The HINTS exam is a 3-part test designed by Dr. Newman-Toker to clinically exclude a central cause of acute vertigo in patients who present with vertiginous symptoms. HINTS involves most famously, the head impulse test(see diagram below), a maneuver which involves torquing your patients head from midline to 20 degrees of rotation(or Vice Versa) and evaluating for tracking deficits. Not only is this exam rather difficult to assess, it seems likely to produce a vertebral artery dissection in the patient (assuming they do not already have one). The other two components of the exam include testing for ocular skew and multi-directional nystagmus, both far less fear inducing maneuvers for patients and physicians alike. Since Dr. Newman-Toker’s article published in STROKE in 2009 claimed 100% sensitivity and 96% specificity, those of us secretly obsessed with vertigo in all its manifestations were waiting for some confirmation. Confirmation arrived this month in a new paper published by Newman-Toker himself examining HINTS vs the ABCD2 score for ruling out posterior stroke (2).
This validation set of sorts was published in Academic Emergency Medicine in October 2013 and was entitled “HINTS Outperforms ABCD2 to Screen for Stroke in Acute Continuous Vertigo and Dizziness”. The fact that the HINTS exam outperforms the ABCD2 score to identify strokes of posterior origin should not instill confidence that the HINTS exam can effectively identify posterior stroke. The ABCD2 score performs very poorly when used to distinguish central vs peripheral origin of vertigo (3), but more importantly the rule was never derived to differentiate ischemic from non-ischemic causes of a stroke-like presentation. The ABCD2 score was built to risk stratify TIA patients into high-risk and low-risk categories for secondary stroke, and even in this role it has questionable utility (4). There is no way the ABCD2 rule would perform with any accuracy in identifying vertigo of central origin. Physicians inherently know that it is the quality of the vertigo and physical exam findings that help us to differentiate between peripheral and central origin. It is obvious that a rule built specifically to differentiate these two pathologies will perform far better than a rule built for an entirely different purpose. We may as well compare the performance of the PERC rule to the HINTS exam at identifying posterior fossa ischemia. This is the equivalent of a magician’s flair, meant to distract us from the truly important findings of this trial, how the HINTS exam performed independently to its straw man comparator. PRESTO!!!!
In Dr. Newman-Toker’s original article in which he presented the HINTS exam (2), he tested each of the 3-parts on a cohort of 101 “high-risk” patients with acute vestibular syndrome (AVS). High risk was defined as rapid onset of vertigo, nausea, vomiting, and unsteady gait with or without nystagmus with ?1 stroke risk factor (smoking, hypertension, diabetes, hyperlipidemia, atrial fibrillation, eclampsia, hypercoagulable state, recent cervical trauma, or prior stroke or myocardial infarction). Patients were excluded if they had a history of recurrent vertigo, an obvious peripheral source, or those with obvious central or oculomotor signs. The authors described the cohort as a prospective cross-sectional study, which is somewhat misleading as the study population was gathered in a hybrid fashion. The population was drawn from two sources, the first being the traditional and externally valid cohort of those presenting with symptoms of AVS, who once identified were shuttled off to the authors who then performed the HINTS exam. The second far more methodologically questionable group were patients admitted to the hospital after radiographic findings of posterior fossa ischemia. This group was retrospectively identified and subsequently prospectively evaluated with the HINTS exam. Of note is the fact that all the HINTS exams were conducted by two highly experienced operators who performed full histories and physical exams before completing the HINTS exam. Essentially these trials are not an evaluation of the HINTS exam’s ability to identify central vertigo, but rather an evaluation of two experienced clinicians’ proficiency in differentiating central from peripheral lesions revealed by a thorough history, physical exam, as well as HINTS. Of the 101 patients included in this cohort 76 (76%) had a central cause for their AVS. 67 were identified prospectively and 9 in this questionable pseudo-retrospective fashion. Overall in this population the HINTS exam performed extremely well with a sensitivity and specificity of 100% and 96% respectively. In fact it outperformed the initial diffusion weighted MRI performed within the first 48 hours which surprising had a sensitivity of 88% when compared to the “gold standard” of MRI performed after the first 48 hours.
