Rudolph's Pediatrics, 22nd Ed.

CHAPTER 581. Special Tests of the Eyes

Daniel J. Karr and Alex V. Levin


The electroretinogram (also known as full-field ERG) is a measure of retinal photoreceptor integrity and thus retina function. The basic technique consists of measuring the action potential produced by the retina when stimulated by light of variable intensity and color. ERG allows differentiation between the responses of the retinal rod and cone systems. A contact lens with an imbedded electrode is placed on the cornea, or a skin electrode is placed on the lower lid. A reference skin electrode is placed on the forehead. The ERG can be performed awake in infancy and later childhood but usually requires sedation or general anesthesia between the ages of 1 and 6 years. The first step of the ERG is dark adaptation, which is done by double patching both eyes for at least 20 minutes. Then, in a completely dark environment, the test is started. Sleeping or awake, quiet infants and children under sedation or anesthesia lie supine while the bowl-shaped machine is brought over their face. Older children can sit upright and their chin is placed on a rest within the bowl. As variable light stimuli are presented within the bowl, the electrical potential between electrodes is then measured and recorded as a waveform. In anxious children dim light adaptation (mesoptic vision) can be substituted.

The normal ocular ERG recording is biphasic, consisting of a negatively deflected a-wave directly generated by photoreceptors and a positively deflected b-wave produced by the cells that transmit the electrical potential generated by the rods and cones to the ganglion cells, which ultimately carry the message to the visual cortex. The amplitude and configuration of the wave is a measure of photoreceptor integrity and function. The ERG consists of a series of dark-adapted (scotopic) and light-adapted (photopic) recordings. The scotopic ERG measures the rod system function, and photopic stimulation assesses the cone system. The photopic ERG occurs after 10 minutes of relatively bright light adaptation, which causes the rods to be bleached out and nonresponsive. Likewise, the cone system is assessed by the flicker response using 30 Hz light stimulation, to which rods cannot cycle quickly enough to record.

The ERG is helpful in diagnosing several conditions, such as generalized (eg, retinitis pigmentosa) and localized genetic retinal degeneration (eg, macular dystrophies). It is also helpful in assessing retinal function following retinal vascular occlusions and in determining potential retinal function when the retina cannot be clearly viewed due to opaque media (eg, cataract). Although the ERG does not test visual acuity (which is a measure of foveal function, an area of retina too small for the ERG to assess, as it is a mass retinal cone or rod response), it is useful in assigning the anatomic location of visual compromise in children with poor vision and nystagmus. We recommend that all children with no other obvious cause for nystagmus receive an ERG. Specific pediatric considerations include Leber congenital amaurosis, rod monochromatism (achromatopsia), and congenital stationary night blindness (CSNB), all of which can present with nystagmus, decreased vision, and a normal retinal examination. Even when the ERG is completely flat, the central vision can be 20/20 if the tiny fovea is spared.

Centers that perform ERG must have control values matched for age, as the infant ERG normally has far less amplitude than that of older children.1 Sometimes it is necessary to repeat the ERG annually or every 2 to 4 years to differentiate between static and progressive conditions.


The multifocal ERG results in a topographical three-dimensional map of retinal function and is used to assess retinal function in much smaller areas than possible with full-field ERG. For example, the mfERG can distinguish responses directly from the fovea, which is useful in patients with reduced vision (although the test requires vision good enough to allow fixation) but otherwise normal-appearing retina and a normal full-field ERG. Multifocal ERG responses can be recorded simultaneously from many individual retinal locations, generating a response curve for each location that can then be presented as a topographical map of retinal function. This may be used with any disorder affecting retinal function but is most helpful for disorders affecting the macula.2 As the test does require some fixational compliance, it is usually reserved for children over the age of approximately 5 years old. Techniques for mfERG are not commonly available for use under sedation or general anesthesia.


The electrooculogram is a measurement of the standing potential between the positively charged cornea and the electrically negative posterior segment of the eye. It is used to measure, in particular, the activity of the retinal pigment epithelial cell (RPE) layer. In order to affect the EOG response, very widespread disease of the RPE is needed.

An electrode is placed on the skin at the medial and lateral angles of the lid. Testing occurs in both light- and dark-adapted states. The patient then looks back and forth rhythmically between left and right gaze. The patient must be awake and compliant; therefore, the test is usually reserved for older children. As the positively charged cornea moves toward each electrode, it makes the nearer electrode positive relative to the other side. The electrical potential difference between the two electrodes is recorded as a numerical ratio and compared to normal values. The EOG is generally used to investigate possible genetic degenerations of the retina, in particular those that affect the macula. The paradigm disorder that demonstrates a very abnormal EOG with a normal ERG is Best disease (vitelliform macular dystrophy).


Visual evoked potential testing evaluates the postretinal pathway to the visual cortex. Stimulation of the retina is used to generate recordings of the electrical activity from the visual cortex. The stimulation used is typically a black-and-white checkerboard pattern that reverses at a specified rate of reversal (pattern VEP); graded, alternating black-and-white vertical lines moving across a monitor (sweep VEP); or a flash of light (flash VEP). Each technique has specific diagnostic uses.3,4 Electrodes are placed on the scalp over the visual cortex. The waveform is evaluated both in terms of its amplitude and its latency to maximum response following the stimulus.

