The systematic monitoring of transfusion microbiology test kit performance

Abstract

The Transfusion Microbiology Test Systems Monitoring Group (TMTSMG) was established as a National Blood Service (NBS) working group to monitor the performance of the microbiology screening assays used within the NBS Testing Laboratories. The group's primary objective was to ensure that technical performance (especially sensitivity, specificity and wastage) remains consistent with that established during validation. This includes the identification and investigation of significant variation in performance and any untoward incidents. The group is also responsible for optimizing transfusion microbiology working practice across the NBS through nationally agreed standards and procedures. Over the past 9 years, a total of 44 assays from 15 suppliers have been monitored. Five assays have been withdrawn from use as a result of identified poor performance; two hepatitis B virus surface antigen assays owing to poor sensitivity, two syphilis agglutination assays with nonspecific (false) reactive rates sustained above contract limits and one human cytomegalovirus antibody assay that persistently failed the manufacturer's quality control criteria. This approach has enabled the differentiation of genuine kit performance issues from 'natural variation' in kit performance, and local instrumentation or training issues. The NBS has been able to address the issues with suppliers much earlier and resolve minor issues before they became major problems. In addition, a lot release system has been developed and implemented, comprising a formal, centralized initial scientific assessment of each new manufacturer's lot, followed by 'delivery acceptance' testing at each site. This system helps to ensure that the evaluated minimum sensitivity and specificity of the assays is maintained from 'lot to lot'.

The donor line break cannula: effect on the donation process, blood component quality and transfusion microbiology testing of an important new blood bag safety feature

“…This is to be expected owing to the higher protein levels and platelet constituents in plasma, known to cause non‐specific reactions in immunoassays. Therefore, for assays showing a higher specificity with plasma, this may reflect assay improvements mediated by the NHSBT Transfusion Microbiology Test System Monitoring Group which strives through manufacturers to improve their assay performance (Nightingale et al ., ).…”

Application of quality principles in transfusion microbiology

“…It is therefore essential not to ignore proper quality control measures on this area. Recently, Nightingale et al reported on the UK approach in systematic monitoring of transfusion microbiology test kit performance [3]. Then Kitchen and Newham further summarized their experience on lot release testing for serological infectious disease assays [4].…”

Twenty-Eight Grams of Prevention Is Worth a Half Kilogram of Cure

“…Lots successfully passing this initial testing are distributed to testing laboratories for "delivery acceptance" testing. During a 9-year period, initial validation testing at the reference laboratory rejected 7 lots, and 5 failures occurred in delivery-acceptance testing (6 ). As a national entity, the NBS has the critical mass to centralize some aspects of validation testing and to aggregate data from distributed testing, thereby improving the costeffectiveness and robustness of validation studies.…”

“…The NBS instituted a prospective monitoring system across all laboratories to assure ongoing high sensitivity for infectious disease agents while also maintaining high specificity (6 ). Indicators of poor assay performance include initial and repeat reactive rates, discrepancies between duplicate test results, QC failures or trends, instrument problems, wastage, quality incidents, discrepancies with subsequent donor testing, and more.…”

“…Aggregating data from all testing laboratories increases the sensitivity for early detection of inherent or emerging problems. Five assays have been withdrawn on the basis of findings from prospective monitoring (5,6 ). Additional approaches for prospective monitoring include the use of patients' means or medians in average-of-normals algorithms (8 ), the proportion of patients' results outside of reference intervals (2 ), ␦ checks, specimen and tube type verifications, and absurdity checks (9 ).…”