The (in-)validity of volatile POCT parameters from patients beyond normothermia

A common characteristic of all blood gas analyzers on the market is that measurements are processed at 37°C, not at the real patients´ temperature. Subsequently temperature-sensitive parameters can be mathematically corrected (alpha-stat method) or used directly (pH-stat method). National rules in Germany (Rili-BAEK) demand defined accuracy and precision without any restriction to samples´ temperatures or corrections. As consequence in the investigation at hand we tried to find out whether blood gas analyzers can fulfill the regulations for pCO₂ and pO₂ when normothermia of the matrix is not given.

Five matrices (blood from intensive care unit (ICU) patients, blood from healthy donors and 3 levels of bovine based quality control material) were tonometered at "high" and "low" partial pressures of O₂ and CO₂ within the RiLi-BAEK controlled range at 32, 37 and 40°C. One mL material was aspired into each blood gas (BG) syringe and analysis was accomplished immediately after. The procedure was repeated 10-fold for "high" and "low" gas concentrations and run on 4 different analyzers. At 18°C instead to the "high" one a "median" gas (n=10 as well) was employed. Every condition which constitutes of temperature (4), matrix (5), analyzer (4) and level of the partial pressure (2) led to a total of 1600 measurements.

At 32°C or 37°C matrix temperature 7.5% to 27.5% of the pCO₂(T) and between 14.5% and 28.1% of the pO₂(T) results were outside the borders required by the RiLi-BAEK. At 18°C or 40°C the number of results beyond the allowed borders grows up to 82.5% for pCO₂(T) and 73% for pO₂(T) depending on the partial pressure (PP) level.

High precision in automated quality control (at a constant matrix temperature) is given in modern BGAnalyzers but is counteracted in practice by non normothermic patient's temperature and unavoidable sample handling effects.