SED navigation bar go to SED home page go to SED publications page go to NIST home page SED Home Page SED Contacts SED Projects SED Products and Publications Search SED Pages

contents     previous     next

3.2.2 Determination of Chloride Concentrations in Human Serum SRM's

William F. Guthrie
Statistical Engineering Division, ITL

Johanna B. Smeller
Analytical Chemistry Division, CSTL

Jennifer C. Colbert
Standard Reference Materials Program, TS

The ability to accurately measure electrolyte levels in human blood is critical to the diagnosis of many diseases and medical disorders. Among other things, electrolytes play a key role in maintaining the homeostasis of fluids in the body. Due to their medical importance, NIST offers several different human serum Standard Reference Materials (SRM's) with certified electrolyte concentrations. These materials are used by medical labs and in physicians' offices for measurement assurance and calibration of ionic selective electrode potentiometry (ISEP) measurement systems.

Recently, in a response to a request from the NCCLS, the National Committee of Clinical Laboratory Standards, NIST agreed to certify the concentration of chloride in an existing SRM, SRM 956a, Frozen Human Serum, for the convenience of medical community. Other human serum SRM's certified for chloride have been available for some time, but are in physical forms that are more difficult to handle than the frozen serum.

The chloride concentration was determined by coulometry in SRM 956a. In this method vials of standard reagents must be combined with the unknown samples to determine the chloride levels of the unknowns. Each vial of standard reagents can only be used to make about six measurements of this SRM material, however, requiring the use of a nested experiment design to collect the data. The design allowed estimation of potential day-to-day variability, reagent-to-reagent variability and material inhomogeneity for three levels of chloride in serum. In addition, it accounted for material perishibility and included measurements of a previously certified material, SRM 909b, Lyophilized Human Serum, as a control and potential calibrant. Analysis of the 956a material was uneventful and showed the material to be homogeneous and the measurements to be of high precision. No extra day-to-day or reagent-to-reagent variabilty was observed. Analysis of the control material, however, was somewhat surprising. It showed that the current chloride measurements are not in agreement with the certified values, which are based on the agreement of two independent chemical methods, coulometry and isotope dilution thermal ionization mass spectrometry.

As a result of these control measurements, a more detailed experiment with the control material was designed and run. The goal of this experiment was to determine whether the disagreement was due to instrument calibration (which we could then account for by using the control data as a calibrant), a change in control material handling between the current experiment and the original 909b certification experiment or a change in the chloride concentration of control material. This experiment compared three control materials, the 909b material and the two previous generations of SRM 909, and two methods of sample dissolution.

Analysis of the control check experiment was complicated by the presence of several potential outliers, but showed that the current chloride measurements of the 909b material are not in agreement with the certified values in any case. In contrast, the current measurements and original certification measurements did agree for the two earlier generations of this SRM.

The outcomes of these two experiments will result in certification of chloride in SRM 956a using only the automatic calibration of the coulometric measurement system. Further investigation of the stability of chloride in SRM 909b will also be made. This work will likely include verification of the hypothesized changes in the material using an independent chemical method. If necessary, new instructions on the use of the 909b material will be provided to customers via the SRM Program's product registration service.


Figure 8: Comparison of the 909b (Level II) results from the two current experiments with the certification data. The uncertainties for each experiment are approximate 95% confidence intervals.

contents     previous     next

Date created: 7/20/2001
Last updated: 7/20/2001
Please email comments on this WWW page to