Analytical performance comparison of Siemens Atellica CI1900 and Roche Cobas Pure integrated analyzers across spectrophotometric, ion-selective electrode and immunoassay parameters

Abstract

Background: Harmonization of laboratory measurements across different analytical platforms is essential to ensure accurate and clinically reliable results. Variations between analytical systems may lead to discrepancies that could affect clinical decision-making and patient management. Objective: This study aimed to evaluate the analytical agreement between the Siemens Atellica CI1900 and Roche Cobas Pure integrated analyzers across different analytical methodologies, including spectrophotometry, ion-selective electrode (ISE), and immunoassay techniques. Materials and methods: Fourteen analytes were included in the comparison: spectrophotometric assays (glucose, creatinine, alanine aminotransferase [ALT], calcium [Ca], total bilirubin [TB], and direct bilirubin [DB]), ISE assays (sodium [Na], potassium [K], chloride [Cl]), and immunoassays (thyroid-stimulating hormone [TSH], free thyroxine [fT4], free triiodothyronine [fT3], human chorionic gonadotropin [hCG], and N-terminal pro-brain natriuretic peptide [NT-proBNP]). Analytical verification was first performed according to Clinical and Laboratory Standards Institute (CLSI) EP15-A3 guidelines. Subsequently, method comparison between the two analyzers was conducted following CLSI EP09-A3 recommendations using 464 serum samples. Agreement between the platforms was evaluated using Passing-Bablok regression, linear regression analysis and Bland-Altman plots. Results: Spectrophotometric assays demonstrated excellent correlations for ALT, total bilirubin, glucose and creatinine (R-2 >= 0.95), while calcium and direct bilirubin showed strong correlations (0.95 >= R-2 > 0.7) between the two systems. Among the ISE parameters, sodium and potassium showed strong correlations, whereas chloride demonstrated moderate agreement (0.7 >= R-2 > 0.3). For immunoassays, TSH and NT-proBNP demonstrated excellent correlations, fT4 and hCG showed strong correlations, and fT3 demonstrated moderate agreement. Bland-Altman analysis demonstrated generally small mean biases, and the corresponding 95% limits of agreement supported the overall analytical agreement between the two analytical platforms. Conclusion: Overall, the Siemens Atellica CI1900 and Roche Cobas Pure analyzers demonstrated good analytical agreement across the evaluated measurement principles. These findings support the comparable analytical performance of the two platforms for routine clinical laboratory testing. Nevertheless, laboratory-specific method verification remains essential before implementing different analytical systems in routine clinical practice.