TABLE 2.

Assay quality and performance indicators that may be considered standard reporting requirements of laboratory practices and characteristics in nutritional research¹

Category	Definition	Explanations and examples
Protocols for specimen collection and handling and laboratory procedures	Detailed outline of procedures and materials sufficient to enable another investigator to replicate the analysis.	Specimen information should include special considerations where appropriate (e.g., trace mineral–free blood collection materials) and details of specimen storage relevant to analyte stability (e.g., number of freeze–thaw cycles). Specific information about commercial kits should include the manufacturer and product number. Detailed protocols and procedures, including QA and QC methods, may be included in supplemental file(s).
LOQs and reportable range	LLOQ and ULOQ-lowest and highest concentrations, respectively-of analyte that can be repeatedly measured with acceptable accuracy and precision (17). Reportable range is the range of values across which results may be quantified and reported for a specific assay in a particular laboratory, including values generated by any standardized pretreatment procedures (e.g., sample dilution) (16).	LLOQ typically refers to the concentration of lowest standard on the calibration curve. LLOQ is distinguished from the LOD, which is lowest concentration of analyte that can be reliably and feasibly differentiated from an acknowledged blank concentration. LLOQ can be ≥LOD but not <LOD (17). Approaches for defining, imputing, or otherwise handling values above/below LOD/LLOQ and ULOQ should be reported.
Precision	Closeness of individual repeated measurements of the same sample, usually described empirically as a measure of imprecision (45), and determined by both within- and between-assay comparisons of results of 2 or more replicates.	SDs and CVs (inter- and intra-assay) of individual repeated measurements under controlled conditions may be used to express precision. CVs may be used to convey within-run as well as between-run variation across batches, personnel, etc. Single CV values for each analyte are less informative than multiple estimates spanning detectable or clinically relevant ranges (e.g., low-, medium- and high-concentration control materials).
Accuracy	Extent to which assay produces “true” results relative to the gold-standard. Bias is average systematic difference between the test result obtained and accepted reference value; also known as systematic measurement error, as distinguished from random error (49).	Accuracy/bias is typically estimated by use of external reference material for which a “true” assigned value is known for the sample. Generally accepted range for variation from true value is ≤5%.
Participation and performance in external quality assessment program	Where applicable, participation in accuracy-based performance testing and/or external quality assurance schemes is encouraged and should be reported.	Results of any proficiency tests should be reported, e.g., VITAL-EQA program (48), DEQAS (50).

¹DEQAS, Vitamin D External Quality Assessment Scheme; LLOQ, lower limit of quantification; LOD, limit of detection; LOQ, limit of quantification; QA, quality assurance; QC, quality control; ULOQ, upper limit of quantification; VITAL-EQA, Vitamin A Laboratory—External Quality Assurance.