Use of refractometry for determination of psittacine plasma protein concentration

Previous studies have demonstrated both poor and good correlation of total protein concentrations in various avian species using refractometry and biuret methodologies. The purpose of the current study was to compare these 2 techniques of total protein determination using plasma samples from several psittacine species and to determine the effect of cholesterol and other solutes on refractometry results. Total protein concentration in heparinized plasma samples without visible lipemia was analyzed by refractometry and an automated biuret method on a dry reagent analyzer (Ortho 250). Cholesterol, glucose, and uric acid concentrations were measured using the same analyzer. Results were compared using Deming regression analysis, Bland-Altman bias plots, and Spearman's rank correlation. Correlation coefficients (r) for total protein results by refractometry and biuret methods were 0.49 in African grey parrots (n=28), 0.77 in Amazon parrots (20), 0.57 in cockatiels (20), 0.73 in cockatoos (36), 0.86 in conures (20), and 0.93 in macaws (38) (P <=.01). Cholesterol concentration, but not glucose or uric acid concentrations, was significantly correlated with total protein concentration obtained by refractometry in Amazon parrots, conures, and macaws (n=25 each, P <.05), and trended towards significance in African grey parrots and cockatoos (P=.06). Refractometry can be used to accurately measure total protein concentration in nonlipemic plasma samples from some psittacine species. Method and species-specific reference intervals should be used in the interpretation of total protein values.