| Sumario: | Background: Severe bleeding is a leading cause of ICU admission and mortality. Fibrinogen plays a crucial role in prognosis, yet optimal thresholds and supplementation targets remain unclear. Method: Patients with major bleeding were extracted from the MIMIC-IV database. Restricted cubic splines (RCS) identified the optimal pre-treatment fibrinogen threshold, and propensity score matching adjusted for confounders. Multiple analytical methods, including multivariable regression and machine learning models, were applied. Post-treatment fibrinogen levels were stratified based on guideline recommendations, and Cox regression assessed survival outcomes. Results: Among 7,063 patients (6,666 survivors, 397 non-survivors), RCS analysis revealed a nonlinear relationship between pre-treatment fibrinogen and ICU mortality (P-non-linear < 0.001), with a threshold at 1.3 g/L. Patients with Fib > 1.3 g/L had a significant 28-day survival benefit (OR = 0.65, 95% CI: 0.48–0.87, p < 0.01). Post-treatment stratification showed that fibrinogen ≥ 1.3 g/L was associated with improved survival (p < 0.01). RCS analysis identified an optimal post-treatment target of 2.0–2.5 g/L. Conclusion: Fibrinogen levels are predictive of ICU outcomes in massive hemorrhage. A pre-treatment threshold of 1.3 g/L indicates poor prognosis, while post-treatment levels of 2.0–2.5 g/L may optimize survival.
|
|---|