The risk of donor blood transfusion was unacceptably high in the Short Duration Group

To establish whether a short duration of tourniquet application (from cement fixation to cement setting)is associated with better patient functional outcomes compared to a long duration of tourniquet application (from surgical incision to cement setting).

It is hypothesised that tourniquet application during cement fixation only (approximately of 20-30 min duration) will be associated with less pain and impairment than a longer tourniquet application (> 45 minutes).

Background:

Tourniquets are commonly used during total knee replacement (TKR) surgery to provide a bloodless field to improve visualisation [1]. The optimum timing of tourniquet release (ie prior to or after wound closure) has been the subject of many randomised controlled trials with a recent meta-analysis [1] concluding that whilst early release may increase blood loss, it may protect patients from regional complications requiring re-operation such as wound dehiscence, haematomas requiring drainage, knee stiffness and infection. Thus, there is good evidence to suggest that tourniquet use of short duration may be optimal as far as complications are concerned. Few studies have looked at the effect of tourniquet duration on functional recovery, but one [2] observed that quadriceps lag recovered more quickly if short duration tourniquet was used. There has been no research to date on using a tourniquet just during cement fixation of the prosthesis when good visualisation is key to preventing blood-cement mixing. Such an approach is important to investigate as it could further reduce the risks of tourniquet use. However, tourniquet use of such short duration may increase blood loss and increase the risk for donor blood transfusion.

Aim: To establish whether a short duration of tourniquet application (from cement fixation to cement setting)is associated with better outcomes compared to a long duration of tourniquet application (from surgical incision to cement setting).Research Design: A randomised controlled trial will be used to assess the research hypothesis. The research hypothesis is that tourniquet use during cement fixation only will be associated with better physical and functional recovery. However, it may be associated with greater need for donor blood transfusion. Methods: Consenting patients will be randomly allocated to have tourniquet use of either short or long duration in the operating theatre after the patient has been anaesthetised.The tourniquet will be applied at the time of the first skin incision (standard practice for the surgeons involved) for those allocated to the Long-Duration Group, and applied just during cement fixation of the prosthesis for the Short-Duration Group. In both groups, the tourniquet will be removed when the cement has set.

Assessments will be conducted at baseline (up to 6-weeks before surgery during a pre-admission visit), at discharge, then at 2,10, 26 and 52 weeks post surgery. The outcomes of interest include patient-centred measures (Oxford knee score -primary outcome)and other physical measures such as stair climbing test quadriceps lag, and knee range of motion. As the risk of donor blood transfusion may increase with short tourniquet duration, the use of donor blood is also a primary outcome. Reinfusion drains will be used routinely on all patients in order to reduce the need for donor blood. All enrolled patients will undergo a screening leg doppler and another at 2 weeks post surgery for detection of DVTs. Length of stay, and discharge destination (home versus rehabilitation),will also be recorded. Complication type and frequency (including evidence of prosthesis loosening in the first year) will be monitored. The study will be powered to detect a significant difference in Oxford scores. Thus, a sample size of 200 patients will provide sufficient power (80%) at 0.05 significance level to detect a 0.5SD difference in Oxford scores between the 2 treatment arms. This provides for a 20% loss to follow-up and also provides sufficient numbers to detect a large absolute difference (> 20%) in transfusion rate between the groups.

Inclusion Criteria:

• primary, unilateral knee replacement;
• capacity to understand the protocol;

Exclusion Criteria:

• inability to read English;
• current warfarin therapy (or similar);
• significant PVD, peripheral arterial disease or diabetes rendering tourniquet use undesirable;
• evidence of chronic or recent DVT in index limb;
• thigh circumference > 100 cm (approximately);
• Jehovah's Witness (ie refusal to receive donor blood)