Discrimination and other statistical intervals for the interpretation of in vivo patient monitoring data.

A calibration line is used to define the relationship between a new clinical technique and a standard in vitro laboratory methodology. Discrimination intervals quantify the reliability of inverse estimates obtained from the calibration line. Applied to transcutaneous PCO2 monitoring, a new in vivo measurement, discrimination intervals for inverse estimates of arterial PCO2 range in size from 10 to 50 torr (1.33-6.67 kPa), or more, depending on the values of statistical parameters selected. Discrimination interval analysis shows that the inverse estimates are more reliable near the means. Also, it indicates that the primary value of transcutaneous PCO2 monitoring lies not in providing accurate inverse estimates of arterial PCO2, but instead, in monitoring short-term clinical trends.