Colour change in cyanosis and the confusions of congenital colour vision deficient observers

Abstract

Visual recognition of cyanosis is an important clinical activity. While pulse oximetry is almost universal in the hospital environment, there are circumstances where it is not available or may be unreliable. Cyanosis recognition is affected by lighting colour. In addition, there is, mainly anecdotal, evidence that people with greater colour vision deficiencies (CVDs) have particular difficulty and there is no effective lighting strategy to assist in the observation. The change of blood colour with oxygenation has been shown to lie close to the direction of colour confusions made by congenital red-green dichromats. The important sites of observation are lips, nail beds and palm creases. 10 subjects who were known to be chronically hypoxaemic were recruited from the chronic respiratory program. Their blood oxygen saturation (SpO(2)) varied from 84% to 96% pre-exercise, and 61-84% post-exercise. Ten normal subjects were recruited whose SpO(2) was 99% or 100%. The spectral radiances of lips, nail beds and palm creases were measured using a telespectroradiometer and compared with the spectral radiances of a white tile of known spectral reflectances measured in the same location. This is a non-contact method of measurement, avoiding the blanching caused by pressure of contact methods. The spectral reflectances were calculated, and the chromaticities calculated for a Planckian radiator T = 4000K. Measurements on lips yielded the most consistent results. The colour changes pre- and post-exercise and compared with normal colour lie generally along a deutan confusion line. These results show the direction of the colour change and confirm the, previously anecdotal, difficulties in detecting cyanosis by observers with CVDs.

Recognition of simulated cyanosis by color-vision-normal and color-vision-deficient subjects

“…It has been shown that the change with oxygenation in the color of blood does follow quite closely the protan (in particular) and deutan confusion lines [8]. The major effect was that the chromaticity difference Δu 0 v 0 available to the protanope was shown to be 10% that of the normal.…”

“…Institutional ethics committees are reluctant to approve such studies. Measurements have been made of lip, nail bed, and palm crease color in vivo on chronically cyanosed subjects [8]. In this case there are some differences causes by individual differences in skin color, etc.…”

The specification of color limits in eye protection lenses for use when color‐contingent clinical observations are made

“…It was found the blood could be held at stable conditions for the color to be measured, unlike trying to measure in vivo. Subsequently, and with faster instrument, in vivo measurements were also possible 29 . The COI remained part of the AS/NZS standard in the 2018 revision.…”

“…26 It would be logical to assume that the irradiance in the region of 475 nm would have an important influence on the visual ability to detect jaundice but only one study, which uses an outdated metric, gives any numerical clue to limits to be placed on the spectral irradiance of the light source or colorimetric measures. 25 In addition, it is evident that the color in cyanosis changes almost entirely in colorfulness [27][28][29] while in jaundice, it changes primarily in hue and less so in colorfulness. 27 It is possible that one of these dimensions is more sensitive to changes in the light source or tinted eyewear.…”

The Eumelanin Human Skin Colour Scale: a proof‐of‐concept study

“…The visual impression of skin tones through the photoreceptors of the eye is dependent not only on lighting conditions, but also on cultural factors, and whether the observer is colour blind. Those without trichromatic vision may not be able to recognize the subtleties of a change in mood expressed by facial skin colour changes, sunburning of the skin, or cyanosis 12 . Observational bias can be overcome using quantitative measures of skin colour using reflectance, 13 colourimeters and spectrophotometers, 14 although generally there is a good correlation with visual observation 15 …”