Article Text

Penlight-cover test: a new bedside method to unmask nystagmus
  1. D E Newman-Toker1,
  2. P Sharma2,
  3. M Chowdhury3,
  4. T M Clemons4,
  5. D S Zee5,
  6. C C Della Santina6
  1. 1
    Department of Neurology, The Johns Hopkins University School of Medicine, Baltimore, Maryland, USA
  2. 2
    Department of Neurology, The Johns Hopkins University School of Medicine, Baltimore, Maryland, USA
  3. 3
    Krieger School of Arts and Sciences, The Johns Hopkins University, Baltimore, Maryland, USA
  4. 4
    Department of Otolaryngology Head and Neck Surgery, The Johns Hopkins Hospital, Baltimore, Maryland, USA
  5. 5
    Department of Neurology, The Johns Hopkins University School of Medicine, Baltimore, Maryland, USA
  6. 6
    Department of Otolaryngology Head and Neck Surgery, The Johns Hopkins University School of Medicine, Baltimore, Maryland, USA
  1. Dr D E Newman-Toker, Department of Neurology, The Johns Hopkins Hospital, Pathology Building 2-210, 600 North Wolfe Street, Baltimore, MD 21287, USA; toker{at}


Background: Most patients with acute vestibular syndrome have vestibular neuritis or labyrinthitis. Some harbour strokes that can only be differentiated on the basis of subtle eye movement findings, including nystagmus. Peripheral nystagmus should be enhanced by removal of visual fixation. Current bedside methods for removing fixation require expensive equipment or technical skill not routinely available. We sought to test a new method for blocking fixation.

Methods: Proof-of-concept study for a new bedside oculomotor diagnostic test using an established physiological measurement of eye movements (electro-oculography (EOG)) as the reference standard. We sampled unselected patients undergoing caloric testing (surrogate model for neuritis) in an academic vestibular clinic. During the brief (30–60 s) decay phase of caloric-induced peripheral vestibular nystagmus, we shone a penlight in the left eye while intermittently occluding the right. We assessed nystagmus intensity (slow-phase velocity) clinically in all subjects and quantified change in two exemplar cases.

Results: Caloric responses frequently decayed before the test was complete, and artefacts rendered many EOGs uninterpretable during the short decay period. A clinically evident increase in nystagmus was seen 18 times in 10 patients and corroborated by EOG in 15. In quantified cases, slow-phase velocity increased as expected (mean change +42%) with fixation blocked.

Conclusion: The penlight-cover test could offer a low-cost, simple means of disrupting visual fixation in clinical settings where differentiating peripheral from central vestibular disorders is crucial, such as the emergency department. Prospective studies are needed to determine the test’s utility for excluding dangerous central causes among patients with suspected peripheral lesions.

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

Acute vestibular syndrome is characterised by the rapid onset of vertigo, nausea/vomiting, nystagmus, unsteady gait and head-motion intolerance over seconds to hours, lasting days to weeks. Most patients have a self-limited, presumed-viral cause for their symptoms known as vestibular neuritis or labyrinthitis, classified together as acute peripheral vestibulopathy (APV). An estimated 145 000 APV patients are seen annually in US emergency departments,1 but not all patients with the acute vestibular syndrome have APV. Brainstem and cerebellar strokes often mimic APV.26 Small studies suggest that 25% or more of acute vestibular syndrome presentations to the emergency department represent posterior circulation strokes.3 6 Since CT scans image the posterior fossa poorly7 and timely brain MRI is not uniformly available, bedside methods are needed to identify patients with these acute central vestibulopathies. In cases without obvious limb ataxia, dysarthria or other clear neurological signs, careful eye-movement assessment may be the only way to identify cerebellar stroke at the bedside.7

