Comparison of selected serum biochemistry measurements between the Nova Prime Plus VET, Nova pHOx Ultra, and Beckman Coulter AU680 analyzers in dogs. (20th September 2021)

• Record Type: Journal Article
• Authors: Nelson, Courtney; Constable, Peter D.; Connolly, Sara L.
• Is Part Of: Veterinary clinical pathology. Volume 50:Number 3(2021)
• Page Start: 327
• Page End: 341
• Publication Date: 2021-09-20
• Subjects: blood gas analyzers -- canine -- clinical chemistry -- method comparison; Veterinary pathology -- Periodicals; 636.089607
• DOI: 10.1111/vcp.13010
• Languages: English
• ISSNs: 0275-6382

Abstract

Abstract: Background: Blood gas chemistry analyzers typically produce results faster and use smaller sample volumes than reference chemistry analyzers. However, results may not be comparable between blood gas chemistry analyzers and reference chemistry analyzers or between different models of blood gas chemistry analyzers. This could suggest the use of separate reference intervals and, thus, has implications when making clinical decisions. Objective: We aimed to perform method comparison studies to evaluate selected canine serum biochemical values obtained using the Nova Stat Profile Prime Plus VET (Prime Plus VET), Stat Profile Nova pHOx Ultra (Ultra), and Beckman Coulter AU680 (Beckman) analyzers. We hypothesized that the three analyzers would be identical within inherent imprecision. Methods: Jugular venous blood samples were collected from 103 endurance‐trained sled dogs, and serum was harvested and stored for analysis. Results for serum chloride, potassium, sodium, creatinine, and urea nitrogen concentrations obtained from the Prime Plus VET and Ultra analyzers were compared with results from the Beckman analyzer, which was considered to be a reference method. Results for serum chloride, potassium, sodium, creatinine, urea nitrogen, and L‐lactate concentrations obtained from the Prime Plus VET and Ultra analyzers were compared. Passing‐Bablok regression and Bland‐Altman plots were used for method comparison. Results: Significant ( P < 0.05) constant or proportional bias was found for many analytes for all three method comparison studies. Conclusions: Due to the presence of statistically significant differences between all three analyzers that may be clinically relevant, it is recommended that reference intervals be created for new blood gas analyzers, even when similar methodologies are used.