Total knee arthroplasty (TKA) is one of the most successful procedures in orthopaedic surgery. Nevertheless, significant postoperative blood loss and requirement of blood transfusion are still problematic. Total blood loss in TKA can be divided into visible and invisible blood loss. Visible blood loss (VBL) means blood loss from the surgical field and wound drainage while invisible blood loss (IBL) means residual blood in the knee, extravasation into the tissues and loss due to haemolysis.

In usual practice, TKA is performed with a bloodless field using a tourniquet. Thus intraoperative blood loss can be negligible and postoperative drainage is only considered as VBLvisible. In terms of IBL, Sehat et al. found that TKA carried a substantial IBL. Their results shown the mean IBL was 765 ml or 49% of the mean total blood loss after TKA. Therefore, the true total blood loss was underestimated if not takes IBL into account.

Modified Robert Jones bandage (MRJB) is a bulky compressive dressing that often used in orthopaedic practice. Various techniques of application have been proposed.

From the previous study, MRJB could make and maintain the anterolateral muscle compartment pressure for at least 24 hours after TKA. Therefore, theoretically, this bandage can cause the tamponade effect that helps to reduce tissue edema and postoperative bleeding especially IBL after TKA. However this potential benefit of MRJB is unclear and the use of this bandage after TKA is still controversy in clinical practice.

In preoperative period, all eligible participants are admitted to the hospital for undergoing TKA. The detailed protocol, including how to rate visual analog pain score, times to take the blood for checking hematocrit levels, criteria for blood transfusion, time to measure the knee and thigh circumferences and plan of ambulation program, is informed again. The routine preoperative preparation is done. The participant is asked for answering preoperative Oxford Knee Score questionnaire. Participants' baseline characteristics are also recorded.

During intraoperative period, all participants receive the same anesthetic method (spinal anesthesia without morphine and peripheral nerve block). Two experienced surgeons (CK and NR) perform all procedure with the same surgical technique. A 750 mg intravenous tranexamic acid is administrated before inflating tourniquet and 3 hours after the operation. A tourniquet pressure of 300 mmHg is used and inflated before skin incision. Pre-incisional local injection with 10 ml of 1% lidocaine with adrenaline is done. A mini-medial parapatellar approach is performed. Cemented TKA prostheses (Nexgen LPS-Flex, Zimmer, Warsaw, Indiana) are implanted in all participants without patellar resurfacing. Periarticular analgesic injection is done with 20 ml of 0.5% bupivacaine, 20 ml of normal saline, 30 mg of ketorolac and 1 ml of 1% lidocaine with adrenaline. A number-10-gauge drain is placed intra-articularly and connected to the Ultravak pressure drainage bottle (Poly Medicure Limited, India). Extensor mechanisms are repaired and the wound closure is performed in the routine fashion.

At the end of operation, the randomized sequence is opened by a scrub nurse. Then, the participants are randomly assigned into two groups; Group 1 or MRJB group and Group 2 or non-compressive dressing (NCD) group. A tourniquet pressure is deflated after the dressing is applied.

In postoperative period, MRJB is left in place for 24 hours and then changed to NCD. All participants obtain the three-hour interval drain clamping protocol. After tourniquet release, the drain is clamped for three hours, released for three hours, re-clamped for three hours, and then the clamp is run continuously. Mechanical prophylaxis of deep vein thrombosis by ankle pumping exercise and ambulation on the bed is started as soon as possible. After 24 hours postoperatively, range of motion exercise and out-of-bed ambulation program are started. Ice packs are placed around the knee for at least 8 hours per day. The drain is removed at 48 hours after operation.

For operational definition, MRJB is defined as a three-layers of thick cotton wool and two-layers of elastic bandages. The wool layers are put on firmly and overlapped the previous one by half at each turn. The elastic layers were pulled snugly with more tension distally than proximally. Before wrapping in each turn, the elastic bandage was stretched approximately 2 and 1.5 inches at below and above tibial tuberosity level, respectively. The whole bandage attains a thickness of about two inches and extends above the ankle joint to six inches above the knee joint. Before applying this bandage, the sterile gauze pads were placed over the wound and followed by WebrilTM padding (Covidien, Mansfield, MA, US). While NCD is made by placing the sterile gauze pads over the wound and covering with the hypoallergenic self-adhesive, non-woven fabric tape.

