Ординатура / Офтальмология / Английские материалы / Clinical Ocular Pharmacology 5th edition_Bartlett, Jaanus_2008

It is not clear, and it may be unreasonable to draw conclusions from these studies as to the community standard in local areas; however, gonioscopy appears to be an underperformed procedure.

Expected Gonioscopic Findings. From anterior to posterior, the following structures are present in the angle: Schwalbe’s line (representing the posterior border of Descemet’s membrane), the anterior trabecular meshwork (often less pigmented than the posterior trabecular meshwork), the canal of Schlemm within the boundaries of and deep to the trabecular meshwork (typically only visible if filled with venous blood), the posterior trabecular meshwork, the scleral spur (to which the ciliary muscle is attached), and the ciliary body band.

Care must be taken to distinguish these structures from clinical entities that can simulate normal anatomy. For instance, pigment from the structures of the anterior chamber can accumulate on and adjacent to Schwalbe’s line. This pigment deposition can give the false impression of a normal trabecular meshwork and an open angle. This pigmented band is referred to as Sampaolesi’s line. The appearance of the trabecular meshwork can also mislead the practitioner into believing that the nonpigmented or lightly pigmented anterior trabecular meshwork, followed posteriorly by a pigmented portion of the trabecular meshwork, is actually the scleral spur and ciliary body.This appearance is due to the fact that the meshwork extends anteriorly beyond the region that is primarily responsible for outflow of aqueous. In the region closest to the outflow, pigment tends to accumulate in greater amounts than in the region adjacent to it.

As a general rule, the width of the ciliary body band is generally equal to or less than that of the trabecular meshwork. If the width is greater, it is typically symmetric between the eyes or may represent an angle anomaly such as angle recession. In addition, the width of the ciliary body band is generally greatest in the inferior quadrant and at its thinnest in the superior quadrant.

Gonioscopic Instruments. Three-mirror and four-mirror lenses are the most commonly used in clinical practice. The four-mirror lens has the advantage of less mess (gonioscopic solutions such as Goniosol are not required), a more rapid procedure, and greater patient comfort. Care should be taken, at least initially, to not press too firmly on the cornea to avoid mechanically opening the angle during observation. Indenting the cornea subsequent to this initial observation (indentation gonioscopy) may be useful in determining the actual location of iris insertion if it is not otherwise visible. The three-mirror lens requires a contact solution but has the advantage of a more stable image in a blepharospastic patient (once the lens is on) and having additional mirrors, which serve other purposes (such as contact funduscopy).

Gonioscopic Assessment. The assessment of the angle by either technique should include the entire angle circumference and should be augmented by changing light levels to simulate the angle architecture in different environments (e.g., dimming the lights may demonstrate a crowding of the angle by the dilating iris that might have been missed under brightly lit conditions). In addition to documenting the posterior-most structure and the presence of angle pathology (e.g., angle neovascularization, neoplasm, peripheral anterior synechia, heavy pigmentation, angle recession), an assessment of the peripheral iris profile (e.g., steep, regular, concave, and plateau), including the presence of iris–trabecular meshwork contact, should be made. This profile may vary in different levels of illumination and during indentation gonioscopy, which assist in differentiating apparent angle depth from actual depth and appositional versus synechial iris–trabecular contact.

New Anterior Chamber Technologies. Anterior segment ocular coherence tomography allows for precise evaluation, measurement, and analysis of the anterior segment, including anterior chamber depth, anterior chamber angles, and the angle-to-angle distance (anterior chamber diameter). It can also assist in postoperative evaluation because it allows imaging and measurement of intraocular lenses and ocular implants.The procedure is relatively fast and does not contact the eye. It can be performed in complete darkness as well as in brightly lit surroundings (to assist in the dynamic assessment of the angle). The images are digitally documented, so they can be magnified, enhanced, transmitted, and measured. In addition, a technician can take the image, freeing the doctor to focus time on assessing the results.

It seems likely that manufacturers will develop archived clinical databases in future permutations of this technology. This would permit comparison of parameters such as the anterior chamber depth and configuration of the anterior chamber angle with an internal database. Probability analysis could be generated to determine the extent of deviation from a norm or the risk of angle closure.

Clinical Pearls

•Gonioscopy is important for every patient with glaucoma (unless contraindicated).

