/ /

  • linkedin
  • Increase Font
  • Sharebar

    Glaucoma: Visual field interpretation for the busy practitioner

    A step-by-step guide breaks down visual field data for quicker detection of glaucomatous defects

    Danica J. Marrelli
    Houston—Automated perimetry plays an important role in both the diagnosis and the follow-up care of patients with glaucoma. Structural changes to the optic nerve or retinal nerve fiber layer are usually considered to occur before functional changes (perimetric loss). In some patients, however, visual field damage may be the first clinical sign of glaucoma.

    The amount of data printed on a standard visual field printout can seem overwhelming to the busy practitioner, but a systematic way of interpreting the test will make things quicker and easier. Three simple steps can speed things along and allow for careful but rapid detection of a glaucomatous visual field defect. These include an assessment of reliability, searching for glaucomatous defects, and examining the Glaucoma Hemifield Test (GHT) and global indices.

    STEP 1: Assess reliability.

    Diagnostic and management decisions should not be made on the basis of unreliable data. The three measures of reliability are fixation losses, false negatives, and false positives.

    • Fixation losses of 20% or higher may indicate an unreliable field, but may be due to incorrect plotting of the blind spot. If the technician observes good fixation throughout the visual field examination, a message of "high fixation losses" can be disregarded.
    • False negatives occur when a patient does not respond to a stimulus that should be visible for that patient. Because false negatives increase with worsening visual field loss, a high number of false positives in a glaucomatous field do not necessarily represent unreliable data.
    • False positives, on the other hand, are always an indicator of an unreliable test. They occur when the patient responds when a response was not expected. False positives make the visual field look better (more sensitive) than it actually is and may mask shallow depressions. Patients with high false positives also may have a "white" grayscale and/or abnormally high threshold values, especially in the peripheral parts of the field. An examination with a false positive rate of 15% or higher should be considered unreliable. It is important to recognize that it may take two or three visual field examinations for the patient to produce a reliable and valid visual field test.

    STEP 2: Look for glaucomatous defects.

    Once the field has been deemed reliable, it should be evaluated for glaucomatous defects. The grayscale should not be used for decision-making in glaucoma, but it is helpful to draw attention to areas that need further evaluation using number and probability plots.

    Deviation plots and their corresponding statistical probability plots demand the most attention. These plots identify areas of the field that are abnormal compared with an age-matched normal population for each point tested. The information is given both in decibels and in statistical probability values.

    In the probability plots, increasingly dark squares represent increasingly significant deviations from normal. The pattern deviation plot gives the same type of information as the total deviation plot, after the visual field has been adjusted for any overall depression or elevation. This plot is helpful in patients who may have a combined overall depression (from media opacity, for example) as well as localized loss from glaucoma.

    When evaluating the deviation plots, the clinician should look for clusters of abnormal points in locations typical for glaucomatous loss (nasal step, arcuate defect, paracentral scotoma). A single statistically significant point may not be clinically significant, but a group of connected points that all reach statistical significance is unlikely to be normal. In fact, a cluster of three or more connected points on the same side of the horizontal meridian that all reach statistical significance, with at least one of the points reaching the p < 1% significance level, is diagnostic of a glaucomatous visual field if the defect is repeatable.

    STEP 3: Look at the GHT and global indices.

    The GHT is based on the fact that glaucoma damages the superior and inferior fields asymmetrically. The GHT compares mirror-image clusters of points in the superior and inferior fields, and it alerts the clinician when significant differences are found between the two hemifields. A GHT message of "outside normal limits" on two occasions with corresponding optic nerve or retinal nerve fiber layer loss is very strong evidence that glaucomatous visual field loss is present.

    The two global indices used in SITA testing are mean deviation (MD) and pattern standard deviation (PSD). The MD is a measure of the average deviation between the patient's sensitivity and that of age-matched normals. This measurement gives a sense of the overall "height" of the hill of vision and is not specific or sensitive for glaucomatous damage. MD is affected by media, uncorrected refractive error, and small pupils, and it may not reach statistical significance from a small or shallow localized glaucomatous defect.

    PSD is a measure that reflects the shape or "smoothness" of the hill of vision and is much more reflective of localized loss, such as that found in glaucoma. It is important to realize, however, that other disease processes can cause localized loss and may result in an elevated pattern standard deviation. A PSD that reaches the 5% statistical significance level on multiple examinations, in the presence of other suspicious clinical findings, is very strong evidence that a glaucomatous visual field defect is present.

    Once a visual field has been determined to be abnormal, the field loss should be classified as mild, moderate, or severe, based on the size and depth of the defect and whether or not central vision is affected. From this point forward, the clinician's job is to evaluate the field for changes (progression). This is a much more challenging skill, which will be addressed in a future article. Fortunately, new software can now make this task easier.

    Danica J. Marrelli, OD, FAAO, is clinical associate professor at the University of Houston. She received honoraria from Carl Zeiss Meditec for speaking engagements and advisor meetings. She can be reached at 713/743-1945 or

    Optometry Times A/V