Zeiss HUMPHREY 720i, User Manual

Page: 131 / 328

STATPAC ANALYSIS & PRINTING RESULTS 7-15
REV. B 02/03 PN 51680-1 HUMPHREY FIELD ANALYZER
II -
i
series
The glaucoma change The Glaucoma Change Probability Analysis (GCP) is designed to facilitate interpretation of
probability analysis Central 30-2 and 24-2 threshold test results in patients with suspect or manifest glaucoma. It
printout is intended to allow maximum use of available test results. This analysis is particularly
useful in determining change over time, that is, in separating random variation from true
change. The GCP is only available when the Full Threshold strategy is used.
The Glaucoma Change Probability Analysis works from baseline data for the individual patient
to create change probability maps and to calculate significance limits for measured changes in
mean deviation. Because the global indices (MD, SF, PSD, and CPSD) are not necessarily
sensitive to important localized changes, the Glaucoma Change Probability Analysis offers
point-by-point significance limits. This allows analysis of smaller areas of the visible field defect
and enables early detection of change.
The Glaucoma Change Probability plots identify those locations in the visual field which have
changed by more than what would be expected simply due to normal variability. The signifi-
cance limits for this analysis were obtained by testing a large group of glaucoma patients four
times in the course of a month. Normal variability in these patients was found to depend on the
depth of the original defect at baseline, the location in the visual field, and the overall Mean
Deviation of the visual field. Points changing from baseline by more than the empirical
significance limits are highlighted with small triangles.
In general, the Glaucoma Change Probability Analysis will use the average of the first two
selected tests as a baseline and all subsequent tests as follow-up. There are two exceptions:
1. If only two tests are selected, the first will be used as the baseline and the second as
follow-up.
2. If the mean deviation of the first test falls significantly below the regression line of
those of the other tests (p < 5%), and five or more tests are to be analyzed, STATPAC
will discard the first test, use the second and third tests to calculate the patient’s
baseline, and analyze the subsequent tests as follow-up tests.
Note: Baseline tests should be representative of the actual baseline status of the patient.
Baselines established from tests in which the patient was obviously inattentive,
inexperienced, or too eager to press the response button can lead to false positive or false
negative conclusions upon follow-up. Create a new pair of baseline tests if significant
change has occurred (cataract surgery, for example).
The Glaucoma Change Probability Analysis was designed to allow comparison of a series of
follow-up tests with the baseline findings in order to detect and confirm changes in the visual
field. Prudence requires that changes detected in one follow-up test be confirmed in at least
one additional test before medical therapy is significantly changed or surgery is ordered.
The glaucoma change probability printout (Figure 7.7 and 7.8) includes the patient informa-
tion that appears on other STATPAC printouts. Two data presentations (graytone and total
deviation plot) for the baseline tests are printed on the upper left section of the printout.
A plot of the Mean Deviation (MD) for each test plus the linear regression analysis of mean
deviation, which is discussed below, occupy the upper right section of the printout.
The first
column on the left of the printout contains the graytone presentation of test results. The total
deviation plot appears in the second column. These are the only two data presentations given
7.htp 3/14/03, 3:17 PM 15