HINTS to Identify Stroke in ED Patients with Dizziness
Version of Record online: 13 FEB 2014
© 2014 by the Society for Academic Emergency Medicine
Academic Emergency Medicine
Volume 21, Issue 3, page 347, March 2014
How to Cite
Kohn, M. A. (2014), HINTS to Identify Stroke in ED Patients with Dizziness. Academic Emergency Medicine, 21: 347. doi: 10.1111/acem.12339
- Issue online: 13 MAR 2014
- Version of Record online: 13 FEB 2014
To the Editor:
In a recent article, Newman-Toker et al. compared the discrimination of the HINTS (head impulse, nystagmus type, test of skew) examination to the ABCD2 score in diagnosing stroke in emergency department (ED) patients with acute continuous vertigo and dizziness. The ABCD2 score is a tool for estimating the risk of stroke in patients after transient ischemic attack (TIA). It combines readily available clinical factors to inform early clinical decisions such as whether to obtain a computed tomography (CT) angiogram and whether to hospitalize the patient for observation. Predictably, this general risk stratification tool for TIA patients performed poorly in identifying posterior circulation stroke when compared with a specialized diagnostic test for patients with continuous vertigo.
Leaving aside the comparison to the ABCD2 score, the sensitivity and specificity of the HINTS exam for posterior fossa stroke were very high, 96.5 and 84.4%. Given the frequency of dizziness as a presenting complaint, emergency physicians certainly need a good tool for distinguishing central from peripheral causes. However, the 190 patients included in this study were not typical of ED patients with dizziness. They were identified over a period of 13 years from 86,000 visits per year; all had at least one stroke risk factor (none had an ABCD2 score < 2); all had acute, persistent vertigo with nystagmus plus nausea/vomiting, head motion intolerance, and new gait unsteadiness—findings that are generally sufficient to justify hospitalization from the ED, and all were, in fact, hospitalized. The proportion with posterior fossa stroke (59.5%) or other central causes (5.8%) was almost two-thirds, whereas the typical proportion of unselected dizziness patients with stroke or TIA is 3.2%. The authors acknowledge the high prevalence of stroke in their sample. Contrary to their expectations, the sensitivity of HINTS for stroke would probably be lower and the specificity higher when applied to a lower-prevalence sample.
Most ED patients with dizziness do not have a posterior circulation stroke. The few that do usually have an easily identified neurologic deficit or gait ataxia. Nevertheless, pending validation for use by emergency physicians in a more typical population of ED patients with dizziness, HINTS may be a powerful tool that helps identify the most difficult to diagnose stroke patients.