Sir,

We like to congratulate Dr Okada and associates for their work on the microperimeter (MP 1). With the commercial unavailability of the SLO microperimter, MP 1 is the only alternative to study macular functions including light sensitivity threshold, fixation pattern, and stability.1, 2, 3 Our initial study with the MP 1 on eyes with macular pathology revealed similar correlation of mean retinal sensitivity and visual acuity.2 In another study, we found that the mean sensitivity of the macular area was approximately 18 dB in normal subject, which reduced with increasing age (unpublished data). 18 dB is higher than median sensitivity (15 dB) reported by the authors.

We would like to make certain comments regarding this study. In the methods section the testing conditions have not been described. This may affect the retinal sensitivity measured by the MP 1. If the test room is lighted, the retinal sensitivity measured could be lower than when measured in the darkroom with less interference of surrounding light during the test. The demographics of the patients have not been mentioned, especially the age. It is well known that the retinal sensitivity reduces with age in the normal subjects both by the SLO microperimeter and conventional perimetry. Whether the control normal subjects were age matched or not would affect the results and interpretation. The other reason the retinal sensitivity could be lower in this study compared to our data is the consequence of learning effect. Subjects tend to do better in subsequent field tests than the initial one due to the learning effect. In our prospective study, we allowed every patient experience with machine in form of test stimuli before starting the test and we would only start the real test once the patient feels comfortable with the whole procedure. We feel lack of familiarity with the machine will affect the mean retinal sensitivity. The authors have not mentioned the initial level of sensitivity they used as the MP 1 allows the examiner to select this setting before test is initiated. This is important as in patients with diabetic macular oedema with mean sensitivity of 2 dB. If the test was started at 16 dB, then it would take a longer time to complete the test. The prolonged time could result in decreased patient cooperation parameters such as false-positives or fixation stability and these parameters have not been reported in this study. This data would be helpful in interpreting the reliability of the results.

The 4–2 strategy is faster but we believe when measuring the macular sensitivity 4-2-1 strategy is superior. We did not understand the rationale behind using the 12° cross for fixation. It would have helped significantly if the authors compared the retinal thickness at each quadrant surrounding the fovea and correlated the retinal sensitivity to thickness both in diabetic macular oedema and normal eyes.

We read this paper with great interest and would like to once again congratulate the authors on their important work in establishing anatomic and functional correlation in the diabetic macular oedema eyes.