Neonatal Laboratory Blood Sampling: Comparison of Results from Arterial Catheters with Those from an Automated Capillary Device

Purpose: To compare neonatal laboratory results from capillary blood samples drawn using the Tenderfoot automated capillary sampling device with those drawn through arterial catheters.

Design: Prospective, paired comparisons of laboratory results from capillary and arterial blood.

Sample: Twenty-one infants being cared for in an NICU and having indwelling arterial catheters through which a variety of predominantly glucose-containing fluids were being administered.

Main Outcome Variables: Statistical comparisons of paired capillary and arterial results of pH, PO₂, PCO₂, lactate, sodium, potassium, ionized calcium, and hematocrit.

Results: No capillary-arterial differences were observed for pH, PCO₂, lactate, or sodium. Although capillary results were slightly, but significantly (p <.01), higher for potassium (+0.4 mEq/liter), ionized calcium (+0.47 mg/dl), and hematocrit (+4 percent), these differences fell within acceptable Clinical Laboratories Improvement Act (CLIA) performance criteria. Markedly lower PO2 (−30.2 mmHg) and glucose (−61 mg/dl) values were observed with capillary sampling. With the exception of results for PaO₂ and plasma glucose, capillary blood drawn using the Tenderfoot automated device yields laboratory results comparable to those from blood drawn from arterial catheters as assessed by CLIA performance criteria.