We thank our colleagues for commenting on our article relating local structure and function in the macula. With the ever-increasing speed and resolution of frequency-domain OCT as well as the accuracy of segmentation algorithms, it seems that we are on the brink of exciting developments in glaucoma, and this attention to the macular region is important. The points made in these comments are all interesting and, in general, we are in agreement. Commenter 2 raised questions about the utility of structure-function relations. We did not intend for this paper to address the question of whether combining structure and function data provides more diagnostic information. However, we have recently shown1 that each can be converted to probability plots and have provided examples suggesting that these combined probability plots will be more diagnostic. In fact, we are currently testing this hypothesis in a study designed with a larger population. Further, we certainly did not intend to imply that a very high correlation between structure and function was necessarily a good thing. For example, because of true individual differences, a structural test might be more sensitive for detecting glaucomatous damage in one patient and a functional test in another.2 On the other hand, it is worth noting that even if the two were perfectly correlated, obtaining both measures would nonetheless be useful as one would expect the measurement error involved in structural and functional tests to be independent, which is undoubtedly not true for two functional measures. In any case, a deeper understanding of the implication of structure vs. function relationships requires a model that takes both measurement error and individual differences into consideration.3
Finally, we agree with the suggestion in comment 3 regarding the need for a better, more detailed map relating the visual field to structural damage in the macula, especially as we move towards more elegant ways of combining structure and function information in glaucoma. The Shafi et al. paper, referenced in comment 1, makes it clear that prominent maps from the literature do not agree. We, as well as other groups, are currently tackling this problem from a variety of angles.