•Gonioscopy is important for every patient who fails angle screening (e.g., van Herick technique) at the slit lamp.

•Gonioscopy should be performed statically (dim lights, no indentation) and dynamically (increased illumination and/or indentation as needed).

•Gonioscopy should include an observation of all 360 degrees of each angle.

•The clinician should be aware of anatomic masqueraders:

Sampaolesi’s line appearing as trabecular meshwork

Lightand dark-banded trabecular meshwork appearing as scleral spur and ciliary body

676 CHAPTER 34 The Glaucomas

•The peripheral iris profile should be observed and documented as steep, regular, concave, or plateau and the presence of iris–trabecular meshwork contact (either appositional or synechia).

•The clinician should observe and document secondary etiologies of glaucoma.

•This procedure should be repeated every 3 to 5 years unless otherwise indicated.

Structural Assessment of the Optic Nerve Head and Retinal Nerve Fiber Layer

In many instances, structural changes of the optic nerve head or retinal nerve fiber layer provide the first clinical evidence of glaucoma.The assessment of these structures has improved dramatically over the past decade as scanning laser ophthalmoscopy has become more available. A consensus document by the Association of International Glaucoma Societies (which includes the Optometric

Glaucoma Society and the American Glaucoma Society) suggests that the introduction of these devices has enhanced the community standard by enabling clinicians with less experience to function at a level that is closer to their experienced counterparts.The use of new technologies is becoming increasingly common.These new instruments augment but do not replace a careful clinical examination and will likely play an increasing role in management decisions in the future.

Methods of Clinical Assessment of the Optic Nerve and Retinal Nerve Fiber Layer. The clinical assessment of the optic nerve and retinal nerve fiber layer is typically conducted using indirect ophthalmoscopy at the slit-lamp through a dilated pupil. This affords a stereoscopic assessment of the deviations from normal optic nerve architecture that could be overlooked with the direct ophthalmoscope or retinal photography. However, these latter two techniques often provide very useful information that could be missed during indirect ophthalmoscopy. Therefore all three devices have a role in the assessment of the optic nerve and retinal nerve fiber layer. Further assessment of optic nerve and retinal nerve fiber layer parameters may be augmented using any of a variety of scanning laser ophthalmoscopes. The clinician assesses the data produced by these devices and correlates the results with the clinical gestalt acquired by assessing these structures directly. Although the scanning laser devices make comparative analyses against an internal normative database, the clinician makes a more comprehensive analysis against his or her clinical experience.

Direct Ophthalmoscopy. The direct ophthalmoscope is perhaps an underutilized instrument in the assessment of glaucoma. It can provide information regarding pupil function, an estimation of the anterior chamber angle depth, spherical refractive error of the patient, presence of media opacity, and a magnified view of the optic nerve

that can be enhanced with the use of filters (e.g., red-free). The size symmetry of the nerve can be assessed by using the 5-degree spot size as a reference, and the nerves can be compared with each other because the procedure allows for a relatively rapid assessment between each eye. The magnified view of the optic nerve head enables the practitioner to carefully assess the vasculature of the nerve in ways that could be overlooked by other means of assessment. Monocular cues to depth such as deflection of vessels, although not as robust as true stereoscopic view, can augment the clinical assessment. In addition, the direct ophthalmoscope is portable, the image is “right side up,” and the instrument is more accommodating to patients who have difficulty at the slit lamp.

Indirect Ophthalmoscopy. Indirect ophthalmoscopy of the optic nerve head and retinal nerve fiber layer affords a three-dimensional view of these structures, which provides the observer with a sense of depth that is often lacking with the direct ophthalmoscope. Most experienced practitioners acknowledge that their impression of the integrity of the neural retinal rim of the optic nerve can be tremendously different during a stereoscopic versus a monocular assessment. Too often, color cues afforded by direct ophthalmoscopy can mislead the practitioner into believing that the neural retinal rim is more intact when compared with the stereoscopic assessment. Indirect ophthalmoscopy (i.e., the stereoscopic assessment of the optic nerve, typically with condensing lenses used at the slit lamp), although generally thought of as the gold standard and an essential component of the assessment of a patient with or expected of having glaucoma, is not without its limitations. Interobserver reliability (the degree of agreement in the assessment of ophthalmoscopic findings between two or more practitioners) is not great (even between experienced practitioners). Additionally, the assessment of the retinal nerve fiber layer is, at times, very challenging, especially in the presence of a lightly pigmented retina or media opacity. The presence of optic disc (Drance) hemorrhages, arguably one of the most significant clinical findings suggestive of future compromise of the neural retinal rim and corresponding visual field, can also be overlooked. Remarkably, a significant percentage of these hemorrhages are missed by experienced practitioners but are easily observable on a retinal photograph (even in the absence of a red-free filter).This may be due, in part, to using a light source at the slit lamp that bleaches the image of the hemorrhage. In summary, although indirect ophthalmoscopy should be viewed as a highly recommended procedure for all glaucoma patients, it should, whenever possible, be augmented by other techniques.

