Acute dizziness: the HINTS exam at the bedside

A 62-year-old woman arrives at the emergency department with intense rotational dizziness since she woke up. She has vomited several times, has trouble standing and prefers to keep her eyes closed. Her blood pressure is a little high and she has diabetes. The head CT scan done on arrival shows no changes. The team thinks of labyrinthitis and considers discharge with symptomatic treatment as soon as she improves.

This is one of the most treacherous situations in emergency medicine. Most patients with continuous acute dizziness have a benign peripheral cause, such as vestibular neuritis. A minority have a stroke in the posterior circulation, in the cerebellum or brainstem, which can present with almost only dizziness, nausea and imbalance. A normal CT scan reassures, but it often misleads.

The HINTS exam was created for this crossroads. It brings together three oculomotor tests done at the bedside, without special equipment, and looks for signs that separate a peripheral lesion from a central one. This text shows what studies from several countries say about the accuracy of HINTS, about who can apply it well, about its limits and about how it fits into the reasoning in the face of acute dizziness.

Key points

• HINTS applies to acute vestibular syndrome: continuous dizziness of acute onset, with nystagmus, nausea and instability.
• In the hands of trained examiners, HINTS had higher sensitivity than early MRI for stroke.
• Head CT has low sensitivity for central causes of dizziness.
• The exam's performance depends heavily on the training of the person applying it.
• Assessing gait and trunk ataxia adds information to HINTS.

Acute vestibular syndrome: when HINTS applies

HINTS was designed for a specific group of patients. Acute vestibular syndrome is dizziness or vertigo of sudden onset, continuous for hours to days, accompanied by nausea or vomiting, intolerance to head movement, gait instability and nystagmus. The exam loses meaning when applied outside this picture.

This seems like a detail, but it is one of the most common sources of error. Anyone who applies HINTS to a patient without nystagmus, or to someone with positional vertigo lasting seconds, gets results that do not have the interpretation described in the studies. The head impulse test, for example, is meant to show a peripheral lesion in someone with continuous dizziness; in someone without acute vestibular syndrome, a normal impulse can be wrongly interpreted as a central sign.

The dizziness groups usually separated in the reasoning are:

• Acute vestibular syndrome: continuous dizziness, with spontaneous nystagmus, where HINTS applies.
• Triggered episodic vestibular syndrome: short attacks provoked by a change in position, as in benign paroxysmal positional vertigo.
• Spontaneous episodic vestibular syndrome: attacks that arise without a trigger, as in vestibular migraine and in some transient ischemic attacks.
• Chronic dizziness: symptoms persisting for months, with various causes.

This classification by timing and trigger, and not by the word the patient uses to describe the symptom, helps choose the right exam. A review of bedside algorithms shows that the approaches with the best accuracy combine a structured history and subtle oculomotor signs (Tarnutzer and Edlow, 2023).

Other causes of acute vestibular syndrome

Vestibular neuritis and posterior circulation stroke are the two poles of the discussion, but there are other possibilities. Vestibular migraine can produce long attacks, sometimes with nystagmus, and usually appears in people with a history of migraine. True labyrinthitis, with associated hearing loss, follows the same logic as neuritis, but adds the cochlear component. Drug and alcohol intoxication can cause nystagmus and ataxia, usually without the typical peripheral signs. Demyelinating diseases can mimic neuritis in younger people.

Remembering this list keeps the reasoning from becoming a binary choice. HINTS helps separate peripheral from central, but the next question, about which central or peripheral cause, depends on the history, the full examination and, often, on how things evolve over the following hours.

The three components of HINTS

HINTS stands for head impulse, nystagmus and test of skew. The logic is to look for the reassuring peripheral pattern; any sign outside that pattern points to a central cause.

