Assessment of Instrument-Specific Variation between Measured and Calculated Bicarbonate

Area of Science:

• Clinical Chemistry
• Acid-Base Balance
• Diagnostic Accuracy

Background:

• Bicarbonate concentration is crucial for assessing acid-base disorders.
• Discrepancies between measured (mHCO3-) and calculated (cHCO3-) bicarbonate can lead to clinical confusion and misdiagnosis.
• Accurate bicarbonate measurement is vital for patient management.

Purpose of the Study:

• To assess the agreement between calculated bicarbonate (cHCO3-) from a Radiometer blood gas analyzer and measured bicarbonate (mHCO3-) from three different chemistry instruments.
• To identify potential biases in bicarbonate measurements across different analytical platforms.
• To improve the reliability of acid-base disorder assessment.

Main Methods:

• Multi-institutional study involving three different chemistry analyzer manufacturers.
• Collected de-identified patient data including plasma mHCO3- and blood gas cHCO3- (arterial and venous) within a 20-minute window.
• Utilized Deming regression and Bland-Altman analysis to determine correlations and biases.

Main Results:

• Good correlations were observed between cHCO3- and mHCO3- across all platforms.
• Bland-Altman analysis revealed the greatest bias between Radiometer and Abbott Architect (-2.63 mmol/L).
• Siemens Advia (0.49 mmol/L) and Beckman AU680 (-0.45 mmol/L) showed minimal bias compared to Radiometer.

Conclusions:

• This study is the first to compare Radiometer cHCO3- with mHCO3- from multiple chemistry analyzers.
• Radiometer blood gas analyzer, Siemens Advia, and Beckman AU680 demonstrate good agreement.
• Abbott Architect may exhibit negative bias in bicarbonate measurements compared to other methods, warranting clinical attention.