Evaluation of point‐of‐care analysers for blood gas and clinical chemistry in Hermann’s tortoises (Testudo hermanni)

Objective To assess the agreement between point‐of‐care and laboratory analysers in measuring biochemical and blood gas analytes in venous samples from tortoises and to define preliminary reference intervals for venous blood gas analysis in Hermann's tortoises (Testudo hermanni). Materials and Methods Jugular venous blood samples from 47 Hermann's tortoises underwent paired analysis with a portable gas analyser (i‐STAT 1, Abaxis), a portable chemical analyser (VetScan VS2, Abaxis), and with the respective reference analysers. Agarose gel electrophoresis was used to determine albumin concentrations on 12 specimens. Agreement was evaluated with Bland–Altman plots and regression analysis using the Passing–Bablok method. Results Point‐of‐care analysers had variable agreement with the reference analysers, presenting constant or proportional bias depending on the analyte. Relevant analytes in reptiles, such as ionised and total calcium, had acceptable agreement. The method for determining albumin concentration currently available in both point‐of‐care and laboratory analysers significantly overestimated albumin concentrations as compared to protein electrophoresis. Clinical Significance While the use of POC analysers is extremely advantageous in small animal primary care facilities, agreement between point‐of‐care and laboratory analysers varies depending on the analyte. For certain analytes, interchangeability of results is limited and specific reference intervals for point‐of‐care analysers are required. Veterinarians should be aware of the size and the direction of the bias of each analyte.