About 81,000 total knee replacements are implanted in the United Kingdom annually. Conventionally pneumatic tourniquet and surgical drains have been used in total knee arthroplasty. However, no definite evidence or guidelines exist to support their use.

The present study aims to look for any difference in perioperative recovery and complications following use of pneumatic tourniquet and surgical drains in knee replacement surgery.

We retrospectively analysed 60 total knee arthroplasties performed by 3 different surgeons at our hospital during January 2007 to June 2010. A sample size of 20 patients in each group was evaluated. Group 1: used tourniquet and drain; Group 2: used tourniquet but no drain; Group 3: didn't use tourniquet but used drain. Data was collected and analysed using statistical means, t-test & analysis of variance.

Parameters compared were surgical (anaesthetic) time, mean postoperative haemoglobin at 24 hrs, blood transfusion rate, haematoma formation, wound infection, pain, deep vein thrombosis and infection at discharge. At 6 weeks clinical follow up letters were analysed.

Our results reveal that there is no significant difference in duration of surgery in either group. Use of tourniquets didn't improve on surgical time (123–133 minutes).

Our Local Hospital Pain Score was used (1 no pain, 2 mild, 3 moderate, 4 severe and 5 worst imaginable pain). No difference at discharge was seen in either group.

Similarly no difference was seen in either group with postoperative haematoma formation with one case (5%) in Group 1 and 3 respectively.

Although the immediate postoperative haemoglobin (Hb%) was comparable in all 3 groups, Group 3 patients received more blood transfusion compared to other groups based on clinical grounds. About 50 % of Group 3 needed blood transfusion compared to 10% in Group 1 and 5% in Group 2.

There was only one proven case of superficial infection at 6 weeks (Group 2) which responded well to antibiotics. There were 4 cases of cellulitis noted at 6 weeks (3 in Group 2 and 1 in Group 3) which responded to oral antibiotics. No deep infection was noted.

The mean hospital stay in hospital was 6.5 days in Group 1, 5.5 days in Group 2 and 8 days in Group 3. There was no difference in rate of post-operative deep vein thrombosis (5% in Group 1 and 3).

The present study with the use of tourniquet and surgical drain in total knee arthroplasty showed no significant difference in surgical time and incidence of postoperative haematoma, deep vein thrombosis. Pain score at discharge was comparable and so was the infection rate. Patients in Group 2 (used tourniquet but no surgical drain) had least hospital stay (mean 5.5 days) and least blood transfusion rate (5%). Decision to use pneumatic tourniquet & drain still lies with the operating surgeon. We recommend a randomised control study for further evidence.