Abstract Number: PB0494
Meeting: ISTH 2020 Congress
Theme: Diagnostics and OMICs » Epigenetics, OMICs and Bioinformatics
Background: Laboratories must participate in an interlaboratory comparison program, such as external quality assessment (EQA) that substantially fulfills the relevant requirements of ISO/IEC 17043. In regard to ISO 17043, it is satisfactory if z-scores are in the range [-2; +2] for each EQA sample, questionable if in the range [-3;-2 or +2; +3] and unsatisfactory if < -3 or > + 3. The ISO did not propose a bivariate analysis of z-scores, although this would be contributive to identifying any defect in accuracy or precision.
Aims: Taking an ECAT DDimer survey as an example, the present study aimed to explain the usefulness of improved bivariate z-score analysis to statistically identify any problem of accuracy: i.e. negative bias with respect to the specific peer group on an analyzer in a set of 8 devices.
Methods: The Lyon Hospitals Board’s multisite hemostasis laboratory has a set of 8 analyzers belonging to the peer group ACL TOP family IL HemosIL HS 500 in the ECAT DDimers program. In this survey, there were 2 ECAT samples measuring very close to the deep venous thrombosis (DVT) exclusion cut-off value, at respectively 440 and 580 mg/l FEU. One of the 8 analyzers showed z-scores of respectively -2.60 and -1.65 on the 2 ECAT samples. Bivariate z-score analysis, taking account of the correlation between the EQA results, was used to determine whether the device in question tended to return significantly lower DDimer rates than the others, thus being unable to rule out DVT.
Results: The elliptic acceptability limits taking account of correlation phenomena showed that the device presented no statistical bias jeopardizing patient care (figure 1).
Conclusions: The forthcoming implementation of this improved bivariate z-score analysis in ECAT reports will be of great benefit for laboratories.
[Fig 1: z-score bivariate analysis, indicating that the bivariate z-score for the device shown in the blue rectangle was acceptable]
To cite this abstract in AMA style:
Sobas F, Tsiamyrtzis P, Nougier C, Jousselme E, Geay Baillat MO, Beghin M, Meijer P. Interest of Improved Z-Score Bivariate Analysis: Application to a ECAT DDimer Survey [abstract]. Res Pract Thromb Haemost. 2020; 4 (Suppl 1). https://abstracts.isth.org/abstract/interest-of-improved-z-score-bivariate-analysis-application-to-a-ecat-ddimer-survey/. Accessed March 21, 2024.« Back to ISTH 2020 Congress
ISTH Congress Abstracts - https://abstracts.isth.org/abstract/interest-of-improved-z-score-bivariate-analysis-application-to-a-ecat-ddimer-survey/