Choose the examination from the dizziness pattern, not from the word vertigo. Continuous symptoms, spontaneous attacks and brief position-triggered episodes create different diagnostic questions. HINTS is not a universal stroke screen, and a negative head CT does not make every dizzy patient safe for discharge.
The clinical framework here is SAEM's GRACE-3 guideline, checked on 27 September 2026. It concerns adults with acute dizziness or vertigo in the emergency department. The fictional case pairs below are original teaching exercises for ABEM-style reasoning, not reproduced examination items or evidence that a reader can perform specialist bedside tests without training.
Classify timing and triggers before selecting a manoeuvre
Ask when symptoms began, whether they are continuously present, how long individual episodes last and what actually initiates an episode. A movement that worsens ongoing dizziness is not necessarily the trigger of a new episodic syndrome. That distinction prevents a common error: calling any movement-sensitive dizziness positional vertigo.
The GRACE-3 framework separates acute vestibular syndrome, spontaneous episodic vestibular syndrome and triggered episodic vestibular syndrome. Its published definitions and recommendations are the source for the following original decision map. SAEM GRACE-3.
| Pattern in the history | First examination question | Interpretation trap |
|---|---|---|
| Persistent continuous dizziness | What eye-movement, hearing, gait and neurological findings accompany it? | Treating worsening with movement as proof of BPPV |
| Episodes without a clear trigger | What happened during the attack, and are vascular or other central features present? | Using a normal interval examination to erase the history |
| Brief reproducible position-triggered episodes | Does appropriate positional testing support the suspected canal disorder? | Performing HINTS on a person without the relevant syndrome |
| Light-headedness with circulatory or metabolic features | Is this actually the vestibular problem the rule addresses? | Forcing every use of dizzy into a vestibular algorithm |
The final row is deliberately outside the three vestibular patterns. Before applying a focused guideline, establish that the presenting problem belongs within its scope. Syncope, hypoglycaemia and an unstable circulation require their own assessment rather than a ritual sequence of eye tests.
Choose a focused examination for the pattern
A useful examination links each observation to a question. Eye movements, nystagmus, hearing, gait, cranial nerves, limb coordination and other neurological findings provide different information. Do not reduce the encounter to one named manoeuvre while ignoring an inability to sit or walk safely.
For acute vestibular syndrome with nystagmus, GRACE-3 recommends HINTS when performed by a trained clinician. Without nystagmus, its recommendation instead emphasises the severity of gait unsteadiness. For spontaneous episodic symptoms, the neurological history and examination guide concern about transient ischaemia and other causes. For a suspected posterior-canal positional syndrome, Dix-Hallpike has a different role. GRACE-3 recommendations.
These distinctions explain why an examination can be technically completed yet diagnostically misused. A test result does not travel independently of the population in which it was obtained, the state of the patient during testing or the examiner's competence.
Define when HINTS is applicable
HINTS combines head-impulse testing, assessment of nystagmus and testing for skew. Learning the names is not the same as learning the technique and interpretation. An article can explain the eligibility question, but cannot observe whether a reader identifies subtle eye movements correctly.
The practical pre-test check is therefore: does the patient have the relevant ongoing syndrome, is nystagmus present, can the examination be performed and interpreted reliably, and is a suitably trained clinician available? If a prerequisite is missing, an apparently reassuring result should not be treated as a validated exclusion of stroke.
A central or equivocal HINTS result prompts MRI/MRA in the GRACE-3 pathway. Its recommendation against routine first-line MRI when a trained HINTS clinician is available is not an instruction to avoid imaging when the examination is uncertain or concerning. SAEM's published recommendations.
Four original case pairs: change the history, change the task
Pair one: movement worsens symptoms versus movement starts them
In version A, a fictional adult reports continuous dizziness since the previous morning, vomiting and persistent nystagmus. Turning the head makes the existing sensation worse, but the person is not symptom-free between movements. The appropriate task is an acute-vestibular-syndrome assessment, including an appropriately trained bedside examination and attention to neurological findings.
In version B, the person feels normal while sitting still but has brief, reproducible episodes when rolling in bed. There are no supplied persistent symptoms or focal neurological findings. This history instead raises a triggered episodic problem in which positional testing may be appropriate.
