A pilot study comparing bone marrow aspirates and venous blood for emergency point-of-care blood parameters in healthy dogs.

Objective: To compare intravenous and intraosseous blood aspirates using point-of-care diagnostic equipment available in veterinary hospitals.

Design: Prospective study.

Setting: Private referral hospital.

Animals: Dogs undergoing a tibial plateau leveling osteotomy or extracapsular anterior cruciate ligament stabilization procedure were enrolled.

Methods: Under general anesthesia, simultaneous 0.5 mL intravenous and intraosseous blood samples were collected from the jugular vein and proximal tibia, respectively. Samples were evaluated in duplicate within 10 minutes of collection and averaged for each of the following parameters: blood urea nitrogen (BUN), glucose, packed cell volume, total plasma protein (TPP), plasma lactate, sodium, potassium, chloride, urea, glucose, pH, anion gap, pO_2, and pCO₂ . Normalcy was tested with Kolmogorov-Smirnov test. A Student's t-test and Bland-Altman plot were used to compare intravenous and intraosseous samples.

Results: Twelve dogs were recruited into the study. There were statistically significant differences between intravenous and intraosseous samples for sodium (P = 0.0216), chloride (P = 0.0225), BUN (P = 0.014), and potassium (P < 0.0001), respectively. No significant differences were detected for the other parameters evaluated.

Discussion: The intraosseous space provides an easily accessible, noncollapsible alternative for assessing blood parameters. Omitting potassium, the statistically significant differences noted between sites was not felt to be clinically significant. Although statistically insignificant, the large difference in hematocrit values indicates that the samples should not be used interchangeably.

Conclusion: Intraosseous aspirates, excluding potassium and hematocrit, appear to be a reliable alternative for assessing most point-of-care analytes in healthy dogs, although a larger sample size should be investigated. The application of these data in shock patients is unknown.