Pre-Analytical Errors in Laboratory Testing of Canine and Feline Samples. Occurrence of Hemolysis, Lipemia, and Icterus

Pre-analytical errors are common issues in veterinary and human clinical pathology. The interference of hemolysis, lipemia, and icterus analyzing serum (S) or heparinized plasma (HP) samples affect results of several biochemistry parameters. A plan to investigate these pre-analytical errors in two veterinary settings was arranged. We retrospectively investigated the occurrence of pre-analytical errors in canine and feline S or HP samples in two university laboratories between years 2010–2014. We based the search looking at the following keywords within the labs database: lipemia/lipemic, hemolysis/hemolytic, and icterus/icteric entered by technicians at the analysis' time. The comparison between the dosage and visual estimation by the technician of cholesterol or triglycerides vs. lipemia and bilirubin vs. icterus was also evaluated. A prospective study was also performed to measure hemoglobin, cholesterol/triglycerides, and bilirubin content in 27 hemolytic, 20 lipemic, and 12 icteric samples, respectively. The effect of centrifugation (10', 3,000 rpm, 4°C) on albumin, total protein, alkaline phosphatase, and alanine aminotransferase activities, cholesterol, triglycerides, urea, and creatinine was also assessed in 20 HP-lipemic samples. In Pisa 551/10,905 (5.05%) canine and 147/1,981 (7.42%) feline S/HP samples with pre-analytical alterations were detected, while in Nantes 301/7,773 (4.44%) canine and 181/5,905 (3.55%) feline S/HP-samples with pre-analytical alterations were instead identified. Retrospective comparison between analyte dosages vs. visual estimation by technician in S/HP-samples was not statistically significant (regression coefficient ≤ 0.3, p > 0.05 for all parameters). Moreover, hemoglobin, cholesterol/triglycerides, and bilirubin concentrations investigated prospectively were not significantly related with the initial visual estimation made by the technician in hemolytic, lipemic, and icteric S/HP-samples respectively (R ≤ 0.3, p > 0.05 for all parameters). In 20 lipemic HP-samples, only cholesterol, triglycerides, and total protein values were significantly lower after the additional centrifugation (p < 0.05). Pre-analytical alterations were quite frequent, around 5% of canine and feline S/HP-samples. A greater number of S/HP-samples with pre-analytical alterations were recognized in Pisa lab compared to Nantes' lab. An additional centrifugation in lipemic HP samples could be appropriate to assess cholesterol, triglycerides, and total protein concentrations. As the visual pre-analytical estimation of the hemolytic, lipemic, and icteric samples was not significantly related with the real amount of hemoglobin, cholesterol/triglyceride, and bilirubin in the sample, a visual scale for inspection and estimation of these pre-analytical errors was arranged.