Evaluation of measurement uncertainty in direct and estimated serum osmolality according to ISO/TS 20914: implications for clinical diagnostics.

Abstract

This study examines the measurement uncertainties (MU) associated with osmolality (OSM), a crucial parameter in clinical diagnostics and treatment. The research compares the uncertainties of directly measured osmolality (dOSM) with those of estimated osmolality (eOSM), which is calculated based on sodium, glucose, and urea levels. Using ISO/TS 20914 guidelines, the study evaluates the MUs of eOSM across different analyzer models and internal quality control (IQC) lots and compares these values with dOSM. In the materials and methods section, sodium, glucose, and urea analyses were performed using two different analyzers (Atellica CH and Advia Chemistry XPT, Siemens Healthineers). OSM was measured using the freezing point depression method (Osmomat Auto, Gonotec) and data were collected to calculate eOSM. MU (k = 2, 95% confidence) calculations were conducted according to the ISO/TS 20914 standard for each system. The results show that the expanded standard MU (k = 2, 95% confidence) for dOSM is low at 5.56 mOSM/L, while the MU for eOSM is 8.54 mOSM/L for the Atellica CH system and 11.13 mOSM/L for the Advia Chemistry XPT system. These findings indicate that eOSM has higher uncertainty, suggesting it should be used with caution in clinical practice.