Fresh frozen plasma transfusion in acute variceal haemorrhage: Results from a multicentre cohort study

• Published in: Liver international 2021-08, Vol.41 (8), p.1901-1908
• Main Authors: Mohanty, Arpan, Kapuria, Devika, Canakis, Andrew, Lin, Honghuang, Amat, Maelys J., Rangel Paniz, Graziella, Placone, Nicholas T., Thomasson, Reggie, Roy, Hemant, Chak, Eric, Baffy, Gyorgy, Curry, Michael P., Laine, Loren, Rustagi, Tarun
• Format: Article
• Language: English
• Subjects: acute variceal haemorrhage; Best practice; Bleeding; Clinical outcomes; Cohort analysis; Confidence intervals; fresh frozen plasma; gastrointestinal bleeding; Hemorrhage; Length of stay; Liver diseases; Mortality; Patients; portal hypertension; Statistical analysis; Transfusion
• ISSN: 1478-3223; 1478-3231
• DOI: 10.1111/liv.14936

Background Fresh frozen plasma (FFP) transfusion is often used in the management of acute variceal haemorrhage (AVH) despite best practice advice suggesting otherwise. Objective We investigated if FFP transfusion affects clinical outcomes in AVH. Design, setting and patients We performed a retrospective cohort study of 244 consecutive, eligible patients admitted to five tertiary health care centres between 2013 and 2018 with AVH. Main outcome measurements Multivariable regression analyses were used to study the association of FFP transfusion with mortality at 42 days (primary outcome) and failure to control bleeding at 5 days and length of stay (secondary outcomes). Results Patients who received FFP transfusion (n = 100) had higher mean Model for End Stage Liver Disease (MELD) score and more severe variceal bleeding than those who did not received FFP transfusion (n = 144). Multivariable analysis showed that FFP transfusion was associated with increased odds of mortality at 42 days (odds ratio [OR] 9.41, 95% confidence interval [CI] 3.71‐23.90). FFP transfusion was also associated with failure to control bleeding at 5 days (OR 3.87, 95% CI 1.28‐11.70) and length of stay >7 days (adjusted OR 1.88, 95% CI 1.03‐3.42). The independent association of FFP transfusion with mortality at 42 days persisted when the cohort was restricted to high‐risk patients and in patients without active bleeding. Limitations and conclusions Fresh frozen plasma transfusion in AVH is independently associated with poor clinical outcomes. As this an observational study, there may be residual bias due to confounding; however, we demonstrate no benefit and potential harm with FFP transfusions in AVH.