The components can be described like this:

1. Head impulse: the examiner quickly turns the patient's head while the patient keeps looking at the examiner's nose. In vestibular neuritis, the eye loses fixation and makes a corrective saccade. A normal impulse in someone with acute vestibular syndrome is a warning sign for a central cause.
2. Nystagmus: in a peripheral lesion, the nystagmus is horizontal and beats to the same side in all directions of gaze. Nystagmus that changes direction with gaze, or pure vertical nystagmus, suggests a central cause.
3. Test of skew: with the alternate cover test, the examiner looks for vertical misalignment of the eyes. The presence of this deviation suggests a central lesion.

An expanded version, called HINTS plus, adds the assessment of hearing. New hearing loss can indicate ischemia of the anterior inferior cerebellar artery, which also supplies the inner ear and can mimic a peripheral lesion.

The forgotten sign: gait

HINTS assesses the eyes, but the ability to stand and walk also informs. Patients with vestibular neuritis can usually walk with help, while patients with cerebellar stroke may be unable to sit or stand without support. Many studies recommend grading truncal ataxia together with HINTS, because this simple sign adds information.

What studies show about accuracy

In an Argentine study with 114 patients with acute vestibular syndrome, 72 with vestibular neuritis, 32 with stroke in the posterior inferior cerebellar artery territory and 10 in the anterior inferior cerebellar artery territory, HINTS had sensitivity of 100 out of 100 and specificity of 94.4 out of 100 (Carmona et al., 2016). In the same study, grade 2 or 3 truncal ataxia had sensitivity of 92.9 out of 100 and specificity of 61.1 out of 100 (Carmona et al., 2016).

An overview published in 2018 gathered 25 articles on the topic, published from 2009 to 2017, and discussed the role of the oculomotor examination in separating central from peripheral causes (Kattah, 2018). The review of bedside algorithms notes that, in the hands of a trained otoneurologist, HINTS showed sensitivity of 100 out of 100 and specificity of 96 out of 100 for stroke (Tarnutzer and Edlow, 2023).

More recent studies tested combinations. In a Chinese study with 121 patients seen from December 2022 to June 2023, 69 of them with vestibular neuritis, HINTS combined with grade 2 or 3 truncal ataxia had sensitivity of 100 out of 100, against 88.5 out of 100 for HINTS alone (Liu et al., 2024). In another Chinese study with 239 patients, the sensitivity of the head impulse, gaze-evoked nystagmus and test of skew was 73.5, 61.2 and 26.5 out of 100, with specificity of 97.9, 92.6 and 93.2 out of 100, respectively (Qiu et al., 2022).

This last finding shows that no single component is enough. The strength of HINTS lies in the combination, and the test of skew, although specific, appears in few patients. So the correct reading is: any central component changes the hypothesis, but the absence of a single component has little power to rule out stroke.

Comparison with imaging

The reason for so much interest in HINTS is the weakness of early imaging. The algorithm review cites a study in which, of 185,338 emergency visits for dizziness, 46,852, or 25 in 100, had a CT scan, and only 6,469, or 4 in 100, had an MRI (Tarnutzer and Edlow, 2023). The same review cites a meta-analysis of 6 studies and 771 patients in which the sensitivity of CT for central causes of dizziness was only 28.5 in 100 (Tarnutzer and Edlow, 2023).

Even diffusion-weighted MRI has a limit. The authors note that it misses about one in five vertebrobasilar strokes presenting as acute vestibular syndrome when performed in the first 24 to 48 hours (Tarnutzer and Edlow, 2023). And obvious focal neurological signs have a sensitivity of only 44 in 100 for a central cause in this setting (Tarnutzer and Edlow, 2023).

Study	Patients	Main finding
Carmona and colleagues, 2016	114 with acute vestibular syndrome	HINTS with sensitivity of 100 and specificity of 94.4 in 100
Liu and colleagues, 2024	121, of whom 69 had neuritis	HINTS with truncal ataxia: sensitivity of 100 versus 88.5 in 100
Qiu and colleagues, 2022	239	Individual components with sensitivity between 26.5 and 73.5 in 100
Korda and colleagues, 2022	46, of whom 11 had stroke	vHINTS with accuracy of 94.2 in 100
Siepmann and colleagues, 2021	30, of whom 11 had stroke or TIA	Low specificity among less experienced examiners
Cortese and colleagues, 2025	67	Adapted algorithm with specificity of 57.9 in 100

Who performs the exam makes a difference

The good results of HINTS come largely from studies with neuro-ophthalmologists and neuro-otologists. When the exam passes into other hands, performance drops. In a German study of 30 patients, 11 of them, or 36.7 in 100, with stroke or transient ischemic attack, HINTS plus with video head impulse had sensitivity of 81.8 and specificity of 31.6 in 100, versus 72.7 and 36.8 in 100 on the conventional exam (Siepmann and colleagues, 2021).

