Research Project
Full description This prospective observational study was conducted at a single veterinary hospital. The investigation was approved by the institutional Animal Ethics Committee (Approval number 2021/AE000818). Forty-eight healthy dogs were enrolled in the study. Animals were either privately owned or part of the institutional colony and written informed owner consent was obtained prior to inclusion of dogs in the study. Dogs were eligible for inclusion if they were healthy based on results of a physical examination, were not currently receiving medication beyond routine prophylactic care, had not undergone blood sample collection in the previous 3 months and body weight was > 3 kg. Dogs then had a CBC, biochemical profile, PT, and aPTT performed and were excluded if values were >10% outside of the laboratory or device specific reference interval. Dogs that became distressed during sample collection or that required more than two attempts for blood draw were excluded. Blood sampling The area over the jugular vein was clipped and disinfected using a chlorhexidine-alcohol swab. Blood samples were collected from the jugular vein with a 23-gauge needle attached to a nonheparinized 10 mL syringe. Blood was immediately transferred to a 2 mL serum tube, a 2 mL 3.2% sodium citrate tube (ratio of 1 mL sodium citrate to 9 mL blood), and a 2 mL ethylenediamine tetra-acetic acid (EDTA) tube (in that order), ensuring proper filling of collection tubes for correct ratios of anticoagulant to blood. For 11 dogs, tubes were filled as described but an additional 2 mL 3.2% sodium citrate tube was collected. Sodium citrate and EDTA tubes were immediately inverted at least three times to ensure adequate mixing. Serum and EDTA tubes were analyzed for CBCb and biochemical analysisc. The 3.2% citrate tube had 0.2 mL removed to perform the PT and aPTT using an IDEXX Coag Dx Analyzerd. The citrated samples were kept at room temperature (20 - 25°C) until viscoelastic testing was performed. ClotPro analysis All viscoelastic measurements were performed as per the Partnership on Rotational ViscoElastic Test Standardization (PROVETS) recommendations3. To determine the device specific reference intervals, a single operator performed each of the following five tests in parallel 30 minutes after sample collection: EX-test, IN-test, FIB-test, ECA-test and TPA-test (Table 1). To determine the effect of storage time, the additional 11 citrated samples were analyzed as described at 90 and 150 minutes after collection for the EX-test, IN-test and FIB-test assays. All tests were performed according to the standard test protocol outlined in the ClotPro User Manual7. Briefly, following entry of the patient details into the ClotPro electronic system, each cup and pin were loaded into the test positions on the 37°C prewarmed instrument bench. The appropriate test tip was loaded onto the electronic pipette, and 340 uL of citrated blood was transferred from the collection tube into the cup. To ensure adequate mixing, the blood was then redrawn into the pipette and replaced back into the test cup. The pin was then connected to the cup to initiate the test. Each test was run for 60 minutes, and parameters were automatically measured and graphically displayed by the ClotPro device (see Table 2 for description of parameters measured). Statistical methods Reference intervals were calculated following Friedrichs, et al. 16 American Society for Veterinary Clinical Pathology (ASVCP) veterinary species reference interval guidelines using the Reference Value Advisor v2.117 Excel Ad-in. Normality was assessed with Q-Q plots, histograms and the Anderson-Darling tests. For normal distributions, the robust method was used to calculate reference intervals. For non-normal distributions, robust Box-Cox transformed (symmetric data without outliers) or non-parametric (non-symmetric and/or outliers) were used. Pearson correlation coefficients were estimated for the correlations of clotting time (CT), aPTT and PT. Due the presence of outliers or non-normal distributions, Spearman correlation coefficients were estimated for the correlations of CT, clotting formation time (CFT), amplitude at 10 minutes (A10), α angle (α) and maximum clot formation (MCF) with hematocrit, total protein, and platelet count. Correlation analysis p-values were adjusted for multiple comparisons with the linear step-up false discovery rate method. Interpretation of the magnitude of the observed correlation co-efficient was as categorized as negligible correlation to very strong correlation as described by Schober, et al. 18. Linear mixed models were used to test for effects of storage times and ClotPro parameters for each parameter and reagent. The models had a single fixed factor of time and a random intercept for each sample. Conditional model residuals and Q-Q plots were examined to confirm the assumption of normality. If normality was not confirmed, a Friedman test was used. Post-hoc testing was performed using Dunnett's test for the linear mixed models and Conover tests for the Friedman tests. All analyses were performed using SAS 9.4e. A significance threshold of 0.05 was used.