[Comparison of arterial blood sample kits]

Most inaccuracies in the analysis of gases and electrolytes in arterial blood samples are due to preanalytic factors, among which is the type of equipment used for blood collection. Our objective was to compare arterial blood gas sample kits used under clinical conditions and to evaluate the impact of delay in estimation on variability in results. In 2 types of study we compared 5 kits (Radiometer's Pico 70, Becton Dickinson's Preset, SIMS Portex's Pro-Vent, SIMS-Concord's Pulsator, and Marquest's Quick ABG). In the first study kitsyringe assignment was randomized for collecting arterial blood samples from 160 consecutive patients to evaluate practical aspects of using them and the presence of bubbles in the samples taken. The second study evaluated the effects of delays of 30 and 60 minutes in estimation and of the type of heparin used in 54 blood samples. The kits which produced the fewest bubbles, gave samples with the greatest stability, and had the least impact on ion concentration were Radiometer's Pico 70 and SIMS-Portex's Pro-vent.