A-344 Delivering better and timely diagnosis for critical patients: POC clinical performance versus Central Laboratory assays in key cardiac markers

Abstract

Abstract Background Cardiac markers are often used in critical settings, such as emergency department and coronary units. Clinica Redsalud Santiago implemented a reference Cardiovascular center to improve patient length of stay and clinical output, supported by diverse procedures and opportune diagnosis. Our aim was to implement POC devices for measuring key Cardiac Markers. Here we validated clinically and analytically the performance of POC devices versus Central Laboratory results for Troponin I, hs Troponin I, BNP and NT-proBNP markers. Methods Forty-nine patients admitted at emergency department were included in the study. Forty samples were obtained from male patients (69.7±9.8 years) and 9 from female patients (67.8±3.9 years). Samples were drawn and processed in parallel using Triage SOB panel, hsTroponin I and NT-proBNP tests, and the other samples were analyzed in a VITROS system, using Troponin I ES, hs Troponin I and NT-proBNP. Results Clinical agreement was assessed for different methodologies, using kappa values. Almost perfect agreement was achieved for Troponin I and NT-proBNP between Triage and VITROS systems (Figure 1). For harmonization results reporting purposes, regression fits and Bland-Altman plots were analyzed for both assays (data not shown). Hs Troponin methods did not show a good linear correlation fit, but it improves significantly on quadratic regression. For NT-proBNP methods, a good linear correlation was achieved. In both comparisons, major differences were observed for high values, while those values near to cut-off or low-to-mid in pathological range were comparable. Conclusions Timely decisions are important but is difficult to find perfect solutions for end-to-end needs. We identified two key cardiac biomarkers used routinely in emergency department and coronary care that can be harmonized for clinicians, in terms of quality and traceability. Based upon results, we can expect value discrepancies between different methodologies, hence the need for creating proper guidance for supporting key decisions for clinicians.