Abstract
Bicarbonate balance is usually calculated from indirect total CO2 values determined by blood-gas analysis. Total CO2 is computed from the Henderson-Hasselbalch equation, under the assumption that the pK' is constant or at most varies insignificantly. In pediatric patients in intensive care, we have found instances of large errors in this assumption. We join Natelson and Nobel in asserting that total CO2 should be measured. Little reliance should be placed on bicarbonate estimation.