Bacterial Contamination of QC Specimens as a Cause of Artefactually Poor Performance in the UK EQAS

Abstract

SUMMARY.During 1984-85, users of a solid-phase radioimmunoassay kit for LH and FSH had problems of both variability and bias in the assay of some EQAS specimens despite adequate internal QC. The cause has been identified as contamination of these specimens with Pseudomonas fluorescens.From mid-1984 onwards some users of the Chelsea Hospital for Women (CHW) solidphase kit for the immunoassay of luteinising hormone (LH) and follicle stimulating hormone (FSH) experienced difficulty in the assay of some (though not all) specimens distributed by the UK External Quality Assessment Scheme (EQAS) for peptide hormones. These laboratories were repeatedly 'flagged' as being outliers and shown to have positive bias and poor reproducibility. Most of the users of this kit experienced the problem to a greater or lesser extent, although it was variable. No other kit method in the LH and FSH scheme appeared to be affected. Several users contacted the manufacturers (CHW), being concerned for the reliability of their assays for local patient care; the kit manufacturer was naturally concerned to exclude a possible kit problem and to maintain public professional confidence in the method. Since in both users' and manufacturer's laboratories internal quality control procedures were satisfactory and had revealed no changes in assay performance, suspicion fell on the EQAS specimens. Two of us (LRK and KMF) had discussions with the UK EQAS organisers regarding the potential problem and this is a report of an investigation and delineation of its cause.Methods lmmunoassays for LH and FSH were performed in the Endocrine Laboratory, Barking Hospital using the CHW kit l with the SacCd (Wellcome Reagents) solid-phase separation 182 system. In some experiments the Pharmacia 'Decanting Suspension' (second antibodycoated microscpharose beads) was used for comparison.Bacteriology. The isolation and identification of the bacterial contaminant was done in the Department of Microbiology, Barking I lospital.Aggregation screening lest. Since visible aggregation of the immunoassay incubation mixture was a readily observable effect of the contaminated specimens this was used as a screening test. The procedure was as follows: the test sample «()·I mL) was incubated with SacCel (0,1 mL) for 3() min at room temperature. The CHW kit wash solution (1 mL) was then added and the occurrence of aggregation noted.Experimental contamination of serum and plasma with the bacterium isolated from the EQAS specimens was studied for its effect on (a) aggregation in thc incubation medium and (h) LH and FSf I potency estimates using the CHW kit. ResultsPerformance in UK EQAS. Poor performance with certain UK EQAS LH and FSII specimens was experienced in the Endocrine Laboratory, Barking Hospital from mid-11.)84 onwards. For example, during the 6-month period MarchAugust 11.)85 (distributions 64-61.), specimens