A number of physicians in the Emergency Medicine community have voiced their unease with these findings. 76% of the patients in this study had a central lesion as the underlying cause of their vertigo. The inclusion criteria states, all of these patients are required to have vertigo with gait instability. This is obviously a highly selected population. Secondly, though the authors claim 100% sensitivity, the sample size is far too small to use the rule with any clinical certainty. Third, this study was done by clinicians who were very experienced and comfortable using these special tests, with no formal evaluation of the inter-rater reliability. Finally this is essentially a derivation set of a diagnostic aide performed on a very small sample size. Further studies are required to examine the HINTS exam’s stability for us to even begin to feel comfortable with its diagnostic capabilities. The recent article published in Academic Emergency Medicine is Dr. Newman-Toker et al’s attempt to address these concerns.
The Newman-Toker et al article used similar methods to their previous trial with an enrollment of 192 patients. If you look at the time period in which these patients were enrolled (1999-2012) it actually overlaps the time period of enrollment of their initial study (1999-2005)(1,2,4). In reality this is not a completely new cohort but rather the same 101 patients from their initial trial with 91 “new” patients. This is obviously the result of just how infrequently centrally caused AVS occurs and the highly selective nature of the population. Overall 65.3% of the cohort was diagnosed with AVS central in origin. The HINTS criteria performed admirably well with a sensitivity of 96.8% and a specificity of 98.5%. In an attempt to achieve perfection the authors added the criteria of new hearing loss to the HINTS exam. This helped boost its sensitivity to 99.2%, and like any good decision instrument rebranded itself as the HINTS “plus” exam.
The question should not be how accurate is the HINTS exam, but rather is it accurate enough to change decision-making on the undifferentiated AVS patient? The HINTS exam has the potential to supplement us in two ways. First in the high risk patient who you intend to admit for further evaluation, the HINTS exam would be used to rule out a central cause to their symptoms and the patient could be safely discharged. The second, that HINTS could aid our decision making process is in the low risk patient to help identify a subtle presentation of central vertigo which would otherwise go unnoticed. In the Newman-Toker trials, the HINTS exam was performed on a cohort that exhibited either trunk ataxia or gait instability, both of which are ominous findings. These are the high risk patients which we hope the HINTS exam will help us to further risk stratify into those who require admission and those who can be safely discharged. In a group as high risk as these, a sensitivity of 96% is not acceptable to safely rule out a central cause of the patient’s vertigo. Not to mention we have no idea about its inter-rater reliability and how well these tests function in the novice’s hands. Conversely how the HINTS exam performs in a group of AVS patients who are far more benign in their presentation is unknown. When used in a lower risk cohort (similar to the typical population we see in the Emergency Department with AVS), it may in fact lead to more diagnostic testing without a significant increase to our diagnostic yield.
Clearly the HINTS exam is an interesting and useful tool to evaluate patients with AVS. Whether it will ever be refined enough to aid us in our critical reasoning for the undifferentiated ED patient is still unknown. Presently it has been tested on a highly selected population, by a single tester, without any examination of inter-rater reliability. Until we can examine the HINTS performance in the ED, as performed by ED physicians, on our patients, we do not know what, if any benefit it contributes to our clinical decision-making.
1. Newman-Toker et al. HINTS Outperforms ABCD2 to Screen for Stroke in Acute Continuous Vertigo and Dizziness. Academic Emergency Medicine. Volume 20, Issue 10, pages 986–996, October 2013
2. Kattah et al. HINTS to Diagnose Stroke in the Acute Vestibular Syndrome Three-Step Bedside Oculomotor Examination More Sensitive Than Early MRI Diffusion-Weighted Imaging. Stroke. 2009; 40: 3504-3510
3. Babak et al. Application of the ABCD2 Score to Identify Cerebrovascular Causes of Dizziness in the Emergency Department. Stroke. 2012; 43: 1484-1489
4. Perry et al. Prospective validation of the ABCD2 score for patients in the emergency department with transient ischemic attack. CMAJ July 12, 2011 vol. 183 no. 10.