Review question
Human plasma, a blood component, is often transfused to people before undergoing surgery or other procedures (such as inserting a chest drainage tube) when blood tests show that their blood may not clot adequately. Aims of this review were to assess how effective this practice is in reducing subsequent bleeding or need for blood transfusion, and whether this approach increases risk of death or other significantly harmful effects. The review excluded people with inherited bleeding disorders and those taking medication that reduces their blood's clotting ability.
Background
Human plasma is obtained from blood donors. It contains many factors that help blood to clot. Although plasma may be transfused to people based on blood tests suggesting that their blood may not clot adequately, these tests have limitations. A person's blood may clot adequately despite abnormal test results. Abnormal test results also do not clearly predict those people who will go on to bleed. Furthermore, plasma transfusion corrects abnormal blood tests to varying degrees.
Plasma is also expensive, and some countries have issues with blood product shortages, donor reliability, and incomplete screening for infections that could be transmitted through blood product transfusion. Given the potential for life-threatening complications from plasma transfusion, its use in this setting carries risk of harm without clear evidence of benefit.
Study characteristics
We included five trials which were all conducted in high-income countries.
Our search is current up until 28 January 2019. One trial compared plasma transfusion with no transfusion given. Another trial compared plasma or platelet transfusion or both with neither plasma nor platelet transfusion given. One trial compared plasma transfusion with alternative products given to help blood clot. Another trial compared different blood tests to trigger a plasma transfusion, and still another trial compared different transfusion triggers using the same blood test.
Four trials involved adult participants over 18 years old, and the fifth trial did not specify age of participants. In four trials, participants underwent bedside procedures. Only one trial involved some participants undergoing major surgical operations. Two trials included only participants in intensive care, and two trials included only participants with liver disease.
One trial recruited only two participants. Therefore review results include the remaining four trials, incorporating 234 participants. Three further trials are ongoing.
Key results
When plasma transfusion was compared with no transfusion given, we are very uncertain whether there was a difference in major bleeding, number of blood transfusions per participant, or harmful effects from the transfusion (1 trial; very low-quality evidence). When plasma or platelet transfusion or both were compared with neither plasma nor platelet transfusion, we are very uncertain whether there was a difference in mortality within 30 days, or in the number of individuals requiring a transfusion (1 trial; very low-quality evidence).
When plasma transfusion was compared with other haemostatic agents, we are very uncertain whether there was a difference in major bleeding or in harmful effects from the transfusion (1 trial; very low-quality evidence).
When different triggers for plasma transfusion were compared (1 trial; 60 participants), we are very uncertain whether there was a difference in major bleeding or in harmful effects from the transfusion due to very low-quality evidence for these outcomes. The number of people requiring blood products may have been reduced overall, although this is based on low-quality evidence.
No trials reported procedure-related harmful events or quality of life as an outcome.
Quality of the evidence
The overall quality of the evidence was predominantly very low over a range of clinically important outcomes due to combinations of issues within the studies, such as potential for bias, limited clinical settings, and imprecise estimates of intervention effects.
Authors' conclusions
We are very uncertain of the effectiveness and safety of the use of plasma in non-cardiac operations or invasive procedures due to very low-quality evidence. Furthermore, as trials do not cover a wide range of surgical contexts, our confidence in applying study results to the wider surgical setting is limited. Overall limited evidence for the utility of plasma transfused to people within this context is of insufficient quality to support or oppose its use.
Review findings show uncertainty for the utility and safety of prophylactic FFP use. This is due to predominantly very low-quality evidence that is available for its use over a range of clinically important outcomes, together with lack of confidence in the wider applicability of study findings, given the paucity or absence of study data in settings such as major body cavity surgery, extensive soft tissue surgery, orthopaedic surgery, or neurosurgery. Therefore, from the limited RCT evidence, we can neither support nor oppose the use of prophylactic FFP in clinical practice.
In the absence of bleeding, plasma is commonly transfused to people prophylactically to prevent bleeding. In this context, it is transfused before operative or invasive procedures (such as liver biopsy or chest drainage tube insertion) in those considered at increased risk of bleeding, typically defined by abnormalities of laboratory tests of coagulation. As plasma contains procoagulant factors, plasma transfusion may reduce perioperative bleeding risk. This outcome has clinical importance given that perioperative bleeding and blood transfusion have been associated with increased morbidity and mortality. Plasma is expensive, and some countries have experienced issues with blood product shortages, donor pool reliability, and incomplete screening for transmissible infections. Thus, although the benefit of prophylactic plasma transfusion has not been well established, plasma transfusion does carry potentially life-threatening risks.
To determine the clinical effectiveness and safety of prophylactic plasma transfusion for people with coagulation test abnormalities (in the absence of inherited bleeding disorders or use of anticoagulant medication) requiring non-cardiac surgery or invasive procedures.