The wrong answer in version A is "Dix-Hallpike because movement makes it worse". It ignores the continuous baseline. The wrong answer in version B is "reassuring HINTS rules out stroke". It imports a test into the wrong clinical setting. Both errors are failures to classify the task before performing it.
Pair two: persistent nystagmus versus no nystagmus
Version A describes ongoing dizziness with nystagmus and a clinician trained in HINTS. The learner must decide what the complete examination shows rather than selecting one convenient component. Hearing and the broader neurological assessment remain relevant.
Version B describes ongoing unsteadiness without nystagmus. The patient cannot walk independently and the reason is unresolved. Inventing a normal HINTS result is not a way to discharge this patient. The case requires assessment of gait severity, other neurological features and appropriate escalation of the remaining diagnostic concern.
The discriminating fact is not that one version sounds more dramatic. It is that the absence of nystagmus changes the applicability of the specific bedside strategy. This case pair applies the distinction in GRACE-3.
Pair three: a normal interval examination versus a concerning attack history
In version A, a fictional adult has repeated spontaneous dizzy episodes with a recognised recurrent pattern and no new neurological features in the information provided. A careful history still needs to establish the diagnosis and exclude relevant alternatives; the label cannot be assigned from recurrence alone.
In version B, the most recent episode included transient double vision and marked difficulty coordinating movement. The patient now appears better. The interval improvement does not cancel the features during the attack. Assessment for a possible central vascular event remains appropriate, with vascular imaging when indicated by the clinical concern.
A non-contrast CT is not the recommended discriminator between central and peripheral causes in this episodic setting. GRACE-3 recommends CTA or MRA where there is concern for TIA and large-vessel pathology. SAEM guidance.
Pair four: a trustworthy examination versus an uninterpretable one
Version A supplies an appropriately performed examination by a trained clinician, with a clear interpretation that fits the complete presentation. Version B uses the same symptoms but the examination is interrupted, the patient cannot cooperate and one component is uncertain.
The two records should not receive the same conclusion. In version B, the correct summary is that the bedside result is incomplete or equivocal, not negative. The next action must address the residual concern through senior assessment, suitable imaging or another appropriate pathway.
The learning point transfers beyond dizziness: missing evidence is not normal evidence. Repeating a label such as "HINTS negative" in the notes does not repair an unperformed or unreliable component.
Read recommendation strength without turning it into certainty about a patient
| GRACE-3 example, checked September 2026 | Published evidence description | What not to infer |
|---|---|---|
| Trained HINTS use in AVS with nystagmus | High certainty | Any clinician can obtain the same result |
| Hearing and gait assessment in the specified settings | Moderate certainty | One normal finding excludes every central cause |
| Dix-Hallpike for posterior-canal BPPV | Moderate certainty | Every position-sensitive symptom is posterior-canal BPPV |
| Short-term steroids for vestibular neuritis | Very low certainty | Treatment is obligatory or proves the diagnosis |
This small selection comes from SAEM's recommendation list. The certainty label concerns the evidence supporting that recommendation, not a numerical probability that an individual patient's diagnosis is correct.
Match disposition to the unresolved problem
Before deciding on discharge, articulate the working diagnosis, the evidence supporting it, unresolved alternatives, ability to mobilise, symptom control and feasibility of follow-up. A patient with a plausible peripheral explanation but unsafe mobility may still need a different immediate plan from a person who can function safely at home.
A learning debrief should record the timing pattern, the test's prerequisites, the finding that changed the plan and the residual uncertainty. Repeat the case after changing only one of those fields. This is more informative than collecting another generic list of central and peripheral causes.
Per iatroX product information, September 2026, the Socratic Tutor can use targeted follow-up after an attempted question to explore why a test was selected. That is reasoning practice, not certification of eye-movement examination or independent stroke assessment.
Frequently asked questions
Is HINTS appropriate for every patient who says the room spins?
No. Its role depends on the relevant acute vestibular syndrome, examination findings and a trained examiner, not the patient's choice of adjective.
Does a normal head CT exclude a central cause of dizziness?
No. GRACE-3 specifically advises against using routine non-contrast CT as the discriminator between central and peripheral vestibular causes.
Are symptoms worsened by head movement always BPPV?
No. Head movement may aggravate continuous symptoms; true triggering of brief episodes is a different history that requires its own assessment.