All patients receive the same postoperative pain managements, including (1) parecoxib 40 mg administrated as intravenous bolus every 12 hours for 2 days, then diclofenac 25 mg orally every 8 hours until discharge, (2) acetaminophen 300 mg with codeine 15 mg orally every 8 hours until discharge, (3) morphine 2 mg intravenously pro re nata (prn) for pain every 2 hours as a rescue drug.

During the first and second 24 hours after surgery, the maximum of visual analog pain scores at rest and during the ambulation are assessed. The circumferences of knee and thigh at postoperative 24 and 48 hours are measured by single blinded assessor.

To evaluate of blood loss and determine criteria for blood transfusion, hematocrit levels are measures at 24 and 48 hours after surgery. The patients will receive a transfusion of one unit of packed red cells, if their hematocrit levels are less than 30% or if the compromised clinical criteria (such as hypotension, tachycardia, symptoms of anemia that are relative to the preoperative medical condition) necessitated blood transfusion. If hematocrit levels are less than 24%, two units of packed red cells are transfused. The hematocrit levels are reevaluated at 4 hours after the end of transfusion and blood transfusion is considered again using the same criteria. Drainage blood loss and amount of blood transfusion are recorded at 24 and 48 hours postoperatively.

At discharge, range of motion is measured again. Clinical venous thromboembolism, peroneal nerve palsy, infection and wound complications are also assessed.

During follow-up period, range of motion, clinical venous thromboembolism, infection, wound condition and other complications are reassessed at 2 and 6 weeks after TKA. The functional outcomes, including Oxford Knee Score and time up and go test, are evaluated at 6 weeks postoperatively.

Primary outcome in this study is IBL. It means invisible red blood cell (RBC) volume loss that can be calculated according to the following steps; first, the patient's blood volume (PBV) is calculated using the Nadler formula.

PBV = k1 x height (m)3 + k2 x weight (kg) + k3 When k1= 0.3669, k2 = 0.03219, k3 = 0.6041 for men; k1= 0.3561, k2 = 0.03308, k3 = 0.18331 for women Then, multiplying the PBV by the hematocrit (Hct) will give the total RBC volume. The change in RBC volume can be calculated from the change in Hct [20]. In this study, total RBC volume loss at postoperative 48 hours is calculated using the following formula; Total RBC volume loss = PBV x (Preoperative Hct - Postoperative Hct at 48 hours) At 24 hours after surgery, the volume of drained blood is measured and then multiplying by the preoperative Hct will give the drainage RBC blood volume at 24 hours. During 24-48 hours after surgery, the volume of drained blood is measured and then multiplying by the postoperative Hct at 24 hours will give the drainage RBC blood volume at 24-48 hours. The summation of drainage RBC blood volume at 24 hours and at 24-48 hours is total drained RBC volume. If the patient receives blood transfusion, the volume of packed red cell is also recorded. Finally, invisible blood loss is calculated using the following formula; Invisible blood loss = Total RBC volume loss - Total drained RBC volume + Transfused blood volume

Inclusion Criteria:

• Patients who diagnosed primary osteoarthritis of knee and scheduled for primary unilateral total knee arthroplasty
• Patients aged more than 18 years

Exclusion Criteria:

• Patients who have a history of coagulopathy
• Patients who receive antiplatelet drug in the previous week
• Patients who treated other conditions with anticoagulant drug
• Patients who have previous history of thromboembolic event
• Patients who have vascular compromise of the operated limb
• Patients who have chronic kidney disease or liver cirrhosis
• Patients who refuse to participate the study
• Patients who have allergy to tranexamic acid, sulfa or morphine
• Patients who have blood loss per wound after surgery
• Patients who can not receive spinal anesthesia and peripheral nerve block