Commutability of quantitative standards allows patient results to be compared across molecular diagnostic methods and laboratories. This is critical to establishing quantitative thresholds for use in clinical decision-making. A matrix effect associated with the 1st cytomegalovirus (CMV) WHO international standard (IS) was identified using the Abbott RealTime CMV assay. A commutability study was performed to compare the CMV WHO IS and patient specimens diluted in plasma and whole blood. Patient specimens showed similar CMV DNA quantitation values regardless of the diluent or extraction procedure used. The CMV WHO IS, on the other hand, exhibited a matrix effect. The CMV concentration reported for the WHO IS diluted in plasma was within the 95% prediction interval established with patient samples. In contrast, the reported DNA concentration of the CMV WHO IS diluted in whole blood was reduced approximately 0.4 log copies/ml, and values fell outside the 95% prediction interval. Calibrating the assay by using the CMV WHO IS diluted in whole blood would introduce a bias for CMV whole-blood quantitation; samples would be reported as having higher measured concentrations, by approximately 0.4 log IU/ml. Based on the commutability study with patient samples, the RealTime CMV assay was standardized based on the CMV WHO IS diluted in plasma. A revision of the instructions for use of the CMV WHO IS should be considered to alert users of the potential impact from the diluent matrix. The identification of a matrix effect with the CMV WHO IS underscores the importance of assessing commutability of the IS in order to achieve consistent results across methods.

No abstract

In this paper the Kulman-Filter algorithm is applied to the Arrhenius lije model, normally used f o r thermal endurance characterization of insulating materials. It is shown how this algorithm con be used to derive estimates of Temperature Index and Halving Intervals, i. e. the indices recommended by IEC216, using thermal endurance data relevunt to three insulating materials or systems. Depending on sccittering of the experimental points from the assumed linear model, especially of the points at the lowest temperatures, estimates performed by Kalman filter and according to IEC216 may considerably diffeer: Tests of the validity of the model, which are available in the Kalman procedure, show that there is a high probability that the linecir model is not valid at the lowest test temperatures.

No abstract

No abstract