Measuring room illuminance

Abstract

In the article by Wickremasinghe et al.,1 we are concerned by Figure 3 (mean pupil diameter [PD] versus illumination in lux), which would predict a mean PD at 1 lux of approximately 4.5 mm in a study population with a mean age of 38 years. Our work2 shows a mean dark-adapted PD (DAPD) of approximately 7.0 mm in a similar population. A discrepancy of 2.5 mm requires an explanation. Reading the Methods section with care, we noted that there was no preadaptation step, which we have found important in accelerating low-light adaptation once the patient enters the testing room (unpublished data); the duration of subsequent low-light adaptation was also variable. However, our greatest concern was with the measurement of ambient illumination. The investigators used the Sixtomat digital photographic exposure meter (Gossen AG) to allegedly adjust the ambient illumination in their testing room to 0.08 lux. When we perused a digital version of the Sixtomat's technical data sheet and user's manual (English translation) from Gossen's Web page, we did not see a clear indication that this device could function as an illuminance meter with a readout in lux. We contacted Gossen AG and received the following reply, quoted without modification, via e-mail from Gudrun Zeller, Export Area Manager: “Please note that the Sixtomat Digital is a light meter for photography, for ambient light and shows EV, f-stops, exp. times, etc. There is no indication of lux since this meter is not designed to measure illuminance but to assist in making perfect photos.” Our own research has been challenged on the grounds that our light measurement system was both imprecise and inaccurate (Daniel Taylor, Procyon, UK, personal communication). We used the AEMC CA813 illuminance meter (AEMC Instruments), which is calibrated to the spectral sensitivity of the human eye. This meter has a display precision of 0.01 lux, but for purposes of honest realism, we adjusted the ambient illumination to 1 lux (not 1.0 or 1.00 lux). At very low light levels, slight changes in reflected light caused by movement of the examiner or patient will cause variability in the luxmeter reading. When read to 1 decimal place, our “1 lux” ranged between 0.8 and 1.4 lux; this range corresponds with a “very-low-mesopic” level of retinal photoreceptor adaptation in which the rod system is fully active but some color information is retained because there is sufficient light to stimulate cones with low frequency. Tiny differences in low-level illumination (eg, 0.04 lux versus 0.4 lux) are of no practical importance under ordinary low-light activities of daily living. We agree with Wickremasinghe et al.1 that evaluation of the DAPD is a crucial portion of the preoperative refractive surgery workup and hope that future studies will address all conditions needed for accurate DAPD measurement. Regrettably, we see no way in which their study can be salvaged and we feel that clinicians should not use the reported findings as the basis for surgical decision making.