Optimal ROTEM FIBTEM Thresholds for the Identification of Hypofibrinogenemia in Cardiac Surgical Patients: An Analysis of the Influence of Increased Sensitivity on the Rate and Magnitude of False Positives.

OBJECTIVE: ROTEM Sigma is widely used to identify hypofibrinogenemia (eg, <1.5 g/L) in cardiac surgical patients, with a sensitivity of 0.63 to 0.73 using common FIBTEM A5 (A5) thresholds (eg, A5 <7 mm). This equates to a false negative (FN) rate (1 - sensitivity) of 0.27 to 0.37, meaning that true hypofibrinogenemia may be frequently missed. While sensitivity can be improved by choosing a higher A5 threshold, this results in increased false positives (FP), potentially leading to overtreatment. The influence of higher sensitivity on the rate and magnitude of FPs is unknown, however. This study examined the influence of choosing an A5 threshold with sensitivity ≥0.9 (ie, FN<10%) on the rate and magnitude of FP hypofibrinogenemia in cardiac surgical patients. DESIGN: Multicenter observational diagnostic accuracy study SETTING: Public and private cardiothoracic surgical centers PARTICIPANTS: 800 cardiac surgical patients INTERVENTIONS: Adjustment of the A5 threshold to achieve sensitivity ≥0.9 for identifying hypofibrinogenemia MEASUREMENTS AND MAIN RESULTS: The A5 threshold to achieve a sensitivity of ≥0.9 to detect Clauss fibrinogen level <1.5 g/L was <10 mm. This was associated with 197 FPs (25%), which had a median Clauss fibrinogen of 1.8 g/L (interquartile range, 1.6-2.0 g/L); 80% were in the range of 1.5 to 2.0 g/L. CONCLUSIONS: In this study, using an A5 with 90% sensitivity to detect Clauss fibrinogen <1.5 g/L (<10 mm) had an FP rate of 25%. However, most of the FPs (80%) were <2.0 g/L. Given the implications of missing hypofibrinogenemia versus the potential unnecessary replacement of fibrinogen mostly within 1.5 to 2.0 g/L, using a higher A5 threshold to improve sensitivity may be a pragmatic option. Nevertheless, these findings require prospective validation.