Retinal Photography. Photography of the optic nerve head and retinal nerve fiber layer has the advantage of offering varying levels of magnification, filters (e.g., red-free), and a stable image even in the setting of a patient with poor

fixation or nystagmus. In addition, particularly in digital format where the image is quickly accessible, the photograph provides an excellent opportunity to educate the patient about the nuances of his or her particular optic nerve, the effects of glaucoma on the nerve, and the importance of regular monitoring of its structural integrity. These extra few minutes go a long way to demystify this symptomless chronic disease.

A photograph also reveals disc hemorrhages and retinal nerve fiber layer defects that can be overlooked by other methods (including the scanning laser ophthalmoscopes) and is an excellent way of documenting the optic nerve head, particularly if stereoscopic pairs are created. Unfortunately, commercial access to cameras that allow for simultaneous stereo photography is limited. To minimize the effects of photographic stereo artifacts associated with manually offsetting the camera, care must be taken to be as consistent as possible.

Clinical Assessment of the Optic Nerve and Retinal Nerve Fiber.

It is important to approach the assessment of these structures in a consistent and organized manner with several key parameters noted for every optic nerve. One way to keep this assessment organized is the mnemonic CARVES, because glaucoma “carves” out the optic nerve (Courtesy of Nick Holdeman, OD, MD, and Jade Schiffman, MD):

C = Color (e.g., pink or pale)

A = Angle of the disc (e.g., deep, saucerized, tilted) R = Rim tissue and nerve fiber layer

V = Vessels (e.g., bearing, bayoneting, disc hemorrhages) E = Extrapapillary features (e.g.,zone beta,myopic crescent) S = Size

Size, Shape, and Symmetry. The size and shape of the optic nerve influence the appearance of the neural retinal rim in significant ways. Larger discs, in general, have larger cups, and vertically elongated discs tend to have neural retinal rims that appear thinner at the long axis (superiorly and inferiorly) when compared with round rims. Round rims are more likely to follow the ISNT rule of the neural retinal rim: The thickness of the rim tends to be greatest in the inferior quadrant (I), followed by the superior (S), nasal (N), and then temporal (T) quadrants. This phenomenon is due to the convergence of the retinal ganglion cells from the superior and inferior arcades (the larger proportion of the entire population of retinal ganglion cells) onto the superior and inferior rim, in tandem with the branches of the central retinal artery and vein that also occupy this area. Although the ISNT rule is a useful guideline, it is not without its documented limitations and should be used with a degree of caution.

The disc size is also important for other reasons. Whereas a 0.7/0.8 cup-to-disc ratio might have a normal neural retinal rim in a large vertically elongated disc, this same ratio could be quite abnormal in a smaller disc. In addition,a 0.5 cup-to-disc ratio in a small disc would take

on added clinical significance if the previous assessment was significantly less. Said another way, the clinician should monitor small and large discs with equal diligence because subtle changes may easily be overlooked.

Disc symmetry is also a critical element in the assessment of the nerve. A common misinterpretation of “asymmetric cupping” is often asymmetric discs (recall large discs have large cups). Disc asymmetry is difficult to appreciate without some form of measurement from one eye to the next (the measurement need not be quantifiable; it can be a simple comparison of one eye to the next).There are several methods to assess the size of the disc. One simple way is to compare photographs taken at the same level of magnification or comparing the sizes relative to a known area (e.g., the 5-degree spot during direct ophthalmoscopy). Another is at the slit lamp when matching the length of the light beam to the long and short axis of the nerve and comparing this length from one eye to the other. Certain scanning laser ophthalmoscopes also measure the area of the disc.