We searched for randomised controlled trials (RCTs), without language or publication status restrictions in: Cochrane Central Register of Controlled Trials (CENTRAL; 2017 Issue 7); Ovid MEDLINE (from 1946); Ovid Embase (from 1974); Cumulative Index to Nursing and Allied Health Literature (CINAHL; EBSCOHost) (from 1937); PubMed (e-publications and in-process citations ahead of print only); Transfusion Evidence Library (from 1950); Latin American Caribbean Health Sciences Literature (LILACS) (from 1982); Web of Science: Conference Proceedings Citation Index-Science (CPCI-S) (Thomson Reuters, from 1990); ClinicalTrials.gov; and World Health Organization (WHO) International Clinical Trials Registry Search Platform (ICTRP) to 28 January 2019.
We included RCTs comparing: prophylactic plasma transfusion to placebo, intravenous fluid, or no intervention; prophylactic plasma transfusion to alternative pro-haemostatic agents; or different haemostatic thresholds for prophylactic plasma transfusion. We included participants of any age, and we excluded trials incorporating individuals with previous active bleeding, with inherited bleeding disorders, or taking anticoagulant medication before enrolment.
We used standard methodological procedures expected by Cochrane.
We included five trials in this review, all were conducted in high-income countries. Three additional trials are ongoing.
One trial compared fresh frozen plasma (FFP) transfusion with no transfusion given. One trial compared FFP or platelet transfusion or both with neither FFP nor platelet transfusion given. One trial compared FFP transfusion with administration of alternative pro-haemostatic agents (factors II, IX, and X followed by VII). One trial compared the use of different transfusion triggers using the international normalised ratio measurement. One trial compared the use of a thromboelastographic-guided transfusion trigger using standard laboratory measurements of coagulation.
Four trials enrolled only adults, whereas the fifth trial did not specify participant age. Four trials included only minor procedures that could be performed by the bedside. Only one trial included some participants undergoing major surgical operations. Two trials included only participants in intensive care. Two trials included only participants with liver disease.
Three trials did not recruit sufficient participants to meet their pre-calculated sample size. Overall, the quality of evidence was low to very low across different outcomes according to GRADE methodology, due to risk of bias, indirectness, and imprecision.
One trial was stopped after recruiting two participants, therefore this review's findings are based on the remaining four trials (234 participants).
When plasma transfusion was compared with no transfusion given, we are very uncertain whether there was a difference in 30-day mortality (1 trial comparing FFP or platelet transfusion or both with neither FFP nor platelet transfusion, 72 participants; risk ratio (RR) 0.38, 95% confidence interval (CI) 0.13 to 1.10; very low-quality evidence).
We are very uncertain whether there was a difference in major bleeding within 24 hours (1 trial comparing FFP transfusion vs no transfusion, 76 participants; RR 0.33, 95% CI 0.01 to 7.93; very low-quality evidence; 1 trial comparing FFP or platelet transfusion or both with neither FFP nor platelet transfusion, 72 participants; RR 1.59, 95% CI 0.28 to 8.93; very low-quality evidence).
We are very uncertain whether there was a difference in the number of blood product transfusions per person (1 trial, 76 participants; study authors reported no difference; very low-quality evidence) or in the number of people requiring transfusion (1 trial comparing FFP or platelet transfusion or both with neither FFP nor platelet transfusion, 72 participants; study authors reported no blood transfusion given; very low-quality evidence) or in the risk of transfusion-related adverse events (acute lung injury) (1 trial, 76 participants; study authors reported no difference; very low-quality evidence).
When plasma transfusion was compared with other pro-haemostatic agents, we are very uncertain whether there was a difference in major bleeding (1 trial; 21 participants; no events; very low-quality evidence) or in transfusion-related adverse events (febrile or allergic reactions) (1 trial, 21 participants; RR 9.82, 95% CI 0.59 to 162.24; very low-quality evidence).
When different triggers for FFP transfusion were compared, the number of people requiring transfusion may have been reduced (for overall blood products) when a thromboelastographic-guided transfusion trigger was compared with standard laboratory tests (1 trial, 60 participants; RR 0.18, 95% CI 0.08 to 0.39; low-quality evidence). We are very uncertain whether there was a difference in major bleeding (1 trial, 60 participants; RR 0.33, 95% CI 0.01 to 7.87; very low-quality evidence) or in transfusion-related adverse events (allergic reactions) (1 trial; 60 participants; RR 0.33, 95% CI 0.01 to 7.87; very low-quality evidence).
Only one trial reported 30-day mortality. No trials reported procedure-related harmful events (excluding bleeding) or quality of life.