Validation of 2 point-of-care meters for measuring triglycerides in chickens using whole blood and plasma

Abstract

Disorders of the avian reproductive tract are common, yet monitoring their resolution presents a diagnostic dilemma. Reproductive hormones such as luteinizing hormone or estrogen are the best reflection of reproductive status, but the required sample volumes and lack of reference intervals limit their clinical utility. An alternative analyte is blood triglyceride, the concentration of which rises markedly during sustained estrogen release from the ovary. Portable meters for measuring human blood triglyceride concentration offer the advantage of using minimal sample volumes, but these have not been validated for use in birds. We assessed the precision and accuracy of 2 portable meters for measuring blood triglyceride concentration in pooled whole blood and plasma from chickens (n = 42), and performed method comparison using a reference analyzer and determined total error. Within-run repeatability was fair-to-excellent using whole blood and plasma (range: 2.5–11.5%), and between-run repeatability using plasma was similar (3.1–12.2%). The meters performed well in recovery and dilution studies in which almost all readings fell within the preset requirement of 75–125%. Correlations between each meter, using whole blood and plasma, and the reference analyzer, using plasma only, were high to very high (0.86–0.98). Bias determined by Bland–Altman analysis was similar between whole blood and plasma for each meter, yet markedly different between the meters. The calculated total observed error was consequently within our pre-set total allowable error of 25% for one meter but not the other, indicating the requirement for a meter-specific reference interval.