Integrity of the Retinal Nerve Fiber Layer. The utility of the assessment of the retinal nerve fiber layer has been known for several decades. It was not until the wide distribution of digital retinal photography and the introduction of scanning laser ophthalmoscopy that the assessment of this structure became more commonplace. Before digital photography, the nerve fiber layer assessment was conducted either with direct or indirect ophthalmoscopy (often with auxiliary filters) or by nerve fiber layer photography (also with auxiliary filters and, often, very specialized photographic film). Debates ensued over the best way to assess this structure but often with the knowledge that this level of assessment was beyond the community standard outside of academic institutions or well-known glaucoma practices. The retinal nerve fiber layer, especially in lightly pigmented fundi as a backdrop or in the presence of any significant media opacity, was a challenge to visualize.

Documentation was accomplished by drawing. Photography was equally challenging but had the added disadvantage of a delay in the processing time. With the introduction of high-quality digital photography, page proofs are “developed” in seconds, allowing the photographer to make adjustments to lighting and focus in real time. As such, the ability to photodocument the retinal nerve fiber layer has been greatly enhanced. The introduction of scanning laser ophthalmoscopes enhanced this measurement even further. These technologic advances have reawakened the clinical practitioner to the importance of the assessment of this structure. Although limitations of the assessment by ophthalmoscopy remain, the assessment and documentation of the retinal nerve fiber layer have become a community standard in the evaluation of a patient with or suspected of having glaucoma.

678 CHAPTER 34 The Glaucomas

Peripapillary Atrophy. Peripapillary atrophy is not an uncommon condition and is not a sensitive means of differentiating glaucomatous from nonglaucomatous patients (especially in early glaucoma). However, the size, location, and changes in areas of peripapillary atrophy may have some significance for patients with glaucoma.

Beta Zone and Alpha Zone. Peripapillary atrophy is sometimes divided into two distinct zones, each with different underlying histopathologies. One zone, the alpha zone, appears as an irregular hypopigmentation and hyperpigmentation and thinning of the chorioretinal tissue layer. It is bordered anteriorly by the retina and posteriorly by either the beta zone or the scleral ring. Histopathologically, it corresponds to pigmentary irregularities in the retinal pigment epithelium. Psychophysically, this defect corresponds to a relative scotoma.

The beta zone is characterized by marked atrophy of the retinal pigment epithelium and of the choriocapillaris, good visibility of the large choroidal vessels and the sclera, and thinning of the chorioretinal tissues. It correlates histopathologically with a complete loss of retinal pigment epithelium cells and markedly diminished count of retinal photoreceptors. This defect corresponds psychophysically to an absolute scotoma.

In normal eyes both alpha and beta zones are largest and most frequently located in the temporal horizontal sector, followed by the inferior temporal area and the superior temporal region. They are smallest and most uncommonly found in the nasal peripapillary area. If both zones are present, the beta zone is always closer to the optic disc.

Alpha zones are present in almost all normal eyes and are thus more common than beta zones. Alpha and beta

zones must be differentiated from the myopic scleral crescent in eyes with high myopia.

Both zones are often larger, and the beta zone occurs more often in eyes with glaucomatous optic nerve atrophy and may be correlated with thinning of the neural retinal rim and visual field loss (Figure 34-1). In unilateral glaucoma, a beta zone, if present, is found significantly more often in the affected eyes than in the contralateral nonglaucomatous eyes. Increases in the size of the beta zone may suggest progression of glaucoma in some patients.

Clinical Pearls

•Assessment of the optic nerve head and retinal nerve fiber layer should occur for all glaucoma patient and glaucoma suspects.

•Indirect ophthalmoscopic assessment of the optic nerve head and retinal nerve fiber layer at the slit lamp should, whenever possible, be augmented by other techniques such as direct ophthalmoscopy, retinal photography, and/or scanning laser ophthalmoscopy.

•An assessment of the optic nerve and retinal nerve fiber layer should include an assessment of the following: disc size, shape, symmetry, color, angle, vessels, and extrapapillary features such as the presence of a zone beta.

