Trueness Evaluation and Verification of Interassay Agreement of 11 Serum IgA Measuring Systems: Implications for Medical Decisions Federica Braga, 1* Ilenia Infusino, 1 Erika Frusciante, 1 Ferruccio Ceriotti, 2 and Mauro Panteghini 1 BACKGROUND: To identify an IgA deficiency, the avail- ability of reliable IgA lower reference limits is essential, especially in pediatrics. In this study, we reported the results of an intercomparison study aimed to verify the status of standardization of IgA measurements using 11 commercially available measuring systems (MSs). METHODS: After confirming its commutability, the ERM- DA470k/IFCC reference material was used for the true- ness evaluation of IgA MSs. Furthermore, the interassay agreement was verified using 18 patient pools. By com- bining the bias, if any, between the obtained mean of ERM-DA470k/IFCC and its target value and the mean imprecision of MSs with the uncertainty of respective calibrators, we also estimated the mean uncertainty (U) of IgA measurements on clinical samples. RESULTS: Although the majority of IgA MSs were suffi- ciently aligned with each other, the bias against the ERM-DA470k/IFCC target value was unacceptable in 55% of cases. This bias resulted in an excessive U of IgA measurement on clinical samples. Importantly, when the analysis focused on the lower IgA concentrations— typical of children—the situation worsened, with only 4 MSs showing good equivalence. CONCLUSIONS: Although the harmonization among most commercially available IgA MSs is good, the implemen- tation of traceability to higher order references is inade- quate, especially at concentrations 0.7 g/L. This ana- lytical background information needs to be considered carefully when defining traceable reference intervals in the pediatric population. © 2018 American Association for Clinical Chemistry IgA accounts for approximately 13% of the serum immu- noglobulins. Increased IgA concentrations in serum are associated with both polyclonal [e.g., chronic liver disease (especially alcohol-induced), chronic infections, inflam- matory bowel disease] and monoclonal increases. Assess- ment of IgA concentrations in serum is also important to exclude a selective IgA deficiency condition that occurs more frequently in patients with celiac disease (CD) 3 and that may lead to false-negative results for some serologic tests recommended for CD identification, such as tissue transglutaminase IgA antibodies (1). Measurement of se- rum IgA is the first step of the testing algorithm recom- mended for the screening of CD (2). For correctly identifying an IgA deficiency condi- tion, reliable reference intervals are essential, in particu- lar, for the pediatric population in whom CD frequently occurs. As previously discussed for other analytes (3), the recommendation of reference limits as action limits in interpreting abnormal results strongly depends on 2 basic premises. First, fixed limits, be they medical decision thresholds or reference limits, can be used only if the laboratory test in question is standardized. Second, stud- ies producing reference intervals should fulfill rigorous experimental design (4). The Canadian Laboratory Ini- tiative on Pediatric Reference Intervals (CALIPER), using the Abbott Architect c8000 system, defined IgA reference intervals for different groups of multiethnic children and adolescents, showing a lower reference limit (LRL) close to the analytical limit of detection in infants 1 year of age, and IgA LRLs of 0.04 g/L, 0.26 g/L, and 0.47 g/L in children of ages 1 to 3 years, 3 to 6 years, and 6 to 14 years, respectively (5). Similar results were obtained in smaller and more homogeneous groups of individuals using nephelometric techniques (6, 7 ). As a preliminary phase to verify the applicability of these data to the Italian pediatric population, we evaluated the sta- tus of standardization of 11 commercially available IgA measuring systems (MSs) to confirm whether common LRLs, independent of the used commercial MS, can be implemented. 1 Research Centre for Metrological Traceability in Laboratory Medicine (CIRME), University of Milan, Milan, Italy; 2 Clinical Laboratory, Fondazione IRCCS Ca’ Granda Ospedale Mag- giore Policlinico, Milan, Italy. * Address correspondence to this author at: UOC Patologia Clinica, ASST Fatebenefratelli- Sacco, via GB Grassi 74, 20157 Milano, Italy. Fax +390250319835; e-mail federica. braga@unimi.it. Received September 21, 2018; accepted December 7, 2018. Previously published online at DOI: 10.1373/clinchem.2018.297655 © 2018 American Association for Clinical Chemistry 3 Nonstandard abbreviations: CD, celiac disease; LRL, lower reference limit; MS, measur- ing system; U, expanded uncertainty; CLSI, Clinical and Laboratory Standards Institute. Clinical Chemistry 65:3 473–483 (2019) Pediatric Clinical Chemistry 473 Downloaded from https://academic.oup.com/clinchem/article-abstract/65/3/473/5608015 by guest on 04 July 2020