One hallmark of acute vestibular syndrome is spontaneous horizontal nystagmus, but this finding often looks similar in APV and acute unilateral cerebellar infarction.7 Some stroke victims may be identified on the basis of other bedside oculomotor signs, such as skew deviation or a normal head impulse test of vestibulo-ocular function, but this is not always the case.5 6 In contrast with central vestibular nystagmus, peripheral vestibular nystagmus is generally well suppressed by visual fixation.8 Therefore, in challenging cases it might be possible to distinguish APV from central causes of nystagmus by determining if the nystagmus is suppressed by visual fixation (fig 1).9

Figure 1

Peripheral versus central nystagmus in patients with acute vestibular syndrome. Adapted from Hotson JR, Baloh RW. Acute vestibular syndrome. New Engl J Med 1998;339:680–5. The direction of the arrows indicates the horizontal (predominant) direction of the fast phase of the nystagmus in patients with right acute peripheral vestibulopathy (APV) (top) or acute, left unilateral cerebellar stroke (bottom). Smaller torsional and vertical vectors are not shown. The thickness of the arrows indicates the relative intensity of the nystagmus. (A) Two examples of severe nystagmus evident in the primary position of gaze even under conditions of visual fixation. In this panel, central is distinguished from peripheral by two features: (1) change in direction when the patient looks laterally, opposite the predominant fast-phase direction (*); (2) lack of enhancement in amplitude and frequency with fixation blocked. (B) Mild nystagmus evident only in lateral gaze to one side. In this panel, central is distinguished from peripheral solely by lack of enhancement in amplitude and frequency with fixation blocked. The top half of both panels suggests that suppression of peripheral nystagmus by visual fixation is often incomplete. However, in some cases (not shown) peripheral nystagmus is so well suppressed that patients with APV may appear to have no nystagmus at all, even in lateral gaze, until fixation is blocked.

To elicit subtle APV nystagmus at the bedside or to assess if spontaneous nystagmus is the result of a peripheral lesion, it is necessary to examine the eyes both with and without the patient visually fixing on a target. This requires a technique to remove visual fixation. Having the patient close both eyes or examining the patient in total darkness are conceptually simple but require electro-oculographic (EOG) or infrared video monitoring techniques to assess any change in nystagmus. In neuro-otological practice, peripheral vestibular nystagmus is generally unmasked using Frenzel goggles, which place a +20 or +30 dioptre lens in front of each eye, disrupting fixation through visual blur, and giving the examiner a magnified view of the eyes.10 Unfortunately, Frenzel goggles cost ∼$500–1500 (US) and are manufactured by only a handful of companies worldwide,11 making them effectively inaccessible to all but a few subspecialist providers. More practical bedside techniques to block visual fixation are needed.

An ideal technique would permit quick switching back and forth between the conditions of visual fixation and fixation blocking to allow serial comparison of nystagmus with and without fixation suppression. One solution involves intermittently interposing a featureless white piece of paper immediately in front of the patient’s eyes, obscuring their view of the visual surround.10 However, this Ganzfeldt technique requires the examiner to peek at the nystagmus around the paper from a lateral position, which is quite challenging for the untrained observer. An elegant method of assessment, requiring only a handheld ophthalmoscope, was described by Zee in 1978.12 The technique uses a bright light (ophthalmoscope) to disrupt fixation continuously in one eye, while fixation in the other eye is intermittently blocked by manual occlusion. This method of occlusive ophthalmoscopy disrupts fixation in both eyes, but does not obscure the examiner’s view, allowing continuous monitoring of the nystagmus by observing the oscillatory motion of the optic disc. Unfortunately, the technique demands comfort and dexterity with ophthalmoscopy along with favourable funduscopic viewing conditions (eg, absence of cataracts). We sought to develop and test a new method for blocking visual fixation that capitalises on the principles of occlusive ophthalmoscopy but is less reliant on examiner skill.


