advertisement

---

Editors Selection IGR 10-4

Perimetry: Learning effects

Comment by Anders Heijl on:

22987 Is there evidence for continued learning over multiple years in perimetry?, Gardiner SK; Demirel S; Johnson CA, Optometry and Vision Science, 2008; 85: 1043-1048

Find related abstracts

---

Perimetry Learning effects Comment by Anders Heijl, Malmö, Sweden Gardiner et al. (152) studied longterm perimetric learning effects with SAP and SWAP in 80 patients with suspected or early glaucoma. Testing was done annually for 8 years. The outcome parameter was Mean Sensitivity (MS). As expected, the authors found a clear increase of MS (by 0.5 dB) between the first and the second test with SAP, but then no mean improvement until year 5, when the test program was switched from Full Threshold to SITA Standard, which explains this apparent 'late learning'. With SWAP, on the other hand, MS slowly increased until year 6. The authors conclude that SWAP seems to have prolonged learning effects. Smaller prolonged learning effects may also be present at least in some patients with SAP; certainly we have sometimes seen such effects in patients with glaucomatous field loss in the EMGT cohort. As pointed out by the authors

There is a need for untreated control groups in prospective studies evaluating any treatment modalities

the larger learning effects in SWAP are probably due to the higher variability of SWAP as compared to SAP ‐ SWAP is a more difficult test than SAP. What does all this mean, clinically and scientifically? I believe that learning effects after the second test mean very little in everyday clinical practice. But if one were to rely on measurements of MS or MD over time as a method to find early glaucomatous visual field loss, late learning effects could be quite disturbing. The SWAP effects are clearly considerably larger and could be more of a problem. But, both with SAP and with SWAP early diagnosis would of course be based more on detection of localized loss than of diffuse reduction of sensitivity and then any diffuse small-scale increase of sensitivity would not in any way jeopardize detection of early damage. If we instead consider the authors' findings in the context establishing rate-of-progression in manifest glaucoma with field loss, we can probably concluded that the effects of any late learning are probably negligible with SAP (after the 1^st year), since any such effects are probably about one log unit smaller than typical progression rates in established glaucoma ‐ while with SWAP learning effects may certainly be of clinical importance. For research, the findings reported in this paper are clearly important. This paper clearly demonstrates the need for untreated control groups in prospective studies evaluating any treatment modalities. Imagine if these patients had been treated with blueberry soup every morning for their suspect and tested with SWAP only ‐ and no control group ‐ that would have been just wonderful for blueberry soup producers.
NB. Blueberries may also offer significant protection against the development of glaucoma, because of its collagen-enhancing actions.

---

• ← Previous
• Next →

---