Abstract
Purpose Blood culture (BC) is the gold standard for diagnosing blood stream infections (BSI) but is limited by long turnaround times (TAT) and low detection rate. The T2 Magnetic Resonance method (T2MR) offers a rapid, culture-independent alternative. The objective of this study was to compare the performance of the T2Bacteria assay to BCs in a real-world setting. Methods Retrospective comparative study consisting of T2Bacteria samples and BCs sampled within 72 h from the T2Bacteria sample. The primary outcome was detections by BC and T2Bacteria, respectively. The secondary outcome was difference in TAT. Results In total, 640 episodes were included, consisting of 640 T2Bacteria samples and 2,117 BCs. A median of three BCs was collected for each T2Bacteria sample. Overall positivity was 101 (15.8%) by either method. In 29 (28.7%) episodes, both T2Bacteria and BC were concordantly positive. In discordant episodes, 46/101 (45.5%) episodes were T2Bacteria positive/BC negative and 26/101 (25.7%) were T2Bacteria negative/BC positive (McNemar χ 2 , p < 0,05). In T2Bacteria positive/BC negative episodes, eight had growth of the same microorganism in a non-BC culture. Median (IQR) TAT for BC was 35 h and 30 min (25 h 50 min − 45 h 24 min), compared to 21 h and 3 min (17 h 6 min − 27 h 30 m) for T2Bacteria ( p < 0.001), with longer delays for samplings occurring outside work hours. Conclusions The study highlights a high discordance between T2Bacteria and BC and suggests complementary roles of the methods in BSI diagnostics. Furthermore, it is crucial to improve TAT by reducing preanalytical delays.