Limited overlap in FDT, SAP may identify different subgroups

September 15, 2004

Only a small proportion of patients were classified as having progression using both techniques.

Dr. Haymes, a postdoctoral fellow in the department of ophthalmology, Dalhousie University, Halifax, Nova Scotia, presented results of a study in which 71 patients underwent longitudinal follow-up with both FDT (full-threshold C-20 program, Humphrey, Carl Zeiss Meditec) and SAP (full-threshold 24-2 program, Humphrey Field Analyzer, Carl Zeiss Meditec).

Over a period of 3.5 years, one eye of each patient was studied with both tests at 6-month intervals. Two techniques were used to identify visual field change:

2) Linear regression analysis based on mean threshold

For each of those methods, progression was defined using multiple criteria, but ensuring that for each, the number of locations required to be flagged with FDT and with SAP was approximately equal on a percentage basis.

The results showed only a small proportion of patients were classified as having progression using both techniques. For the majority of GCP criteria, more patients showed progression with FDT than SAP, whereas the progression rates were higher with SAP using linear regression analysis.

"While FDT has been shown useful for glaucoma diagnosis, its potential to be used for progression monitoring has not been examined. We were interested in finding out if it could be used longitudinally and if it might be able to detect glaucoma progression earlier than SAP," Dr. Haymes said.

"Our results provide new information in showing that FDT can detect glaucomatous progression. However, in comparing the two techniques, it seems that in the absence of any gold standards, users can find a benefit for either FDT or SAP depending on how the data are analyzed."

In designing the study, the researchers decided to investigate the performance of the two perimetry tests using GCP and linear regression analysis because both analysis methods are fairly well accepted in the literature, and at the present time, there seems to be no preference for one versus the other. The lack of any gold standard was similarly the rationale for using various criteria to define progression.

In the GCP analysis and using the least strict criterion for defining progression-one test location to progress on FDT and two on SAP in two of three consecutive tests-32 (45%) of the 71 patients were classified as having progression with each technique. However, there was limited overlap because only 16 showed progression on both tests.

As the criteria for progression were made stricter by either increasing the number of test locations that were required to be detected as progressing or the number of exams required to confirm progression, a higher proportion of patients were classified as having progression by FDT compared with SAP. In total, 27 criteria were applied, and that pattern of a higher progression rate with FDT was noted in 15 of the analyses.

The technique of linear regression analysis was applied to data from 65 eyes in three ways looking at different subdivisions of the visual field. Irrespective of how the visual field was subdivided, a higher proportion of patients showed progression with SAP versus FDT.

As an example, in applying linear regression analysis to the mean of all test locations in each quadrant, 31% of patients were classified as showing progression on FDT and 35% on SAP. Using that criterion, only 10 patients (15%) had progression on both tests.

"We believe the limited overlap between these two techniques may indicate they are identifying different subgroups of glaucoma patients whose glaucoma is progressing," Dr. Haymes said.