Prevention of fatal and disabling neuroglycopaenia: a comparison of arterial blood sampling methods

Introduction

In 2008, the UK National Patient Safety Agency (NPSA) published a report following 42 incidents and two deaths where glucose-containing flush solutions were attached to arterial lines. Deaths occurred when blood samples contaminated with the flush solution led to artificially high blood glucose readings, inappropriate insulin administration and iatrogenic neuroglycopaenia (the molar concentration of 5% glucose at 277 mmol/l is huge next to physiological blood glucose at <11 mmol/l). The NPSA sought a solution so we applied a bench model to test the performance of three open and three closed arterial line systems in preventing sample contamination.

Methods

All arterial line systems were set up in a standard manner and pressurised to 300 mmHg with 5% glucose used as the flush solution. This was connected to the radial artery using an 18 G needle representing the radial cannula. The radial artery was simulated using a wide bore extension set with blood flow at 60 ml/minute. Blood was simulated by the addition of red dye to Hartmann's solution. Increasing multiples of arterial line dead space were aspirated and discarded. Blood samples were then taken and the glucose concentration measured.

Results

Significant glucose contamination (3 mmol/l ± 3.4) was detected in all open arterial line systems up to an aspiration volume of five times the dead space. No samples from the closed systems recorded glucose concentration >1 mmol/l (0.2 ± <0.1) (Figure 1).

Figure 1

Conclusions

Recommended minimal discard volumes are inadequate in the presence of glucose as the flush solution. Human factors ensure that such errors can persist without a systems-based solution. Our study demonstrates that the closed loop arterial sampling system could be the universal solution sought by the NPSA.