Sir,

The informative comments by Mr Roper-Hall and Dr Mutlukan are valued contributions to understanding the background to much modern psychophysical testing of the central visual field. The original quotes from Professor Amsler are very relevant to current practice in the use of Amsler grids and are sure to educate many contemporary ophthalmologists. Indeed, one cannot help but wonder why the current black grids on white paper were introduced, presumably as they would seem to be easier to print and thus might be more cost-effective in our predominantly state-run healthcare system in the United Kingdom.

We would emphasise, however, that our study was not to determine which type of chart is the best to use, nor the extent of visual field loss it detected, but rather an assessment of methodology that is normal current practice.1 We found that the British National Health Service most often uses the Chart No. 1 by Keeler: a black grid on a white background. In short, our study found this to be an unsatisfactory test and Professor Amsler's original comments may indeed partly explain this. However, we stress that Amsler charts should continue to be dispensed as they do detect a fair proportion of subretinal membranes (approximately 30% in our study using black on white charts).

The significance of this area is increasing considerably with PDT laser and other treatment modalities for age-related macular degeneration that require early reliable detection of subretinal neovascular membranes.

We are in agreement that methods for early detection of submacular neovascularisation need revision. A trial using a white grid on a black background is justified. Certainly, the use of colour field test cards is a physiologically sound suggestion to test macular function. Their feasibility for elderly patients including those living at home needs to be assessed, but this would likely not be a major obstacle, although the cost of colour charts might be a more significant challenge. The benefits colour charts might confer over the black-on-white Amsler grid suggest a need for their inclusion in any rigorous comparison of practically feasible tests of central field.