RESUMO
OBJECTIVE: To determine the relationship between arterial blood colour [as defined by the International Commission on Illumination (CIE) L∗a∗b∗ colour space] and haemoglobin oxygen saturation [functional saturation (SaO2) and fractional saturation (FO2Hb)], and if arterial blood colour can be used to predict arterial haemoglobin oxygen saturation. STUDY DESIGN: Descriptive study as an adjunct to two prospective randomized crossover studies. ANIMALS: A group of 10 wild caught adult female impala (Aepyceros melampus) weighing 34.1 ± 5.2 kg (mean ± standard deviation). METHODS: Impala were immobilized with potent opioids (0.09 mg kg-1 of etorphine or thiafentanil). A total of 163 arterial blood samples were collected anaerobically into heparinized syringes from arterial cannulae and analysed immediately using spectrocolourimetry and co-oximetry. Data were analysed by modelling the relationship between predicted arterial blood colour CIE L∗a∗b∗ components and SaO2 and FO2Hb. The models were then used to predict values for L∗, a∗ and b∗ to produce a colour palette for the range of SaO2 and FO2Hb used. The modified version of the Farnsworth-Munsell hue test was used to assess the subjective ordering of the resulting colour palette by 20 observers. RESULTS: The second-order polynomial (quadratic) model produced the best fit for all three arterial blood colour CIE L∗a∗b∗ components for both SaO2 and FO2Hb. The regression models were used to generate predicted arterial blood colour CIE L∗a∗b∗ components for the midpoint of each decile over a range of SaO2 and FO2Hb percentages (15% to 95%). The resulting colour palettes were correctly ordered by all observers in the SaO2 range of 45-95% saturation. CONCLUSIONS AND CLINICAL RELEVANCE: An association between arterial blood colour (as defined by CIE L∗a∗b∗ components) and SaO2 and FO2Hb exists, and arterial blood colour can be used to give a clinically useful estimate of arterial haemoglobin oxygen saturation in impala.