Patterns of dissimilarities among instrument models in measuring PO2, PCO2, and pH in blood gas laboratories.

STUDY OBJECTIVES

To ascertain the degree of dissimilarities among blood gas and pH analyzer models of the same and different manufacturers in measurement of PO2, PCO2, and pH using fluorocarbon containing emulsion (FCE) proficiency testing material.

DESIGN

Statistically and graphically analyze data from six recent proficiency testing surveys for the 20 more frequently used models of analyzers.

SETTING AND PARTICIPANTS

Over a 2-year period, approximately 900 participants from blood gas laboratories in the United States analyzed similar ampules from each of 30 lots.

MEASUREMENTS AND RESULTS

Both graphic and statistical comparisons were used to demonstrate differences between manufacturers. For each of the four major manufacturers, comparisons revealed statistically significant differences not only for PO2, but also for PCO2 and pH. Additionally, comparison models within each of the three manufacturers (those with multiple models and > 15 instruments per model represented) disclosed statistically significant dissimilarities among models for each analyte in 115 of 153 model pairings. Previously reported tonometered blood differences among analyzer models for PO2 are qualitatively similar to the differences found in these same models in this FCE study. Model differences are important in research studies and may be clinically important in deciding abnormality, selecting oxygen therapy, or the treatment of patients with respiratory failure or severe respiratory alkalosis.

CONCLUSIONS

To minimize the likelihood of misleading clinicians, laboratory directors should consider the degree of dissimilarity among blood gas analyzer models in current use and when changing instrumentation.