Y. Kono et al., Comparative study between pointwise and ranked threshold distribution analyses of change in serial fields for short-wavelength automated perimetry, J GLAUCOMA, 9(6), 2000, pp. 419-427
Purpose: To assess the validity of ranked threshold distribution (RTD) anal
ysis for longitudinal evaluation of short-wavelength automated perimetry (S
WAP).
Methods: Eighty-five patients with glaucoma and with three or more SWAP res
ults were examined. An additional 20 patients with glaucoma and 3 stable st
andard automated visual fields and 3 stable SWAP fields were assessed to ca
lculate variability. An RTD curve was created. For RTD analysis and pointwi
se analysis, values that decreased more than the SWAP variability were cons
idered to have worsened. Using pointwise analysis, four groups were defined
: no change (without any 2 contiguous points decreasing in threshold); foca
l change (at least 2 contiguous points with decreased thresholds in either
hemifield); diffuse change (more than 50% of the entire 52 locations with s
ignificant decrease in thresholds); and focal and diffuse change (less than
50%, but crossing both hemifields).
Results: Using pointwise analysis, 32 (37.6%) patients did not change; 23 (
27.1%) patients had focal worsening; 8 (9.4%) patients had diffuse worsenin
g; and 22 (25.9%) patients had focal and diffuse worsening. Significant dif
ferences were found among the groups in the number of deteriorating locatio
ns (P < 0.0001, analysis of variance). The ranges of deteriorating location
s identified by RTD analysis for focal change and diffuse change did not ov
erlap. A significant correlation was found between the numbers of locations
with decreased threshold by both analyzes (r(2) = 0.83, P < 0.0001).
Conclusion: For evaluation of possible SWAP progression, RTD analysis is cl
osely related with pointwise analysis and is useful for visualizing a diffu
se component of change in the field.