 |
|
|
 |
A Critical Evaluation of Enzyme Immunoassay Kits for Detection of Antinuclear Autoantibodies of Defined Specificities. III. Comparative Performance Characteristics of Academic and Manufacturers' Laboratories
MARVIN J. FRITZLER, ALLAN WIIK, ENG M. TAN, JOSEF S. SMOLEN, J. STEVEN McDOUGAL, EDWARD K.L. CHAN, THOMAS P. GORDON, JOHN A. HARDIN, JOACHIM R. KALDEN, ROBERT G. LAHITA, RAVINDER N. MAINI, WESTLEY H. REEVES, NAOMI F. ROTHFIELD, YOSHINARI TAKASAKI, MERLIN WILSON, MARTHA G. BYRD, LLOYD SLIVKA, and JAMES A. KOZIOL
ABSTRACT.
Methods. EIA kits for detection of ANA from 9 commercial manufacturers were evaluated. The manufacturers were advised that they would be sent coded sera containing mixtures of the Arthritis Foundation/Centers for Disease Control reference reagents, and that they were to use their own test kits to analyze the antibody specificities of these sera and to report the data, in optical density (OD) units or their equivalent. Independently, 12 investigators in academic institutions who have done research in this field agreed to participate in a parallel study. The concentration of the antibodies and the specificities were blinded to the analysts and the coefficients of variation (CV) were computed for each participant. Results. There were statistically significant differences between laboratories in terms of CV for all 9 kits tested. With the exception of one kit, there were no significant CV differences between the various autoantibody kits provided by each manufacturer and, with the exception of kits from 2 manufacturers, there were no significant differences between the various antibody kits in terms of reproducibility (CV). From the point of view of interlaboratory variability, manufacturers could be separated into either a high or low performance group. Conclusion. We found a disconcertingly large range of performance characteristics in the various laboratories, which could be quite detrimental in routine utilization of EIA ANA kits. Clinicians should be aware of the performance issues raised in our study, and should know and be involved in how their service laboratory assesses its own performance and the performance of commercial testing systems utilized. Manufacturers and clinical laboratories need to exercise constant quality assurance and surveillance of kit performance in the hands of medical laboratory technologists involved in routine testing. (J Rheumatol 2003;30:2374-81) Key Indexing Terms:
ELISA
From the Standardization Committee in Rheumatic and Related Disorders of the International Union of Immunological Societies; The World Health Organization; The Arthritis Foundation and Centers for Disease Control and Prevention (CDC), Atlanta, Georgia, USA. M.J. Fritzler, PhD, MD, University of Calgary, Calgary, Canada; A. Wiik, MD, State Serum Institute, Copenhagen, Denmark; E.M. Tan, MD, Scripps Research Institute, La Jolla, CA, USA; E.K.L. Chan, PhD, University of Florida, Gainsville, FL, USA; T.P. Gordon, MD, Flinders University, Adelaide, Australia; J.A. Hardin, MD, Einstein Medical College, New York, NY, USA; J.R. Kalden, MD, University of Erlangen, Erlangen, Germany; R.G. Lahita, MD, St. Vincent Hospital, New York, NY, USA; R.N. Maini, MD, Kennedy Institute, London, UK; J.S. McDougal, MD, CDC, Atlanta, GA, USA; W.H. Reeves, MD, University of Florida; N.F. Rothfield, MD, University of Connecticut, Farmington, CT, USA; J.S. Smolen, MD, University of Vienna, Vienna, Austria; Y. Takasaki, MD, Juntendo University, Tokyo, Japan; M. Wilson, MD, Immunotek Reference Laboratory, New Orleans, LA, USA; M. Byrd, BS, MT(ASCP), CDC; L. Slivka; J.A. Koziol, PhD, University of Calgary. Address reprint requests to Dr. M.J. Fritzler, Faculty of Medicine, 3330 Hospital Dr. NW, Calgary, AB, Canada T2N 4N1. E-mail: fritzler@ucalgary.ca Submitted August 22, 2002; revision accepted March 28, 2003. |