A survey of blood transfusion practice in UK cardiac surgery units

Introduction

Blood is expensive and carries risks. The theoretical risk of variant Creutzfeldt-Jakob disease and the consequent introduction of leucodepletion by UK blood services has significantly increased the cost of blood to the National Health Service. This is with a probable increase in demand because of waiting lists and no concurrent increase in donations. Cardiac surgery is a major user of blood. The use of guidelines and blood conservation techniques such as perioperative cell salvage may result in a reduction in blood use [1]. We do not have guidelines in our unit and have reviewed our blood use over the past 3 years. In the absence of good evidence from which to construct guidelines, we have surveyed UK cardiac surgery units on blood use and the presence of audit and guidelines in order to form an overview of current practice.

Method

Glasgow Royal Infirmary data were analyzed on the basis of percentage of unnecessary transfusions performed. An unnecessary transfusion was defined as having occurred in patients with a discharge haemoglobin of = 11 g/dl transfused one unit of blood or with a discharge haemoglobin of = 12 g/dl transfused two or more units. The questionnaire was sent, in the form of an electronic document, to 41 cardiac units. Only paediatric units were then excluded. Units were also requested to supply data sets on 20 first time coronary artery bypass grafts (CABGs). This was the number thought to be achievable by most units.

Results

At the Glasgow Royal Infirmary, by fiscal year from April 1996 to March 1999, 79, 81 and 80% of all cardiac surgery patients received blood and 26, 22 and 26% of those transfused received unnecessary units of blood. The response to the questionnaire was 43.6% (17 out of 39) of units; 31% (12 out of 39) supplied data sets. Twelve out of 17 units have blood transfusion guidelines; these were sometimes followed in seven and usually followed in five units. Eight out of 17 units have audit, and three of these felt that this had resulted in a reduction in blood use. Seven out of 17 hospitals have guidelines but do not have ongoing audit. The percentage of CABGs transfused ranged between 20 and 95%. Units agreed that there was no evidence on which to base transfusion triggers, but most units accept haemoglobin above 8-9 g/dl during and after intensive care unit treatment. In the data sets, there was no significant difference between units in age, weight and height of patients. For haemoglobin at admission and discharge, there were significant differences between units (P = 000.1 and P = 0.00004, respectively, by analysis of variance). The average haemoglobin at admission ranged between 12.4 and 14.8 g/dl, and at discharge it ranged between 9.8 and 11.4 g/dl. Combining the 11 groups for which discharge haemoglobin was supplied (ie 220 patients), 52% of all patients received blood, of whom 25% received unnecessary (as defined above) units of blood.

Conclusions

Transfusion practice at the Glasgow Royal Infirmary needs to be improved. Despite Health Department guidance [1], there is a wide range in transfusion practices among cardiac surgery units in the UK. Blood transfusion appears to be excessive in some units. An important number of transfusions may be unnecessary. Identifying unnecessary transfusions performed may be a useful index of transfusion practice. Identification of how units achieve low transfusion rates and model guidelines by the Association of Cardiothoracic Anaesthetists may be helpful to units in which transfusion is excessive.