The meta-analysis focused on emergency physicians found data for the STANDING algorithm in two studies with 450 patients examined by 11 physicians, with sensitivity of 0.96 and specificity of 0.88 after prior education and training (Nakatsuka and Molloy, 2022). The authors report that data were lacking for HINTS performed by emergency physicians and cite an earlier review with sensitivity of 83 and specificity of 44 in 100, based on a single study without specific training (Nakatsuka and Molloy, 2022).

Clinicians' own perception confirms the problem. In a survey at two hospitals in the United Kingdom, almost half of the 73 respondents reported limited experience, 33 of them, or no experience, 19 of them, with HINTS, and 98 in 100 identified a need for training (Warner and colleagues, 2022). At a Canadian hospital, among 380 patients with dizziness, 139, or 36.6 in 100, had neuroimaging, and only 17, or 7.1 in 100, of those who had no imaging had a record of HINTS (Quimby and colleagues, 2018).

The same Canadian study found another problem: 44 in 100 HINTS records used the expression "HINTS negative", which is ambiguous because it does not say whether the pattern was peripheral or central (Quimby and colleagues, 2018). Recording each component separately avoids this confusion and helps whoever takes over the shift. The text on organizing clinical reasoning in the medical record addresses this care.

Why training matters so much

Each component of HINTS requires motor skill and interpretation. The head impulse must be quick and of small amplitude, the corrective saccade may be subtle, and the nystagmus may be suppressed by visual fixation. When supervised practice is lacking, the examiner tends to classify as normal what is abnormal, and vice versa. The literature on teaching clinical reasoning with cases shows that skills like these improve with repetition and feedback.

When the exam becomes difficult

There are situations in which HINTS is hard to apply or interpret. Very nauseated patients may refuse head movement. People with significant cervical disease may have limited ability to perform the head impulse. Those who already had strabismus or vision loss in one eye may have a cover test that is hard to read. Sedative and antiemetic medications given on arrival can reduce nystagmus and mask the pattern.

In these scenarios, the best approach is to recognize the limitation. Recording that a component could not be assessed, explaining why, and considering the patient as being at indeterminate risk protects more than interpreting an incomplete examination as reassuring. Serial observation, reassessment of gait and discussion with neurology or neuro-otology complete the evaluation.

Another consideration is the timing of the examination. Nystagmus in neuritis tends to diminish over days, and a patient assessed late may have less clear findings. Knowing how long the symptoms have been present helps interpret what is seen.

Technology: video and telemedicine

To reduce dependence on individual skill, some services have started using video of the head impulse and oculography. In a Swiss study of 46 patients seen from July 2015 to April 2020, 11 of them with stroke, vHINTS had accuracy of 94.2, sensitivity of 100 and specificity of 88.9 per 100, while examination by a specialist had accuracy of 88.3, sensitivity of 90.9 and specificity of 85.7 per 100 (Korda and colleagues, 2022).

Telemedicine has also been tested. In a German study of 81 patients enrolled from June 1, 2019 to March 31, 2020, HINTS by telemedicine was performed in 72, or 88.9 per 100, and the complete examination was feasible in 46, or 63.9 per 100 (von Martial and colleagues, 2021). The results show that the idea is promising, but that technical feasibility still limits its use, as discussed in the text on clinical reasoning in telemedicine.

