Tritium Urinalysis Intercomparison - May 2007

Results

Samples were distributed to fifteen laboratories. Fourteen returned results as requested. Results submitted are presented in Table 1 in the format used by participants in the forms submitted to the Reference Centre. The equations used to evaluate the bias and precision of results are given in Appendix A. The terms used in the equations and in the attached tables and figures are defined thus:

y_jkl =   the result obtained by correcting the measurement for a spiked sample by the value obtained for the appropriate urine blank.

b_jkl =   the relative bias of the result obtained for each aliquot of a particular spiked sample. It represents the difference between the measurement and the true value, normalized by the true value.

b_jk =   the mean relative bias, i.e. the mean of the b_jkl values for a particular spiked sample.

s_jk =   relative precision, calculated as the standard deviation of the relative bias of measurements of aliquots of the same spiked sample.

MDA =   Minimum detectable activity. This is an indication of the detection capability of a participant's analytical method.

MTL =   Minimum testing level. The minimum amount of radioactive material at which a dosimetry service should be tested for it to be in compliance with the Canadian Nuclear Safety Commission (formally the Atomic Energy Control Board) license requirements for performance in an independent testing exercise.

The CNSC has set its performance criteria for bioassay measurements in Regulatory Standard S-106, Revision 1 "Technical and Quality Assurance Requirements for Dosimetry Services"[3]. At levels equal to or greater than the MTL, the limits for acceptable bias and precision are:

-0.25 ≤ Mean relative bias ≤ +0.50

Relative precision ≤ 0.40

Note that the MTL is set at five or more times the MDA. The MTL for the determination of tritium in urine samples is set at 2000 Bq/L in S-106, Revision 1. The analytical performance of Canadian bioassay laboratories has been taken into consideration in setting this specification [3].

Relative Bias:

The y_jkl values and their means are listed in Table 2. The y_jkl values are also presented graphically in Figures 1, 2 and 3.

The b_jkl values are presented in Table 3. Table 4 lists the b_jk values which are also shown graphically in Figures 4, 5, and 6. The b_jkl and b_jk values that are outside the acceptable limits for relative bias are marked with an asterisk in the tables.

Relative Precision:

The s_jk values are presented in Table 5, and Figures 7, 8, and 9. The value that exceeded the acceptable relative precision limit is marked with an asterisk in the table..

For participants who followed the recommended protocol (all in this intercomparison) of only one analyst performing measurements for all the blank and spiked samples, the s_jk values indicate within-analyst variability, in addition to intra-laboratory precision.