[Optimization of blood gas analysis in intensive care units : Reduction of preanalytical errors and improvement of workflow].

Background: Point of care testing with blood gas analysis (BGA) is an important factor for intensive care medicine. Continuous efforts to optimize workflow, improve safety for the staff and avoid preanalytical mistakes are important and should reflect quality management standards.

Aim: In a prospective observational study it was investigated whether the implementation of a new system for BGA using labeled syringes and automated processing of the specimens leads to improvements compared to the previously used procedure.

Material And Methods: In a 4-week test period the time until receiving the final results of the BGA with the standard method used in the clinical routine (control group) was compared to the results in a second 4-week test period using the new labeled syringes and automated processing of the specimens (intervention group). In addition, preanalytical mistakes with both systems were checked during routine daily use. Finally, it was investigated whether a delay of 10 min between taking and analyzing the blood samples alters the results of the BGA.

Results: Preanalytical errors were frequently observed in the control group where non-deaerated samples were recorded in 87.3 % but in the intervention group almost all samples (98.9 %) were correctly deaerated. Insufficient homogenization due to omission of manual pivoting was seen in 83.2 % in the control group and in 89.9 % in the intervention group; however, in the intervention group the samples were homogenized automatically during the further analytical process. Although a survey among the staff revealed a high acceptance of the new system and a subjective improvement of workflow, a measurable gain in time after conversion to the new procedure could not be seen. The mean time needed for a complete analysis process until receiving the final results was 244 s in the intervention group and 201 s in the control group. A 10-min delay between taking and analyzing the blood samples led to a significant and clinically relevant elevation of the values for partial pressure of oxygen (pO2) in both groups compared to the results when analyzing the samples immediately (118.4 vs. 148.6 mmHg in the control group and 115.3 vs. 123.7 mmHg in the intervention group). When using standard syringes the partial pressure of carbon dioxide (pCO2) was significantly lower (40.5 vs. 38.3 mmHg) whereas no alterations were seen when using the labeled syringes.

Conclusion: The implementation of a new BGA system with labeled syringes and automated processing of the specimens was possible without any difficulties under daily clinical routine conditions in this 10-bed intensive care unit (ICU). A gain of time could not be measured but a reduction in preanalytical errors using the labeled syringes with automated processing was found. Delayed analysis of blood samples can lead to significant changes in pO2 and pCO2 depending on the type of syringe used.