Platelet transfusion therapy has played a major role in the supportive care of patients with acute leukaemia for about 30 years. However, debate continues about the value of prophylactic platelet transfusions and the optimal threshold of platelet count for their use. There are accumulating data to suggest that the previously commonly used threshold level of 20 × 109/l may be safely reduced without increasing the risk of bleeding ( Gmur et al, 1991 ; Rebulla et al, 1991; Wandt et al, 1996 ; Heckman et al, 1997 ), and the recent Consensus Conference on Platelet Transfusion has accepted that a platelet threshold of 10 × 109/l is as safe as higher levels for treating patients without additional risk factors for haemorrhage ( Royal College of Physicians of Edinburgh Consensus Conference on Platelet Transfusion, 1998). Use of a lower threshold has been associated with a considerable decrease in the use of platelet concentrates, and has the potential to reduce complications of platelet transfusion therapy such as transmission of infection. However, the extent to which hospitals in the U.K. have changed their policies for prophylactic platelet transfusions is not known.
In the planning of a proposal to study different regimens of prophylactic platelet transfusion support in a future MRC trial of AML in the elderly, hospitals participating in MRC leukaemia trials were sent a questionnaire in 1996 asking for details about their threshold for prophylactic platelet transfusions and if they used different thresholds for young and elderly patients.
There were responses from 99 hospitals. The threshold for prophylactic platelet transfusions was < 5 × 109/l in four (4%) hospitals, < 10 × 109/l in 46 (46%), < 15 × 109/l in 21 (21%), < 20 × 109/l in 25 (25%) and prophylactic transfusions were not used in three (3%) hospitals (see Fig 1). 3/4 (75%) hospitals with a threshold of < 5 × 109/l, 18/46 (39%) hospitals with a threshold of < 10 × 109/l, and 2/21 (10%) with a threshold of < 15 × 109/l increased the threshold in the presence of fever, sepsis, coagulopathy or recent bleeding. Only three hospitals used a different threshold for young and elderly patients.
This survey showed that there was considerable variation in platelet transfusion practice amongst hospitals participating in MRC leukaemia trials. The majority of hospitals had already switched to a threshold of < 10 × 109/l before the recent Consensus Conference on Platelet Transfusion, perhaps because this had already been suggested in BCSH guidelines on platelet transfusions ( BCSH, 1992). There may be scope for a further reduction in the threshold below 10 × 109/l in the future, but more reliable methods for estimating very low platelet counts are required. It may also be worthwhile to consider alternative methods for providing platelet supportive care, such as increasing the dose of platelets to extend the interval between transfusions and to reduce donor exposure, and exploring the clinical effectiveness of platelet substitutes.