• haemoglobin;
  • ischaemia;
  • transfusion

Background and Objectives  Red cell transfusion is commonly used in orthopaedic surgery. Evidence suggests that a restrictive transfusion strategy may be safe for most patients. However, concern has been raised over the risks of anaemia in those with ischaemic cardiac disease. Perioperative silent myocardial ischaemia (SMI) has a relatively high incidence in the elderly population undergoing elective surgery. This study used Holter monitoring to compare the effect of a restrictive and a liberal red cell transfusion strategy on the incidence of SMI in patients without signs or symptoms of ischaemic heart disease who were undergoing lower limb arthroplasty.

Materials and Methods  We performed a multicentre, controlled trial in which 260 patients undergoing elective hip and knee replacement surgery were enrolled and randomized to transfusion triggers that were either restrictive (8 g/dl) or liberal (10 g/dl). Participants were monitored with continuous ambulatory electrocardiogram (ECG) (Holter monitoring), preoperatively for 12 h and postoperatively for 72 h. The tapes were analysed for new ischaemia by technicians blinded to treatment. The total ischaemia time in minutes was divided by the recording time in hours and an ischaemic load in min/h was calculated. Haemoglobin levels were measured preoperatively, postoperatively in the recovery room, and on days one, three and five after surgery.

Results  The mean postoperative haemoglobin concentration was 9·87 g/dl in the restrictive group and 11·09 g/dl in the liberal group. In the restrictive group, 34% were transfused a total of 89 red cell units, and in the liberal group 43% were given a total of 119 red cell units. A postoperative episode of silent ischaemia was experienced by 21/109 (19%) patients in the restrictive group and by 26/109 (24%) patients in the liberal group [mean difference −4·6%; 95% confidence interval (CI): −15·5% to 6%, P = 0·41). There was no significant difference (P = 0·53) between the overall ischaemic load in the restrictive group (median 0 min/h, range 0–4·18) and the liberal group (median 0 min/h, range 0–19·48). In those patients who did experience postoperative SMI, the mean ischaemic load was 0·48 min/h in the restrictive group and 1·51 min/h in the liberal group (ratio 0·32, 95% CI: 0·14–0·76, P = 0·011). The median postoperative length of hospital stay in the restrictive group was 7·3 days [range 5–11; interquartile range (IQR) 6–8] compared with 7·5 days (range 5–13; IQR 7–8) in the liberal group. The numbers were not large enough to conclude equivalence.

Conclusions  In patients without preoperative evidence of myocardial ischaemia undergoing elective hip and knee replacement surgery, a restrictive transfusion strategy seems unlikely to be associated with an increased incidence of SMI. A proportion of these patients experience moderate SMI, regardless of the transfusion trigger. Use of a restrictive transfusion strategy did not increase length of hospital stay, and use of this strategy would lead to a significant reduction in red cell transfusion in orthopaedic surgery. Our data did not indicate any potential for harm in employing such a strategy in patients with no prior evidence of cardiac ischaemia who were undergoing elective orthopaedic surgery.