One of the first strategies was the Fastpac program (Carl Zeiss Meditec, Inc., Dublin, CA), in which the intensity of the light stimulus was gradually increased in three dB steps and the threshold was crossed only once. Although Fastpac reduced the test time by 36% compared with the Full-Threshold strategy, it also artificially decreased the severity of visual field defects and increased intratest variability.3
Recently introduced strategies such as the Swedish Interactive Threshold Algorithm (SITA; Carl Zeiss Meditec, Inc.) and the Tendency-Oriented Perimetry (TOP; OCTOPUS 1-2-3; INTERZEAG, Berne, Switzerland) aim to reduce test time while preserving the quality of information obtained with standard algorithms. SITA includes two programs, SITA Standard and SITA Fast, which were developed to replace the Full-Threshold and Fastpac strategies, respectively.
Simply put, SITA testing is quicker for patients (approximately 50% faster than testing with the Full-Threshold strategy), because this testing strategy adapts throughout the test in response to patients' performance.4-6 A significant limitation of the existing literature regarding SITA testing is that participating research subjects were already familiar with automated perimetry. As a result, the studies provide little information on the practical use of SITA testing in individuals who are inexperienced with perimetry, such as newly diagnosed glaucoma patients or glaucoma suspects. It is well established that information collected during the first perimetric examination is imprecise, however, and patients' learning curves characteristically involve a progressive improvement of automated perimetry's global sensitivity and a reduction of short-term fluctuation.7,8 Because we believe it is key to recognize the quality of information obtained during the first examination, we compared the SITA Standard and Full-Threshold strategies in normal individuals who had no previous experience with automated perimetry.9
METHODOLOGY
Eighty perimetrically inexperienced, normal individuals underwent computerized visual field examinations (30-2 Humphrey 750 system; Carl Zeiss Meditec, Inc., Dublin, CA) with both the Full-Threshold and SITA Standard strategies. We tested one eye of each subject in a random order.
First, we compared the results of all tests. Next, we evaluated patients undergoing SITA Standard as their first examination with those initially undergoing Full-Threshold examinations. Finally, we compared the results of the second test. To calculate specificity (meaning the percentage of normal examinations among a population known to be normal), we applied Anderson's criteria2 to identify abnormal examinations, which we characterized as (1) a cluster of three or more non-edge points on the pattern deviation probability map that deviated at P<5%, with one of these points deviating at P<1%; (2) a pattern standard deviation value occurring in less than 5% of normal, reliable fields (P <5%); (3) or a glaucoma hemifield test outside normal limits.
RESULTS
The number of significantly depressed points deviating at P <5%, P <2%, and P <1% on the pattern deviation probability maps was higher with the SITA Standard strategy (P <.05). The specificity was 50% for SITA Standard and 72.5% for Full-Threshold examinations (P <.01). When we compared first examinations alone, the number of significantly depressed points deviating at P <5%, P <2%, and P <1% on both the total and pattern deviation probability maps was higher with the SITA Standard strategy (P <.05). The specificity was 38.1% and 63.2% for the SITA Standard and Full-Threshold strategies, respectively (P=.04). A comparison of the second tests alone revealed no significant differences between either the number of depressed points or the specificities of both strategies (73.7% for SITA Standard and 71.4% for Full-Threshold, P=.98) (Figure 1).
Figure 1. A 30-2 Full-Threshold examination (left) and a 30-2 SITA test (right) in a patient undergoing a visual field examination for the first time. Note the higher number of suspicious points in the pattern deviation plot of the SITA examination.
ANALYSIS
Among Anderson's criteria, the glaucoma hemifield test performed better in terms of specificity in all SITA Standard tests. In fact, in all three analyses, there was no significant difference between the percentage of glaucoma hemifield test false positives obtained with the SITA Standard and Full-Threshold strategies. Furthermore, when we defined abnormality as the fulfillment of two of the three criteria, the difference in the number of false positives between the two methods was not statistically significant in all analyses.
