We thank Marta Gonzalez-Hernandez and Manuel Gonzalez de la Rosa for their comments on our article “Comparative Resilience of Clinical Perimetric Tests to Induced Levels of Intraocular Straylight.”
1 We do not agree with their statement that the results presented were “obvious” and “well known.” Before this report, there had been one article on the resilience of the multilocation Moorfields motion displacement test (MDT) to additional straylight and none on the resilience of Pulsar perimetry. We agree that there are several articles on the effect of straylight on standard automated perimetry (SAP) thresholds; however, none provide a running order on the relative resilience of each perimeter to increasing levels of straylight.
We would like to take this opportunity to elucidate further the issue raised by Gonzalez-Hernandez et al. regarding “blurring.” This does not account for the majority of visual changes observed with additional straylight. In fact, there was no measurable loss in visual acuity (VA) with any subjects across the range of fog filters—that is with increasing straylight, as outlined in our article. All subjects had greater than 6/6 Snellen VA with every test, although we concur there may have been small but immeasurable reductions in VA.
We agree that the topographic slope of Pulsar is pronounced. Preceding the sentence quoted by Gonzales-Hernandez et al. we stated, “on examining the differences between central and peripheral locations with SAP and PP from this study, it was found that peripheral locations were more affected by IOS (P = 0.05).” The following sentence omits the words “the difference between.” Thank you for pointing this out, and we apologize for this typographic error in the manuscript. However, there was no deliberate misquote, which is clear when put in the context of the previous sentence.
The reduction in sensitivity described in this article with the Pulsar perimeter is far greater than that expected from aging alone. We agree that accurate detection of retinal dysfunction related to age is essential with any perimeter; this is precisely what prompted our study. However, to detect retinal dysfunction accurately we must be able to minimize confounders. Aging of the lens and the resultant increase in straylight can have a large effect on threshold estimates, and this should be separable from aging of the retina or disease progression.
The overall theme of the article was not to point out that the Pulsar was affected by straylight, but rather to examine the relative effect of additional straylight on threshold measures for each perimeter (including Pulsar). Detecting progression is in part a signal-to-noise problem and, therefore, will be confounded by the noise created by increases in straylight. However, since this study did not investigate the performance of each perimeter in detecting progression, we do not agree that this was a suitable place for starting the larger discussion on progression detection using different stimuli types.
Robustness to the effects of straylight does not guarantee clinical utility (or a good signal-to-noise ratio), but it clearly is a positive factor, as there is less noise in the system from the outset. All perimeters included in this study used a simple yes/no response paradigm, and were designed to be commercially available tools for use in the clinical context. Flicker perimetry as indicated by the authors has not proved acceptable in the clinical context and we feel a discussion on temporal phenomena as suggested would distract from the focus of the study.