A Comparison Study: Possible Bias in Troponin I Measurement Obtained with a Point of Care Testing and a Central Laboratory Analyzers Employing Different Biological Matrices and Anticoagulants.

BACKGROUND

high-sensitive cardiac TroponinI (hs-cTnI) is widely used for diagnosis of acute coronary syndromes. The latest recommendation for hs-cTnI determination is the protocol 0-1 h finalized to improve the rule out accuracy of the test. A Point of Care Testing able to guarantee these performances could be very useful due to reducing the turnaround time and ruling out patients suspected of ACS, especially by using biological matrices that are not required for centrifuge. The aim of our work is to compare the results for hs-cTnI obtained using different biological matrices and anticoagulants, obtained between Atellica VTLi hs-cTnI POCT and Access AccuTnI+3 DxI800 performances, in order to establish a possible bias derived directly from these pre-analytical conditions.

METHODS

Li-heparinized pool samples were primary employ for hs-cTnI with Atellica VTLi as whole blood, then centrifuged and tested on Atellica VTLi and DxI800. KEDTA pool samples were centrifuged and measured on DxI800 too. A comparison of methods was performed according to CLSI_EP-09A2 protocol. Constant and proportional errors were investigated with Deming regression. Bias between methods was evaluated with the Bland Altman test.

RESULTS

comparing whole blood lithium heparin results obtained with Atellica versus lithium heparin and KEDTA plasma tested on DxI 800, the Deming regression revealed a proportional error, whereas in both cases Bland Altman highlighted a minimal underestimation. A similar performance was revealed when considering plasma lithium heparin tested on Atellica versus lithium heparin and KEDTA plasma obtained with DxI800, confirming the same underestimation. Considering values close to the cut off, no significant differences were found.

CONCLUSIONS

in the laboratory, the estimation of the bias of two different analyzers is pivotal. Once more this is crucial when different biological matrices and anticoagulants are employed for the analysis. Our study demonstrates that no significant differences among the two matrices are present when comparing Atellica and DxI800 performances.