Glaucomatous Optic Nerve (Drance) Hemorrhages

Disc hemorrhages are an important prognostic indicator in the assessment and management of glaucoma and ocular hypertension. Glaucoma patients who develop disc hemorrhages are more likely to develop optic nerve

head/retinal nerve fiber layer damage and visual field loss sooner than patients who do not develop these hemorrhages.

Detection. Most disc hemorrhages occur in the inferior temporal quadrant of the optic nerve and are of relatively short duration (~1 to 3 months).The OHTS showed that reviewing retinal photographs was considerably more sensitive at detecting disc hemorrhages when compared with clinicians viewing the nerve directly with ophthal- moscopy—even though the optic nerve heads of these patients were examined ophthalmoscopically twice per year versus retinal photographs, which were reviewed only once in the same time frame.

Pathogenesis. The pathogenesis of Drance hemorrhages is incompletely understood.

Differential Diagnosis. Differential diagnoses of disc hemorrhages include posterior vitreous detachment, diabetic retinopathy, hypertensive retinopathy, hemorrhage resulting from optic disc drusen, ischemic optic neuropathy, leukemia, and peripapillary neovascular membrane.

Laser Imaging Devices. Currently available imaging techniques used for examining the retinal nerve fiber layer and/or the optic disc in glaucoma include confocal scanning laser ophthalmoscopy, optical coherence tomography, and scanning laser polarimetry. Each of these techniques uses different technologies and light sources

Table 34-1

Imaging Devices and Their Associated Technology

to characterize the distribution of retinal nerve fiber layer and/or optic disc topography (Table 34-1).

There are distinct advantages to this imaging technology. Optic nerve and retinal nerve fiber layer assessment, by even very experienced clinicians, has a level of subjectivity and agreement between experienced clinicians that is not ideal. Imaging technology compares acquired data with internal databases and assesses the degree of variability from this age-matched norm. This biostatistical analysis can serve as an important adjunct to the clinical assessment of the optic nerve. This analysis also demonstrates progression over time in ways that are more sensitive than clinical observation. At this time evidence does not preferentially support any one of the above structural tests for diagnosing glaucoma. Different imaging technologies may be complementary and detect different abnormal features in the same patients. This information supplements the assessment by the clinician who compares each patient’s findings with his or her clinical experience. It is ill advised to use any of these devices in the absence of sound clinical assessments and judgment.

New versions of imaging devices are available with higher resolution and more rapid acquisition time. Future improvements will include the incorporation of adaptive optics that, at present, can resolve retinal structures at the cellular level. This technology will continue to improve and play a more critical role in the management of glaucoma and eye disease in general.

The Relationship Between Structure and Function. The correlation between the results derived from these structural

devices and the functional (visual field) assessment of glaucoma (particularly early glaucoma) is weak. Recent evidence suggests that this may be due, in large part, to the test-retest variability associated with our current visual fields devices and other sources of noise in the acquisition of data. Several studies have also provided evidence that the time of onset of structural and functional defects, detectable using current techniques, is different. For instance, the OHTS and European Glaucoma Prevention Study both showed that, in many eyes, structural defects develop before functional defects (perhaps in areas of high retinal redundancy), whereas in a similar number of other eyes, functional defects develop first (perhaps a result of retinal ganglion cells becoming dysfunctional before dying). The simultaneous presentation of structural and corresponding functional defect in early glaucoma is much less common. In advanced glaucoma the correlation between structure and function is quite good (a patient whose optic nerve is cupped to the inferior temporal rim will likely have a corresponding superior nasal defect), although exceptions do occur. As glaucoma progresses to end stage, the utility of the structural assessment is limited. Quite often, these optic discs are cupped to the rim in most quadrants, and most of what remains are the papillomacular bundles. At this point, imaging devices have also reached the limits of their usefulness. As such, our ability to assess changes in the structural integrity of end-stage nerves is poor. Under these circumstances, the visual field assessment is still a useful tool because most of these patients have some functional visual ability. If the condition continues to progress, alternative forms of visual fields may be indicated (e.g., 10-2 test pattern, stimulus size V, Goldmann fields).

Clinical Pearls

•In many cases structural defects develop before functional defects.

•In many cases functional defects develop before structural defects.

