Abstract
Determination of the urinary protein-to-creatinine ratio (UPC) is an important tool in the quantification of proteinuria in animals. However, the result may be affected by the different methods of determining the urinary protein concentration. The aim of this study was to compare the turbidimetric method using benzethonium chloride and the colorimetric method using pyrogallol red in the measurement of the urinary protein concentration in dogs, cats, guinea pigs and horses. A total of 464, 192, 216 and 119 urine samples from dogs, cats, guinea pigs and horses were examined in the study, respectively. The group consisted of animals of both sexes and different ages, and, in the dogs and cats, it included both healthy animals and those with various health problems. In the group of horses and guinea pigs, only clinically healthy animals were included. A total of 347, 185, 103 and 100 samples from the dogs, cats, guinea pigs and horses were used in the statistical analysis; the other values were excluded as they were below the detection limit. According to the Passing-Bablok analysis, there was a significant constant and proportional difference in the horses. In the dogs, cats and guinea pigs, there was a significant constant difference, but no proportional difference. The Bland-Altman method showed significant bias between the two methods in the horses and cats, but not in the dogs and guinea pigs. In the dogs and cats, the agreement between the two methods was tested and expressed as Cohen's kappa (κ). In the cats, it was almost perfect for the proteinuric samples (κ = 0.823 3) and significant for the non-proteinuric samples (κ = 0.804 9). In the dogs, the agreement was significant for the non-proteinuric samples (κ = 0.621 5) and only moderate for the proteinuric samples (κ = 0.527 5). The influence of the method used to determine the urinary protein concentration should be taken into account when evaluating the UPC. Repeated examinations in one patient should be performed with the same method.