation in neuroimaging to evaluate dizziness in the ED. Am J Emerg Med 30(5):665?672

6. Wasay M, Dubey N, Bakshi R (2005) Dizziness and yield of emergency head CT scan: is it cost effective? Emerg Med J 22:312

Emerg Radiol (2013) 20:45?49

49

7. Chalela JA, Kidwell CS, Nentwich LM et al (2007) Magnetic resonance imaging and computed tomography in emergency assessment of patients with suspected acute stroke: a prospective comparison. Lancet 369:293

8. Kerber KA (2007) Vertigo and dizziness in the emergency department. Emerg Med Clin N Am 27(1):39?50

9. Labuguen RH (2006) Initial evaluation of vertigo. Am Fam Physician 73:244

10. Edlow JA, Newman-Toker DE, Savitz SI (2008) Diagnosis and initial management of cerebellar infarction. Lancet Neurol 7:951

11. Fiebach JB, Schellinger PD, Gass A et al (2004) Stroke magnetic resonance imaging is accurate in hyperacute intracerebral hemorrhage: a multicenter study on the validity of stroke imaging. Stroke 35:502

12. American College of Radiology (2008) American College of Radiology (ACR) appropriateness criteria: vertigo and hearing loss.

 ................
................

In order to avoid copyright disputes, this page is only a partial summary.

Google Online Preview   Download