This was a proof-of-concept study for a new oculomotor diagnostic test using an established physiological measurement of eye movements (EOG) as the reference standard. Our overall goal was to describe and pilot a new bedside method to block visual fixation using a human model of APV (caloric testing). A convenience sample of non-consecutive, adult (⩾18 years) patients undergoing caloric testing in an academic vestibular clinic in July–September 2008 consented to participate in this Institutional Review Board-approved research study. The sole inclusion criterion was routine caloric testing for clinical purposes. There were no exclusions.

Patients underwent standard binaural, bithermal water caloric irrigation13 in the sequence right ear warm (44.1°C), left ear warm (44.1°C), left ear cool (30.5°C) and right ear cool (30.5°C) using a temperature-switching technique.14 Patients lay in the semirecumbent position with the head of bed elevated ∼25°. Eye movements were recorded using standard EOG techniques and plotted graphically (ICS Charter ENG, GN Otometrics, Taastrup, Denmark). For each irrigation, the EOG trace is normally stopped after the peak slow-phase velocity has passed and nystagmus begins to decay. In this study, we continued the EOG trace during this brief decay period,13 which served as our surrogate model for acute unilateral vestibular hypofunction (ie, APV).

At the start of the decay period, all subjects were instructed to visually fixate on a ceiling target (“X”) at ∼1.5 m. A handheld penlight (∼0.5 W LED; ∼20-lumen bulb; two AA batteries) was shined in the left eye continuously to obscure visual fixation while the right eye was intermittently covered using a plastic occluder (video 1). We used a ∼14 s cycle (∼7 s occlusion; ∼7 s fixation). This cadence was short enough to achieve several cycles during a typical 30–60 s decay period but allowed time for the examiner holding the penlight to visualise any changes in nystagmus under the two viewing conditions.

We gathered basic demographic information from patients along with clinical diagnoses. The examiner identified patients with a clinically evident change in nystagmus and reviewed their EOG tracings. We were interested in proof of concept that this approach would work, so we only formally quantified the change in nystagmus intensity (slow-phase velocity) for two exemplar cases. We compared clinical nystagmus parameters in responders and non-responders to assess differences and controlled for demographic variables using a multiple logistic regression model. Data were analysed using Microsoft Visio 2003 (Microsoft, Redmond, Washington) and Excel 2003, and comparative statistics calculated using Stata v6.0 (College Station, Texas). Two-sided p values <0.05 were considered significant.


We screened 34 patients and 31 (20 female) consented to participate. The age range was 24–93 (mean 47 years, SD 15). Clinical diagnoses were mostly suspected vestibular migraine (10), Meniere disease (eight), or benign paroxysmal positional vertigo (five). None were diagnosed as having cerebellar stroke or other structural central cause. Among 31 participants, 30 had intact caloric nystagmus ⩾6°/s on at least one irrigation cycle.

The test was attempted during 97 irrigations (some patients had absent caloric responses or refused repeated examination). Caloric responses decayed before we could complete even one penlight-cover test cycle in roughly half of these. Many of the remaining EOG traces revealed poor fixation, intrusive eye movements, blinks or other artefacts that made interpretation of the nystagmus difficult during the short decay period. A clinically evident increase in nystagmus during the occlusive phase of the penlight-cover test was noted 18 times in 10 patients. This clinical impression was corroborated by EOG in 15. As shown in the two exemplar cases (fig 2), the increase in nystagmus intensity in the “fixation-blocked” relative to “fixation” phases of the penlight-cover test confirms that subjects suppressed their (caloric-induced) peripheral vestibular nystagmus during fixation, and the bedside test blocked that fixation suppression.

