As blood transfusion is associated with various risks, a prospective study was carried out to see if it was possible to predict patients more likely to require transfusion following TKR.

Data was collected prospectively on 1532 patients undergoing primary TKR between 1998 and 2006. This was collected at a preadmission clinic and various demographics were measured including haemoglobin, BMI, and a knee score. All patients had a tourniquet and the same approach. All received a LMWH until discharge. Patients with a post op haemoglobin less than 8.5 g/dl were transfused as were those less than 10 g/dl who were symptomatic as per unit protocol.

Each of the predictive factors was tested for significance using t-tests and chi-squared tests as appropriate. Multiple logistic regression was used to test for the independent predictive of factors after adjusting for one another.

Results show transfusion is more likely if the patient was older, female, short light or thin. Also those undergoing a lateral release or a bilateral procedure, having a low pre-op haemoglobin or a large post-op drop were more likely to be transfused. There was also a 2 fold difference between surgeons.

After regression analysis 4 important factors were identified. These were a bilateral procedure, low pre-op haemoglobin, a low BMI or having a post-op drop greater than 3g/dl.

Following this all patients with pre-op haemoglobin less than 11g/dl are postponed and investigated and treated as required. For those with the above predictive factors, measures can be taken to try and reduce the rate of transfusion such as pre-donation, cell salvage or tran-sexamic acid.