Sections

Multi Regional Audit of Blood Component Use in Patients with Cirrhosis

Published March 2014

This project was designed to collect information on patients admitted with a diagnosis of cirrhosis with respect to their use of blood components during the course of their admission. All hospitals registered with the British Society of Gastroenterology (BSG) across the United Kingdom were invited to participate.

The BSG in conjunction with NHS Blood and Transplant (NHSBT) invited all staff on the current BSG membership lists. The project was also supported by the British Association for the Study of Liver Disease (BASL) and Regional Transfusion Committees. Interested parties registered with the BSG. They were then sent supporting documentation plus a link to an online data collection tool. The data collection tool was piloted in 3 sites prior to the main audit and modified using an iterative process. An organisational audit was also carried out to supplement the clinical information.

Executive Summary

  1. This was the first national audit of the use of blood components in patients with cirrhosis.
  2. 85 organisations/hospitals provided data on 1333 consecutive cases of cirrhosis during a 4 week period starting between February – April 2013.
  3. Local hospital audit support was highly variable, which had an impact on the number of completed cases submitted from some hospitals.
  4. The final dataset comprised 1313 cases, with 840/1313 (64%) males, mean age 58 years.
  5. The most common aetiology of cirrhosis was alcohol (70%) followed by non-alcoholic fatty liver disease (12%) and viral hepatitis (11%).
  6. 964/1313 (73%) of admissions were due to decompensated cirrhosis. 275/1313 (21%) cases had a positive septic screen
  7. Case fatality during follow-up was 128/1313 (10%) overall, with decompensated cirrhosis reported as the most frequent cause of death 52/128 (41%).
  8. There were 35/1313 (3%) cases of reported thrombotic events of which 29/1313 (2%) were venous thromboses and 6/1313 (<1%) were arterial thromboses.
  9. 391/1313 (30%) patients were transfused at least one blood component.
  10. For 153/391 (39%) cases the main transfusion indication was prophylaxis (not for bleeding) and for 238/391 (61%) cases the main indication was treatment of bleeding.
  11. For the 238 cases transfused for bleeding, gastro-intestinal bleeding was the most common cause 192/238 (81%). 150/238 (63%) received red cell transfusions alone, which were administered at variable haemoglobin concentrations. In patients with gastrointestinal bleeding who received red blood cells, the pre-transfusion threshold was greater than 80g/L prior to red cell transfusion in 54/220 (25%).
  12. For the 153 cases receiving transfusions for prophylaxis, 94/153 (61%) received transfusions when no procedure was planned. The majority of these were red cell transfusions for the treatment of anaemia 85/94 (90%) but a small number received Fresh Frozen Plasma (FFP) 11/94 (12%) or platelets 5/94 (5%).
  13. For the 59/153 (39%) cases receiving transfusion for prophylaxis prior to interventions, the more common procedures requiring cover by transfusion were paracentesis, surgery and central/femoral line insertion. FFP was the most common single blood component transfused in 34/59 (58%) followed by platelets 25/59 (42%).
  14. 4/72 (6%) hospitals reported having guidelines for patients with liver disease.
  15. In 48/185 (26%) of patients transfused with red cells for gastrointestinal bleeding, the pre-transfusion haemoglobin was greater than 80g/L.
  16. In 103/185 (56%) of patients transfused with red cells for gastrointestinal bleeding, the pre-transfusion haemoglobin was greater than 70g/L.
  17. In 81/101 (80%) of patients transfused with red cells prophylactically, the pre- transfusion haemoglobin was less than 80g/L.
  18. In 13/16 (81%) of patients transfused with FFP prophylactically before a moderate/high- risk procedure, the pre-transfusion INR was greater than 1.5.
  19. In 12/18 (67%) of patients transfused with FFP prophylactically before a low-risk procedure, the pre-transfusion INR was greater than 2.
  20. In 16/25 (64%) of patients transfused with platelets prophylactically before a procedure, the pre-transfusion platelet count was less than 50x109/L