Figure 2

Sample electro-oculographic tracings in two patients during the penlight-cover test. C, cover; U, uncover. Shown are two exemplar EOG tracings demonstrating the penlight-cover test conducted during the nystagmus decay period following a caloric irrigation. Each vertical box corresponds to 10° of horizontal eye position (up is right, down is left) and each horizontal box corresponds to 1 s. Three representative segments in each trace were quantified. As anticipated, the patient’s nystagmus velocity (geometric slope of the nystagmus slow-phase) and amplitude (vertical height of the nystagmus fast phase) both decay from fixation period “A” to fixation period “B” as the caloric nystagmus abates. During the fixation blocked period, the “baseline” (unblocked) nystagmus parameters are approximated by the midpoint values between fixation periods “A” and “B.” Although the most obvious shift from anticipated baseline is in nystagmus amplitude (+187%), the more meaningful change is in the nystagmus slow-phase velocity (+57%), since it is this change in velocity that indicates effective blocking of fixation suppression.

Not surprisingly (given so many caloric responses decaying before the penlight test began), the average peak nystagmus slow-phase velocity for penlight-cover test responders was higher than for those without a clinically evident response (32°/s vs 21°/s, t test p = 0.008). The mean slow-phase velocity across the four caloric runs remained a predictor of a positive response even after controlling for age and gender.


This proof-of-concept study under laboratory conditions demonstrates that the penlight-cover test can reveal peripheral vestibular nystagmus by blocking visual fixation. This new bedside method is easy to apply, requiring only a penlight and the examiner’s hand or other occlusive device. While this test offers the potential to help discriminate between benign peripheral (ie, APV) and dangerous central (ie, stroke) causes of the acute vestibular syndrome for physicians without access to expensive neurovestibular equipment, additional studies in relevant clinical populations are essential before this test can be recommended for clinical use.

Clinicians who consider “testing the test” in clinical practice should be aware that neither an increase in nystagmus amplitude with fixation blocked nor an increase in frequency during fixation represents a “positive” (peripheral-type) response. Amplitude will increase during fixation blocking with either peripheral or central nystagmus because of an associated decrease in beat frequency (and vice versa). A true positive response requires an increase in nystagmus intensity (ie, slow-phase velocity), representing the mathematical product of amplitude and frequency. Since it is difficult for a human observer to discern small shifts in nystagmus intensity (velocity) when brisk nystagmus is present, the test will likely prove most helpful when the patient has minimal or no nystagmus under conditions of fixation (fig 1B, top). Under those conditions, the penlight-cover test will unmask a mild peripheral vestibular nystagmus that is mostly or completely suppressed by fixation. In such patients, the examiner would simply need to note the re-emergence of nystagmus during the “fixation-blocked” phase to identify those with a presumed peripheral vestibular cause (video 2).

Limitations include a suboptimal surrogate for APV, non-representative sampling, unmasked observers and lack of comparison to existing fixation-blocking methods such as Frenzel goggles. Our rate of positive responses was low, but mostly attributable to unrealistic test conditions. The penlight-cover test is easier to conduct in real APV patients (video 2), since they have nystagmus (or occult vestibular imbalance) that is continuous rather than rapidly decaying over 30–60 s. From this small sample, we cannot know if the test will perform uniformly across subjects (ages, diseases, etc). Nevertheless, these data represent proof-of-concept evidence that the penlight-cover test should work to unmask suppressed nystagmus in actual practice.

In conclusion, the penlight-cover test offers a low-cost, simple means of disrupting visual fixation. This technique may be more realistic than Frenzel goggles or occlusive ophthalmoscopy for most frontline providers and general neurologists. While the procedure is simple, it has not been rigorously tested in clinical practice and, for now, should be used only in conjunction with proven diagnostic methods in the assessment of acute vestibular patients. Prospective studies are needed to determine the test’s utility for helping exclude dangerous central causes (eg, cerebellar stroke) among patients with suspected peripheral lesions (eg, APV).


Supplementary materials


  • ▸ Additional videos are published online only at

  • Funding: The preparation of this manuscript was supported by grants from the National Institutes of Health (NIH RR17324-01) and the Agency for Healthcare Research and Quality (AHRQ HS017755-01).

  • Competing interests: None.

  • Ethics approval: Ethics approval was provided by the Johns Hopkins Medicine Institutional Review Board.