Clinical Simulation Model of Fibrinogen Decline During Hemorrhage in Major Noncardiac Surgery

Abstract

Background: Monitoring of decrease in fibrinogen levels with surgical blood loss is crucial for timely transfusion of fresh frozen plasma (FFP) to avoid coagulopathic bleeding. Here, we validated a simulation model to predict hemorrhagic reductions in fibrinogen levels during major noncardiac surgery.

Methods: We retrospectively performed exponential regression analysis of intraoperative blood loss and fibrinogen levels to develop a simulation model in the initial 50 patients and applied the model to another 59 patients to compare the measured and simulated fibrinogen levels. We examined the relationship between FFP transfusion and the measured fibrinogen levels or blood loss. The fibrinogen trigger level of FFP transfusion was below 130 mg/dL, although the decision of a perioperative blood transfusion was at the discretion of the anesthesiologists and surgeons.

Results: Application of the simulation model based on the initial 50 patients to another 59 patients showed no difference between the measured and estimated fibrinogen levels (189 ± 61 versus 186 ± 62, P = 0.60, mean difference: -2.28, limits of agreement: -69.42 to 64.84). The estimated fibrinogen level (mg/dL) = preoperative fibrinogen × exp (-1.90 × [blood loss/estimated circulation volume]), in which the estimated circulation volume = (70 [mL/kg] × body weight [kg]). FFP transfusion was significantly related to the measured fibrinogen level (cutoff: 145; 95% confidence intervals: 124-168; P = 0.0003) but not blood loss (P = 0.12).

Conclusions: Fibrinogen level simulation predicted a hemorrhagic fibrinogen decline, thereby guiding FFP transfusion during active surgical bleeding. Further studies on the usefulness of fibrinogen level simulation are warranted.

Keywords: Bleeding; Fibrinogen; Fresh frozen plasma; Noncardiac surgery; Simulation model.