Clinical investigations Graefe's Archive [Or Clinical and Expefimenlal Ophthalmology © Springer-Verlag 1991 Graefe's Arch Clin Exp Ophthaimol (1991) 229:501-504 The learning and fatigue effect in automated perimetry Gianni Marra and Josef Flammer University Eye Clinic Basel, Mittlere Strasse 91, CH-4056 Basel, Switzerland Received November 5, 1990 / Accepted March 1, 1991 Abstract. A visual field test was performed on 100 volun- teers to study learning or fatigue effects during one ses- sion. The test was carried out with the help of the Octo- pus program Jl, which measures the threshold 12 times at 3 test locations. In the majority of cases the sensitivity was quite stable during the session. We noted no differ- ence between trained and untrained subjects or between normal and pathological eyes. However, patients with refractive errors, especially myopes, revealed a larger learning effect than did emmetropes. Introduction Perimetry is a psychophysical test procedure in which patients must be able to cooperate by exercising a high degree of concentration. In manual perimetry the experi- ence of the perimetrist has a major influence on the outcome of the examination. Whereas it has been possi- ble to minimize this influence through automation, this psychophysical test continues to rely on the cooperation Offprint requests to: J. Flammer of the patients. Although the latter may be perfectly willing to cooperate, they may not understand the test properly or may not be able to concentrate for an ex- tended period; some patients get tired during the test. Even when subjects understand the test properly and try to cooperate optimally, the outcome of a second test is often better than that of the initial one. Such a "learning effect" can often be observed clinically. Al- though the term learning effect is widely used in the literature, we do not know whether the patient really learns or exactly what he is learning. Is a physiological phenomenon of the visual system involved or does a psychological feature of the patients influence their de- cision as to which answer to give, seen or not seen? Although it is well known that the second examination can yield better results than the first one, we again turned our attention to this learning effect. We wanted to know whether a time trend can be observed during a perimetric session lasting a few minutes. Furthermore, we tested the question as to whether factors such as age, among others, have an influence on this trend. The literature on this topic is not entirely conclusive, being to some extent contradictory (Table 1). Some au- thors, such as Greve [5] and Parrish et al. [11], have Table 1. Summary of the findings in the literature concerning the learning effect in manual and automated perimetry Authors Learning effect Perimeter used Subjects Greve [5] Yes Goldman Glaucomatous Parrish et al. [11] Yes Perimetron Normal Rabineau et al. [12] Yes Octopus Normal Heijl et al. [7] Yes Humphrey Normal Lehmann and Faggioni [9] Yes Octopus Glaucomatous Wild et al. [14] Yes Humphrey Glaucoma suspects Katz and Sommer [8] No Humphrey Healthy Gloor et al. [3] No Octopus Glaucomatous Werner et al. [13] Only for SF Octopus Glaucomatous Gramer et al. [4] Only for SF Octopus Glaucomatous Aulhorn and Harms [1] Only in some cases Tfibinger Normal Marra et al. [10] Only in some cases Perikon Normal Gloor et al. [2] Only in some cases Octopus Glaucomatous