Comparison of arterial haemoglobin and electrolyte measurements between an arterial blood gas analyser and the laboratory on the critical care unit

Introduction

Arterial blood gas analysers are commonly used in ICUs to provide a measurement of haemoglobin and electrolytes to determine progress and titrate therapy. Comparison between these analysers and formal laboratory measurement has not previously been undertaken in a large number of samples in the setting of a critical care unit. Anecdotal evidence suggests possible poor correlation.

Hypothesis

To compare and contrast two methods of haemoglobin and electrolyte evaluation, expecting a significant difference.

Design

A prospective, paired observational study.

Methods

Patients had paired samples taken from their arterial line for arterial blood gas analysis via a Radiometer PICO 70 arterial sampler, and also samples taken into Vacutainer collection bottles for formal laboratory analysis. A Radiometer Copenhagen ABL700 or ABL625 blood gas analyser measured the total haemoglobin, sodium, potassium and chloride. Formal laboratory samples were analysed on GenS (haemoglobin) and LX20 (sodium, potassium and chloride) analysers (Beckman-Coulter). Samples were compared via a paired t test.

Results

Two hundred and thirty-eight paired samples were collected over a 2-month period. The correlation coefficients between arterial blood gas analysis and formal laboratory analysis for haemoglobin, sodium, potassium and chloride were 0.934, 0.945, 0.817 and 0.922, respectively. There was no significant difference between the two methods of measurement in this study when measuring these variables (P < 0.001)

Conclusion

Arterial blood gas analysis is an accurate method of measuring haemoglobin, sodium, potassium and chloride in the critical care unit. The transfusion practice and electrolyte or fluid resuscitation can therefore be accurately guided by the arterial blood gas measurements. This will be dependent on the accurate and regular calibration of the measuring device.