Broader algorithms attempt to help non-specialists. In a London study of 67 patients, an algorithm combining TiTrATE and STANDING had sensitivity of 90 per 100, specificity of 57.9 per 100 and overall accuracy of 62.7 per 100, with 24 false positives among the 67 patients (Cortese and colleagues, 2025). The authors report that vestibular migraine and chronic vascular problems were frequently misclassified.

What technology solves and what it leaves out

Video objectively records the head impulse, but choosing the right patient, reading the nystagmus and integrating it with the history still depend on the examiner. Technology helps those who already know how to apply the reasoning; it hardly replaces the reasoning.

How HINTS fits into diagnostic reasoning

Returning to the 62-year-old woman, she has continuous dizziness, nystagmus, vomiting and difficulty standing. The picture meets the criteria for acute vestibular syndrome, and a normal CT scan has little value for ruling out a central cause. She also has vascular risk factors, which increases the pretest probability of stroke.

Organized reasoning could follow this sequence:

1. Confirm that the picture is an acute vestibular syndrome, with spontaneous nystagmus.
2. Apply the three components of HINTS and record the result of each one.
3. Assess hearing and the ability to sit and walk without support.
4. Consider vascular risk factors and other neurological signs.
5. If any finding is central, or if the examination is doubtful, treat as possible stroke and discuss appropriate imaging.

The logic of pretest probability helps interpret the result. A typical peripheral pattern in someone young and without risk factors greatly reduces the chance of stroke. The same pattern in an older person with diabetes who is unable to walk deserves more caution, because the initial probability was higher and the examiner may have erred.

The most common error

The most frequent error is using a normal CT scan as proof of a peripheral cause. Another error is labeling any dizziness as "labyrinthitis" and ending the reasoning early. Both are examples of premature closure, a bias discussed in morbidity and mortality sessions. Keeping the question "what else could it be?" until the end of the evaluation is the best protection.

Uncertainty that must be stated

When the examination is doubtful, or when the examiner feels they did not perform some component well, that uncertainty should appear in the record and in the handover. Saying the HINTS was inconclusive is safer than forcing a result. The strategies for deciding under diagnostic uncertainty apply well to this scenario.

How to explain the reasoning to the patient

Patients with intense dizziness usually want a quick answer, and the word "labyrinthitis" often comes ready-made from home. Explaining that the team performed an eye examination to tell an inner ear problem apart from a circulatory problem in the brain helps the patient understand why it may be necessary to observe for longer or perform another test. This conversation also prepares the ground for giving warning signs at discharge, such as new difficulty speaking, swallowing or walking.

How to train the examination

HINTS is learned through deliberate practice. Watching videos of normal and abnormal examinations, practicing the head impulse on colleagues under supervision and comparing your own conclusions with those of an experienced examiner are steps that speed up the learning curve. Reviewing the final diagnosis of the patients examined afterwards closes the feedback loop and helps calibrate confidence. In services without an otoneurologist available, local protocols with clear criteria for requesting help or imaging reduce dependence on each on-call physician's memory.

It is also worth training the decision about when not to apply the examination. Knowing how to recognize that the patient has positional vertigo, and not acute vestibular syndrome, avoids wrong conclusions and directs you to the maneuvers and tests suited to that condition. This recognition depends on a good history, with questions about onset, duration, triggers and associated symptoms.

How NEXO helps train this reasoning

In NEXO, the case of the day is fictional, with a real clinical presentation and up to six progressive clues, and you propose the diagnosis by searching for the ICD-10 code.

The limits of use of the game are described on the NEXO economy page. If you research the teaching of clinical reasoning, the page for researchers explains how to get in touch. Professionals who write or review cases find information under curation and under creators, and institutions can learn about the proposal on the partnerships page. Training with differential diagnosis and clues and searching for the right ICD-10 code complete this practice.

Frequently asked questions

Who is HINTS indicated for?

For patients with acute vestibular syndrome, that is, continuous dizziness of acute onset with nystagmus, nausea and instability; outside this condition, the results lose the interpretation described in the studies (Tarnutzer and Edlow, 2023).

What is the accuracy of HINTS?

