Analytical and clinical performance of the epoc blood analysis system: experience at a large tertiary academic medical center.

OBJECTIVES To describe validation and performance of epoc, a blood gas analysis point-of-care system, in a live clinical setting. METHODS Data were collected for 156 epoc systems over 12 months. Preimplementation precision and correlation studies and postimplementation quality assurance data were collected, including test card, reader, and personal data assistant (PDA) failure rates. RESULTS The coefficient of variation was clinically acceptable for all analytes. Correlation studies yielded an R(2) from 0.901 (for sodium) to 0.994 (for potassium) with the Nova analyzer and from 0.961 (sodium) to 0.991 (glucose) with the i-STAT. Average test card failure rate was 13%. Of the PDA/reader units, 55% needed repair within 1 year. CONCLUSIONS The analytical performance showed high precision and good correlation with the Nova and i-STAT platforms. Test card and instrument failure rates were higher than that of the i-STAT system.