I read with great interest the article by Jang and associates regarding the effect of yellow-tinted intraocular lenses (IOLs) on short-wavelength automated perimetry (SWAP). The study showed that when SWAP was performed, assuming normal structures in both eyes, mean deviation (MD) and pattern standard deviation (PSD) of eyes implanted with yellow-tinted IOLs were significantly different from those implanted with nontinted IOLs, so as the glaucoma hemifield test. These differences, however, were not evident with standard automated perimetry (SAP). The authors thus concluded that yellow IOLs may affect the results of SWAP.
It was noted that patients were neither tested at the identical area of their visual field, nor did they use the same thresholding technique between tests. In this study, SAP was performed with the 30-2 Swedish interactive threshold algorithm (SITA) standard strategy, whereas SWAP was performed using the 24-2 full-threshold (FT) strategy. The ability of Jang and associates, using SWAP with the full-threshold strategy, to detect significant functional differences between eyes might be the result of its intrinsic high long-term variability, that is, the test–retest variation over time. Kwon and associates noted that the test–retest variation was up to 4.07 dB in normal subjects, whereas the MD and PSD differences between eyes in Jang and associates’ study were only 1.61 and 0.48 dB, respectively. It is also debatable whether these small differences could translate to clinical significance. The authors might have considered highlighting the intertest difference of MD and PSD results between SWAP tests, so that the readers might have had a better interpretation of the conclusions. Because the data were obtained at least twice for each type of visual field, the authors might have also elaborated on the proportion of patients having different glaucoma hemifield test scores between tests, especially with SWAP, and how these variations were handled in the analysis.
Sensitivity thresholds usually are lower in SWAP, but even so, the ranges of MD and PSD generated from SAP were quite large for structurally normal eyes (−18.45 to 1.10 dB and 1.0 to 9.15 dB, respectively, combining both groups), reflecting a heterogeneous group of patients. It is uncertain whether the low MD and high PSD values reflect that at least some of the patients had some form of ocular disease or that these results can be attributed to poor performance. For these eyes, testing with the full-threshold strategy may not be the best option, because it is limited by a narrow dynamic range. However, SWAP with the Swedish interactive threshold algorithm may be a better alternative: this strategy would have wider dynamic range, would reduce test duration by 65%, and would result in higher mean sensitivities in normal eyes and lower intersubject variability.
Finally, Eisner and associates determined that patients’ adaptation to the yellow background, via pupil size and retinal light absorptions mechanisms, contributed at least 40% of the variance in the MD differences between SWAP and SAP. Because cataract formation is age related, in future studies on the effect of blue-blocking IOLs on SWAP, it may be inevitable to include older patients. However, when performing SWAP, we need to take into account miosis and age-related degenerative changes commonly encountered in the elderly population, which are potential confounders.