Brain tissue oxygen monitoring: a study of in vitro accuracy and stability of Neurovent-PTO and Licox sensors.

OBJECT: Periods of brain tissue ischemia are common after severe head injury, and their occurrence and duration are negatively correlated with outcome. Accurate and reliable measurement of brain tissue oxygenation (B(ti) pO(2)) may be a key to improve patient outcome after severe head injury. Knowledge of stability and accuracy of the B(ti) pO(2) systems is crucial. We have therefore conducted a bench test study of new Neurovent-PTO (NV) and Licox (LX) oxygen tension catheters to evaluate the sensor accuracy, response time to different oxygen tensions, response to temperature changes and long-term stability.

METHODS: For all experiments five new fluorescent NV sensors and five new electrochemical LX sensors were used. The catheter probes were placed into a container filled with a buffer solution. The solution was equilibrated with five high precision calibration gases. The accuracy of the probes was recorded after an equilibration period of 20 min in O(2) concentrations of 5, 10, 20, 30 and 40 mmHg at 37.0 +/- 0.2 degrees C. The probe response to an increase in temperature from 37.0 degrees C to 38.5 degrees C to 40.0 degrees C in two different gases with O(2) concentrations of 10 and 20 mmHg were analysed. We also recorded the time for reaching 90% of a new oxygen concentration level when switching from one concentration to another. Finally, to test if there was a time-dependant drift in pO(2) recordings, all sensors were left in 10 mmHg O(2) solution for 10 days, and recordings were taken every 24 h.

RESULTS: In all gas concentrations, NV and LX sensors measured pO(2) with high accuracy and stability in vitro (mean differences from calculated values were for NV 0.76-1.6 mmHg and for LX -0.46-0.26 mmHg). Both sensors showed a shorter response time to pO(2) increase (for NV 56 +/- 22 s and for LX 78 +/- 21 s) compared to pO(2) decrease (for NV 131 +/- 42 s and for LX 215 +/- 63 s). NV pO(2) values were more stable for changes in temperature, while LX sensors showed larger standard deviations with increasing temperature (the difference from the calculated values in 19.7 mmHg O(2) at 40 degrees C were for NV probes between 0.5 and 1.7 mmHg and LX between -2.3 and 1.9 mmHg). Both sensors gave stable results with low standard deviations during long-term (10 days) use, but with a slight elevation of measured pO(2) levels by time.

CONCLUSIONS: Both NV and LX were accurate in detecting different oxygen tensions, and they did not deviate over longer recording times. However, LX needed a significantly longer time to detect changes in pO(2) levels compared to NV. Furthermore, LX probes showed an increased standard deviation with higher temperatures.