In 114 patients with acute vestibular syndrome, HINTS had a sensitivity of 100 out of 100 and a specificity of 94.4 out of 100 in the hands of experienced examiners (Carmona et al., 2016).

Does a normal CT rule out stroke in acute dizziness?

CT has low sensitivity in this scenario, 28.5 out of 100 for central causes according to a meta-analysis cited in a recent review (Tarnutzer and Edlow, 2023).

Can emergency physicians apply HINTS?

With training, performance improves, but data are lacking: a meta-analysis found data only with emergency physicians trained in the STANDING algorithm, with a sensitivity of 0.96 (Nakatsuka and Molloy, 2022).

Is it worth assessing gait together with HINTS?

It helps a great deal, and a study with 121 patients showed a sensitivity of 100 out of 100 with HINTS and truncal ataxia, versus 88.5 out of 100 with HINTS alone (Liu et al., 2024).

Take-home messages

• HINTS applies to acute vestibular syndrome, not to any dizziness.
• Any central component shifts the hypothesis to possible stroke.
• A normal CT has little value for ruling out a central cause of dizziness.
• The performance of the examination depends on the training of the examiner.
• Assessing gait, truncal ataxia and hearing adds information to HINTS.
• Recording each component separately avoids the ambiguity of "negative HINTS".

References

1. Carmona S, Martínez C, Zalazar G, et al. The Diagnostic Accuracy of Truncal Ataxia and HINTS as Cardinal Signs for Acute Vestibular Syndrome. Front Neurol. 2016. PMID 27551274.
2. Cortese E, Rochelle P, Patel F, et al. Integrated diagnostic algorithm for acute vertigo combining TiTrATE, STANDING, and HINTS: a validation study in the emergency department. Sci Rep. 2025. PMID 40659780.
3. Kattah JC. Use of HINTS in the acute vestibular syndrome. An Overview. Stroke Vasc Neurol. 2018. PMID 30637123.
4. Korda A, Wimmer W, Zamaro E, et al. Videooculography "HINTS" in Acute Vestibular Syndrome: A Prospective Study. Front Neurol. 2022. PMID 35903121.
5. Liu X, Li Z, Ju Y, et al. Application of bedside HINTS, ABCD2 score and truncal ataxia to differentiate cerebellar-brainstem stroke from vestibular neuritis in the emergency room. Stroke Vasc Neurol. 2024. PMID 38531544.
6. Nakatsuka M, Molloy EE. The HINTS examination and STANDING algorithm in acute vestibular syndrome: A systematic review and meta-analysis involving frontline point-of-care emergency physicians. PLoS One. 2022. PMID 35511910.
7. Qiu T, Dai X, Xu X, et al. A prospective study on the application of HINTS in distinguishing the localization of acute vestibular syndrome. BMC Neurol. 2022. PMID 36199036.
8. Quimby AE, Kwok ESH, Lelli D, et al. Usage of the HINTS exam and neuroimaging in the assessment of peripheral vertigo in the emergency department. J Otolaryngol Head Neck Surg. 2018. PMID 30201056.
9. Siepmann T, Gruener C, Simon E, et al. Video-Oculography-Assisted Head Impulse Test and Caloric Testing for Detecting Stroke in Acute Vertigo Patients via Modified HINTS Plus. J Clin Med. 2021. PMID 34640489.
10. Tarnutzer AA, Edlow JA. Bedside Testing in Acute Vestibular Syndrome-Evaluating HINTS Plus and Beyond-A Critical Review. Audiol Res. 2023. PMID 37736940.
11. von Martial R, Leinweber C, Hubert N, et al. Feasibility of Telemedical HINTS (Head Impulse-Nystagmus-Test of Skew) Evaluation in Patients With Acute Dizziness or Vertigo in the Emergency Department of Primary Care Hospitals. Front Neurol. 2021. PMID 35222226.
12. Warner CL, Bunn L, Koohi N, et al. Clinician's perspectives in using head impulse-nystagmus-test of skew (HINTS) for acute vestibular syndrome: UK experience. Stroke Vasc Neurol. 2022. PMID 34702750.