Introduction: There is increasing awareness and concern among the medical profession, general public and media about the various complications of homologous blood transfusion.

Primary arthroplasty of either the hip or knee has an estimated total bleeding of 1.51 (Lotke et al 1991), commonly resulting in transfusion. In knee arthroplasty, performed with the use of a tourniquet, almost all the bleeding occurs postoperatively. Several studies have shown that salvage of blood after the operation and reinfusion can reduce the need for homologous transfusion (Majkowski RS et al, Newman JH et al).

We studied prospectively 100 consecutive patients for knee replacement to compare post-operative transfusion requirements in blood salvage and reinfusion group and a no drain technique.

Methods: After written consent 100 consecutive patients for knee replacement surgery were randomly allocated by computer generated numbers to either the no drain group (n=50) or the reinfusion group (n=50). The patients were anaesthetised as deemed appropriate by their anaesthetist. All the patients were operated on using a tourniquet. At the end of surgery the wound was closed with or without a deep drain. The drain was attached to the Constavac CBC II closed suction system. Preoperative haemoglobin and haematocrit values were recorded. Homologous blood was transfused to the patients as per the standard protocol depending on the haematocrit and or haemoglobin at 4 & 8 hours post-operatively, as well as days 1,2 & 3. In the reinfusion group blood collected in the reservoir was transferred to the blood bag and reinfused at 5 hours or 500 mls depending which was earlier. At 8 hours the reservoir blood was collected and reinfused, no more blood was reinfused as per the manufacturer recommendations. Cardiovascular stability was assessed by hourly blood pressure and heart rate during the first 24 hours and twice daily thereafter. Blood loss was assesses by measuring the drain loss, assessing the wound ooze serial haematocrits and total transfusion requirements.

Results: 50 patients were completed in each group. The mean preoperative haemoglobin in the drain group was 12.8 g/dl and in the no drain group it was 12.9 g/dl. No difference in predisposing factors for bleeding was recorded in the two groups. The mean volume of blood collected in the drain group was 1008mls and the mean volume of autologous transfused was 864 mls. 14% of patients in both groups had no requirement for homologous blood. There was no significant difference in the homologous blood transfusion in the two groups over the study period (up to day 3 post op), the no drain group requiring and average of 2.1 units of packed cells the reinfusion group requiring an average of 1.8 units of packed cells in total. The homologous blood requirements in the drain group was significantly reduced on day 2 in comparison to the no drain group, but as noted there was no overall reduction in homologous requirements. Cardiovascularly 16% of patients with a drain had at least one episode of hypotension (Bp < 90 systolic) compared with 20% of patients in the no drain group. The group without a drain had a significant increase in wound ooze (70% compared with 44%)

Conclusion: Post-operative blood salvaging and autologous transfusion following primary knee replacement under tourniquet in this study did not significantly reduce the patient’s requirement for homologous blood transfusion in the first 3 days post operatively. There was a significant increase in wound ooze in the no drain group but as shown this has not result in an increase in cardiovascular instability or an increase in transfusion requirements.