The lower variability between individuals obtained with the SITA Standard strategy compared with the Full-Threshold strategy may explain our findings, because it resulted in a smaller range for normal age-adjusted values.10 Discreet changes of sensitivity in relation to the normal database are therefore more likely to be considered significantly depressed on SITA Standard probability maps compared with the Full-Threshold algorithm.
In the comparison of results obtained with the second visual field tests, the previously described differences regarding the number of depressed points on the pattern deviation probability maps, pattern standard deviation values, and specificity disappeared. These findings suggest that, after patients become familiar with the examination, the probability decreases of finding imprecise data, which may erroneously indicate the presence of visual field defects.
CONCLUSION
Our findings may have an impact on epidemiological studies. If investigators decide to define abnormality based on a single examination, we recommend using two of Anderson's three criteria to increase testing specificity, although this approach may decrease sensitivity. In a clinical setting, if physicians indeed must cautiously analyze the first visual field examination with the Full-Threshold strategy because it may provide inconsistent data, then that recommendation becomes stronger when the first examination uses the SITA Standard strategy.
The differences we have outlined tend to vanish during a second visual field examination. We do believe that SITA represents an important evolution in automated perimetry, one that allows patients to undergo a quick examination and provides consistently reliable information in perimetrically experienced individuals.
Rui B. Schimiti, MD, is Assistant Professor of Ophthalmology for the Department of Ophthalmology at the University of Campinas in Brazil. He does not hold a financial interest in any of the technologies and companies mentioned herein. Dr. Schimiti may be reached at +55 19 3788 7380; ruioft@sercomtel.com.br.
Vital P. Costa, MD, is Director of the Glaucoma Service for the Department of Ophthalmology at the University of Campinas and Associate Professor of Ophthalmology for the Department of Ophthalmology at University São Paulo in Brazil. He does not hold a financial interest in any of the technologies and companies mentioned herein. Dr. Costa may be reached at +55 11 3865 9630; vp.costa@uol.com.br.
1. Anderson DR. Perimetry: with and without automation. 2nd ed. St Louis: CV Mosby; 1987:484-489.
2. Anderson DR, Patella, VM. Automated static perimetry. St. Louis: CV Mosby; 1992:10-35.
3. Flanagan JG, Wild JM, Trope GM. Evaluation of FASTPAC, a new strategy for threshold estimation with the Humphrey Field Analyzer, in a glaucomatous population. Ophthalmology. 1993;100:949-954.
4. Bengtsson B, Heijl A, Olsson J. Evaluation of a new threshold visual field strategy, SITA, in normal subjects. Swedish Interactive Thresholding Algorithm. Acta Ophthalmol Scand. 1998;76:165-169.
5. Bengtsson B, Heijl A. Comparing significance and magnitude of glaucomatous visual field defects using the SITA and Full Threshold strategies. Acta Ophthalmol Scand. 1999;77:143-146.
6. Bengtsson B, Olsson J, Heijl A, Rootzén H. A new generation of algorithms for computerized threshold perimetry, SITA. Acta Ophthalmol Scand. 1997;75:368-375.
7. Heijl A, Lindgren G, Olsson J. The effect of perimetric experience in normal subjects. Arch Ophthalmol. 1989;107:81-86.
8. Autzen T, Work K. The effect of learning and age on short-term fluctuation and mean sensitivity of automated static perimetry. Acta Ophthalmol. 1990;68:327-330.
9. Schimiti RB, Avelino RR, Kara-Jose N, Costa VP. Full-threshold versus Swedish Interactive Threshold Algorithm (SITA) in normal individuals undergoing automated perimetry for the first time. Ophthalmology. 2002;109:2084-2092.
10. Bengtsson B, Heijl A. Inter-subject variability and normal limits of the SITA Standard, SITA Fast, and the Humphrey Full Threshold computerized perimetry strategies, SITA STATPAC. Acta Ophthalmol Scand. 1999;77:125-129.
