External quality assurance target setting with NIST SRM 968d material: performance in the 2010 Royal College of Pathologists of Australasia Quality Assurance Program with retinol,α -tocopherol and β -carotene

Since 1999, the Royal College of Pathologists of Australasia (RCPA) Quality Assurance Program (QAP) has conducted an assessment of analytical performance for vitamins, including: vitamin A (total-retinol), vitamin E (α-tocopherol) and β-carotene. ¹

Participating laboratories demonstrate considerable method commonality. ² The calibrators chosen by most participants are commercial secondary calibrators sourced from Bio-Rad Laboratories (Hercules, CA, USA), Chromsystems (Munich, Germany) or Recipe (Munich, Germany); all are referenced to National Institute of Standards and Technology (NIST; Gaithersburg, MD, USA) Standard Reference Material (SRM) 968. ^3,4 A minority of participants prepare in-house calibrators referenced to either NIST or to Sigma (St Louis, MO, USA).

The RCPA QAP aims to provide target values for all analytes where practicable. A standardized procedure of ‘transfer of values’ ^5,6 is used to establish targets, where designated laboratories analyse material for the forthcoming cycles along with the SRM. The raw results are re-calculated against the expected SRM results to set the target for each concentration level.

In previous years, target setting exercises of the RCPA QAP Fat-Soluble Vitamins material have been carried out with NIST-SRM-968c. Targets have been set successfully over successive years for retinol but not for α-tocopherol or β-carotene. In 2008, NIST released the new SRM, 968d, and a decision was made to use this to set targets for the 2010 RCPA QAP Fat-Soluble Vitamins material.

Participant results from the first cycle of 2010 (cycle 22) compared with NIST-SRM-968d target setting results are detailed in Table 1. Target setting was again successful for retinol, but unsuccessful for α-tocopherol and β-carotene, based on linear regression analysis. Hence, the NIST-SRM-968d is not usable as a primary calibrator for α-tocopherol and β-carotene. Therefore, for the second cycle of 2010 (cycle 23) the assigned target was continued for retinol only.

Table 1

RCPA QAP cycle 22 Fat-Soluble Vitamins material results compared with NIST SRM 968d target setting results

	Vitamin A (total retinol)	Vitamin E (α-tocopherol)	β-carotene
NIST 968d – assigned values	1.09 ± 0.17 μmol/L	13.77 ± 1.28 μmol/L	0.145 ± 0.013 μmol/L
RCPA QAP material expected range based on NIST target setting*	0.5–5.9 μmol/L	8–65 μmol/L	0.1–6.4 μmol/L
Linear regression†	y = 1.0032x − 0.0099	y = 0.8020x + 1.5772	y = 0.8031x–0.1065
r 2	0.9987	0.9964	0.9914
Number of labs participating	23	23	12
Target setting outcome	Successful	Not successful	Not successful
Comments	Even spread of results around target. Assessed in the NIST program by HPLC with spectrophotometric detection.	Assessed in the NIST material by HPLC with fluorometric detection, which is different from the detection method of QAP participants. The allowable error reported by NIST approximates the total ALP for the QAP.	The concentration of β-carotene in NIST SRM 968d is too low.
	Wide allowable error of NIST material.
	Reasonable analyte concentration

RCPA, Royal College of Pathologists of Australasia; QAP, Quality Assurance Program; NIST, National Institute of Standards and Technology; SRM, Standard Reference Material; HPLC, high-performance liquid chromatography; ALP, Allowable Limit of Performance

*The RCPA QAP material is provided at six linearly related levels, each level is analysed twice per cycle; with two cycles per annum

^†The linear regression analysis of the NIST SRM 968d target values against all results returned by laboratories in cycle 22 of the RCPA Vitamin QAP

Root cause analysis of the failure to successfully target set α-tocopherol and β-carotene included a review of the NIST material certificates. The concentration of β-carotene is very low in the NIST material. α-Tocopherol is assessed in NIST-SRMs by two methods, both high-performance liquid chromatography, one with fluorometric detection which is different from the detection method of RCPA QAP participants. The median of results from participants of the NIST Micronutrients Measurement Quality Assurance Program is included in the establishment of their SRM values. Furthermore, the allowable error reported by NIST approximates the allowable limit of performance in the RCPA QAP. The allowable error of the primary material needs to be tighter than routine methods, and the methods used for assignment of SRM values need to be derived from higher order methods, ⁷ which is not apparent from the NIST certificate of analysis.

A new NIST-SRM-968e was released in late 2010. This is in relatively quick succession to the NIST-SRM-968d, which was released in October 2008 and has an expiry date of September 2013 on the certificate of analysis. However, as target setting has been unsuccessful on multiple occasions, a decision was made not to pursue target setting with the new material in 2011.

DECLARATIONS

Competing interests: There is no conflict of interest that could be perceived as prejudicing the impartiality of this letter. The manuscript consists of original material and has not been published elsewhere; except in the form of an abstract at the AACB-AIMS Scientific Meeting in Perth, Australia in October 2010.

Funding: This work was conducted as part of the routine activities of the RCPA QAP and the AACB Vitamins Working Party. No additional funding was required.

Ethical approval: Not applicable.

Guarantor: RG.

Contributorship: All authors listed contributed to this work as members of the AACB Vitamins Working Party.

Acknowledgements: None.