The VEP is dominated by activity from the central 5% of the retina. It can therefore be used as an indicator of macular function and vision in children. Unfortunately, the VEP is less helpful in accurately assessing vision in some situations where it is most needed: children with severe neurological compromise, seizures, nystagmus, and very poor vision. It has particular usefulness in preverbal children and children who may have functional visual loss or malingering.

One special VEP technique, the multichannel (also known as 3-lead) VEP, can assess chiasmal function. Characteristic asymmetry of the chiasmal decussations of visual pathway axons is noted in albinism, achiasmia, unilateral severe microphthalmia or anophthalmia, severe amblyopia, and chiasmal compression.


The visual field evaluation is a determination of the limits of side (peripheral) vision but is also used to assess whether a patient may have “blind spots” (scotoma). The normal visual field extends farther in the temporal (90°) and inferior (80°) directions compared to superiorly and nasally (approximately 60° each).5 The optic nerve represents an absolute blind spot, since there are no retinal receptors at that location. Each eye is measured individually. The physiological blind spot from the optic nerve appears just temporal to the midline in each eye. The visual field provides invaluable information regarding both optic nerve and retinal function.

The stimulus used to evaluate the visual field is a light target of varied size and luminance. Accurate testing with either a manual visual field testing system, such as the Goldmann test, or a computerized perimetry, such as the Humphrey test, takes a considerable amount of concentration and compliance. In younger children, visual field testing is often difficult or impossible, with results unreliable and requiring retesting at a later age. As Goldmann testing is done with a technician who can coax and assist the child, it is usually more useful for younger children. Children may be well into their teens before they can accurately do automated testing. Special perimeters and methods have been developed for those unable to be tested by Goldmann perimetry because of age or ability/disability.6

Formal visual field testing is useful when confrontation testing is abnormal or in patients with disorders known to affect the visual pathways (eg, brain tumor, stroke, optic nerve disease) or macula and retina (eg, scars, retinal dystrophy, vascular occlusion). It is also helpful in monitoring the progress of patients with glaucoma.


A-scan ultrasound utilizes a single ultrasound source to provide vertical spike images that indicate distances between specific interface structures. It can be used to measure lens thickness, size (axial length) of the eye, and size of ocular components.7 In awake, compliant children, the probe is placed on the eyelid with a contact gel with topical anesthetic drops. In children under sedation or general anesthesia, the probe can be placed directly on the cornea. An immersion technique where the probe is in a small water bath allows for the most accurate measurements of axial length.

B-scan ultrasound provides information regarding the inside of the eye, in particular those structures behind the lens, including the vitreous, the retina, and the optic nerve (Fig. 581-1). It can also visualize the distal portions of the extraocular muscles and the anterior retrobulbar tissues. It can assess the quality of an intraocular lesion (eg, tumor), including its size and shape and how it relates to other structures. The probe is placed in the same fashion as described for A-scan, although immersion is not needed. B-scan is particularly helpful when it is not possible for the ophthalmologist to view inside the eye (eg, cataract) and is helpful for evaluating intraocular tumors and foreign bodies and delineating abnormal anatomy such as retinal detachment. Depending on the frequency of the transducer utilized, the B-scan can provide imaging information from the anterior segment as well.8 Ultrasound biomicroscopy (UBM) is a direct contact technique that allows for remarkable high-resolution views of the anterior segment (Fig. 581-2).


Optical coherence tomography (OCT) is a technique similar to B-scan ultrasonography, which uses light instead of sound waves. It is a noncontact, noninvasive imaging system that is used to produce very-high-resolution cross-sectional images of intraocular structures, including the retina (Fig. 581-3); optic nerve; vitreous; or, with another version of the machine, the anterior segment. The quality of these images is fine enough to delineate the anatomic layers of the retina or cornea. The OCT is used to diagnose and monitor specific retinal macular problems such as retinal swelling, membranes, and holes.9 It can determine tractional connections between the retina and vitreous and can measure nerve fiber layer thickness (eg, in glaucoma) and optic nerve head anomalies. The test does require patients to be awake and in a sitting position so they can fixate on the light source. As a relatively new technique, normal values are still being generated for children. A similar test—high-resolution tomography (HRT)—is also available.


Fluorescein angiography is a diagnostic technique that provides progressive and sequential visualization and documentation of blood flow through the retina; the choroid; and, if specifically requested, the iris (not commonly indicated). Variations in the blood flow through the structures provide information regarding the understanding and treatment of ocular vascular disorders. The patient first receives pharmacological dilation of the pupils. The basic technique consists of a patient seated in front of a retinal (fundus) camera. Fluorescein dye is injected intravenously after obtaining baseline retinal and optic nerve photographic images. After injection, sequential photographic images are taken with a special filter that allows visualization of the dye in the ocular vasculature (Fig. 581-4). Young children may require this procedure to be done under sedation or general anesthesia using a type of contact retinal photography called the RetCam.

FIGURE 581-1. B-scan ultrasound. The eye is facing to the left; * indicates vitreous. Short arrow indicates retina. Long arrow is optic nerve. Arrow heads indicate extraocular muscles.

FIGURE 581-2. Ultrasound biomicroscopy (UBM) of the anterior segment. C, cornea; I, iris; L, lens; P, pupil.

FIGURE 581-3. Optical coherence tomography (OCT) of retina. Short arrow indicates fovea. The upper layer of the image is closest to the inside of the eye. Long arrow indicates photoreceptor layer (rods and cones); * indicates choroid.

FIGURE 581-4. Intravenous fluorescein angiography of the retina.