Table 34-2

Visual-Function Specific Perimetric Tests

•The relationship between structural and functional loss measured with our current clinical technology is weak, especially in early glaucoma.

•Both the visual field and optic disc must be monitored with equal diligence (in all but end-stage glaucoma).

•Confocal scanning laser ophthalmoscopy, ocular coherence tomography, and scanning laser polarimetry seem to be similarly able to discriminate between healthy and glaucomatous eyes.

•Retinal photography is an important adjunctive tool in the assessment of the optic nerve and retinal nerve fiber layer.

•Disc (Drance) hemorrhages and retinal nerve fiber layer defects that are visible with retinal photography can be overlooked during clinical and/or laser ophthalmoscopy.

Functional Assessment (Visual Fields)

In many instances, functional (visual field) changes provide the first clinical evidence of glaucoma. The measurement and assessment of the visual field has seen several transformations over the past few decades. The conversion from Goldmann kinetic visual fields to static automated perimetry marked a significant milestone in the measure of retinal sensitivity in a clinic setting. No longer requiring the skilled perimetrist and having the theoretical advantage of a more objective measure of visual function, the automated visual field has become a conventional tool in most offices. Since then, these devices have increased in clinical utility with the addition of normative databases, built-in statistical analyses, and faster algorithms, aimed at assisting the practitioner in the diagnosis of glaucoma and the determination of whether a patient’s condition is stable or progressing. In addition, alternative visual field stimuli aimed at specific retinal pathways have been introduced with the hope of “reducing retinal redundancy,” thereby detecting functional changes in advance of conventional white-on-white (achromatic) perimetry (Table 34-2).

Achromatic (white-on-white) static automated perimetry (“standard” automated perimetry or conventional

automated perimetry) presents an achromatic incremental stimulus on an achromatic background. This testing strategy has become very familiar to most practitioners. Since its introduction, much has been learned about the human retina that, we now know, is divided into several distinct retinal ganglion cell pathways that project to specific layers in the lateral geniculate nucleus en route to the visual cortex and other locations.Achromatic stimuli are not tuned to any particular cell type. In fact, any of these pathways is capable of responding to a white-on- white stimulus. The clinical effects of this overlap, or “redundancy,” theoretically means that some percentage of most of these cell types must lose their function in a given location in the retina for a white-on-white stimulus to either not be seen or to require a brighter intensity to be seen. Requiring a brighter intensity to be seen is referred to as an increase in visual threshold (or a decrease in retinal sensitivity). Although there are advantages to a redundant system, this works to our disadvantage if we are trying to detect change in sensitivity as early as possible. If perimetric stimuli could be “tuned” to the frequencies of particular cell types, then this redundancy could be reduced. Theoretically, this reduction in redundancy could produce a perimetric test that was more sensitive to early change because there would be no, or minimum, responses from alternative pathways to stimulate. In addition, if one particular cell type was believed to be affected earlier in the disease process, then tuning stimuli to this cell type may enable us to measure changes in retinal ganglion cell sensitivity in a more timely manner—assuming that early diagnosis has some impact on the long-term outcome of our patients.

Recent evidence, however, does not support the notion that any retinal ganglion cell type is preferentially affected in early glaucoma, and perimetric stimuli tuned to specific types of ganglion cells are not necessarily more sensitive at distinguishing patients with early glaucoma or progressive glaucomatous optic neuropathy. It is often assumed that visual field stimuli tuned to specific ganglion cell pathways are more sensitive than SAP. However, when assessing the ability of these tests to identify glaucoma patients using the presence of glaucomatous optic neuropathy or progressive glaucomatous optic neuropathy as examined by expert observers, static automated perimetry performance has been shown to be equal to or slightly better than short wavelength automated perimetry and not significantly different from frequency doubling technology. Because, in general, no one test appears to be more sensitive at confirming glaucomatous optic neuropathy, perhaps a battery of functional tests that uses some or each of these test strategies may prove to be of greater benefit.

Where We Are Now. Static automated perimetry (white- on-white or conventional automated perimetry) has become the standard for functional testing of the visual field in the clinical setting.There are several manufacturers

of these devices. What distinguishes them from each other is the way in which the stimuli are presented and the data analyzed. There have been many attempts at striking a balance between reliable data and speed of acquisition. Contemporary algorithms are truly an improvement over the early versions, particularly when compared with full threshold data. Care should be taken in choosing appropriate algorithms for testing because some may have more variability between tests than others (e.g., Swedish Interactive Thresholding Algorithm [SITA] FAST version).

Variability. One serious drawback to static automated perimetry analysis is the variability of data within a given examination (short-term fluctuation) and between examinations (test–retest variability). It is well established that test–retest variability increases as a function of decreased retinal sensitivity even in the normal retina. Imagine the island of vision (Traquair’s island of vision) with its peak at the fovea and a relatively gradual slope in sensitivity until the retinal periphery where the slope decreases exponentially. Test–retest variability also increases exponentially in the periphery.This is one reason why testing beyond the central 30 degrees is seldom used in clinical practice. Simply put, it would be very difficult to distinguish any measured changes in sensitivity from the noise of the expected test–retest variability.

Decreased retinal sensitivity can also occur as a result of small pupils and media opacities (preretinal receptor factors) or from disease (e.g., glaucoma). Figure 34-2 shows test–retest variability as a function of sensitivity in the central 10 degrees in a group of patients with glaucoma. Notice that variability is less in areas of high sensitivity (e.g., ~32 dB) and very low sensitivity (~0 to 3 dB). In the former, highly sensitive areas likely remain highly

Mean Sensitivity (dB)

Figure 34-2 Test–retest variability as a function of retinal sensitivity. (From Wyatt HJ, Dul MW, Swanson WH.Variability of visual field measurements is correlated with the gradient of visual sensitivity.Vision Res 2007;47:925–936.)

682 CHAPTER 34 The Glaucomas

sensitive to a given stimulus from one test to the next. In the latter, areas of very low sensitivity likely do not detect a given stimulus no matter how often it is presented (the results are generally similar each time). In between these two extremes (especially between retinal sensitivities between 15 and 20 dB), test–retest variability can be quite large (±15 dB). Variability can also increase dramatically with small changes in fixation, particularly near the edge of a steep scotoma.

Test–Retest Variability in the OHTS and Clinical Practice.

Test–retest variability also proved to be a significant issue in the OHTS,where about 86% of visual fields that were consistent with glaucoma on initial testing were normal on retest. Following two consecutive glaucomatous visual field results, ~66% were subsequently read as normal on the third follow-up.These results are not uncommon in clinical practice (Figure 34-3) and speak to the need to establish a baseline before diagnosis of the disease or its severity.

Additionally, assessments of the stability of glaucoma rely on the quality of the baseline data.

Management of Test–Retest Variability. One way to manage test–retest variability is to increase the size or intensity of the stimulus presented. However, this modification would be at the expense of sensitivity to change. That is, it would take a significantly greater degree of retinal

dysfunction to produce a change in sensitivity, an untenable alternative for most clinicians who are interested in detecting change as soon as possible. This approach is used in some forms of perimetry (e.g., frequency doubling technology).

Another way to deal with variability is to gather more data. That is, repeat the visual fields on several occasions.This may also be untenable for some clinicians.

684 CHAPTER 34 The Glaucomas

However, it is, at present, the basis for most types of serial visual field analysis (e.g., progression analysis). It may take as many as five to eight visual fields to be able to statistically differentiate true change from the noise of test–retest variability, particularly in visual fields with scotomas. Figure 34-4 shows the gray scales of 5 years of visual fields.

Note how the depth of the scotoma appears to vary considerably from one field to the next. This degree of variability is not an uncommon finding in the measurement of visual fields in glaucoma. Clinical decisions regarding the stability or progression of glaucoma based on visual fields must be tempered with an appreciation and understanding of expected variability. In fact, it is difficult to distinguish between progression of glaucomatous

visual field loss and long-term variability unless several visual field tests are obtained over time.Thus, it is necessary to confirm changes to avoid false-positive progressive visual field loss.

The Glaucomatous Visual Field. By the OHTS criteria, a visual field is considered abnormal if the glaucoma hemifield test is outside of normal limits and/or the corrected pattern standard deviation is p < 5% on at least three consecutive reliable tests, with the abnormality in the same location.The patterns of glaucomatous visual fields are summarized in Box 34-3.

It may be more reasonable and consistent in clinical practice to reduce the number of confirmatory examinations