Knee Surg Sports Traumatol Arthrosc DOI 10.1007/s00167-014-2983-8
KNEE
Knee flexion after total knee arthroplasty reduces blood loss Alfredo Schiavone Panni • Simone Cerciello Michele Vasso • Chiara Del Regno
•
Received: 30 May 2013 / Accepted: 1 April 2014 Ó Springer-Verlag Berlin Heidelberg 2014
Abstract Purpose Extensive blood loss after total knee arthroplasty (TKA) may be a potential problem since it leads to anaemia, increased need for transfusion and prolonged hospitalization. Aim of this study was to investigate the effects of postoperative knee flexion after TKA on blood loss and the need for transfusion. Methods One hundred consecutive patients undergoing primary TKA from 2012 to 2013 were randomizely divided into two groups. In one group, the knee was extended for the first 6 h after surgery, whereas in the other was flexed at 90° for the same time. Two doses of endovenous tranexamic acid were administered in all subjects. Patients were homogeneous for all the possible confounding factors. Results Calculated blood loss was 846 ± 197 (ml) in the flexion group and 1,242 ± 228 (ml) in the extension group (p \ 0.05). Drop of haemoglobin levels at 24 h in the study group and the control group was 1.9 ± 0.8 (g/dl) and 3.0 ± 0.5 (g/dl), respectively (p \ 0.01). Drop of haematocrit at 24 h was 4.5 ± 0.2 (%) in the flexion group and 6.7 ± 0.3 (%) in the extension group (p \ 0.05). Blood transfusion was necessary in 5 patients in the control group and was not necessary in any patient of the study group. Average knee flexion at day 7 was 105° ± 4° in the flexion group and 98° ± 7° in the extension group. Conclusion Knee flexion at 90° after TKA, associated with the intraoperative use of tranexamic, acid is an
A. S. Panni S. Cerciello M. Vasso C. Del Regno Department of Medicine and Science for Health, Molise University, Via De Sanctis 1, Campobasso, Italy S. Cerciello (&) Via R. Zandonai 11, 00135 Rome, Italy e-mail: [email protected]
effective method to reduce blood loss and the need for blood transfusion. The routine use of the present protocol is effective in reducing social costs and length of hospitalization of TKA procedures. Level of evidence Prospective comparative study, Level I. Keywords Total knee arthroplasty Knee flexion Blood loss Blood transfusion Tranexamic acid
Introduction Blood loss after the total knee arthroplasty (TKA) is a common occurrence and may result in anaemia and prolonged postoperative recovery [4]. The total amount of blood loss is the result of evident and hidden loss. It can be estimated by doubling the measured wound drainage loss and may reach up to 1.500 ml in average [6, 27]. Thus, it represents an important health issue and also an economical concern. Several aspects contribute to the overall blood loss; the use of tourniquet and the surgery time [3, 25], the use or the type of drainage (suction or not) [29, 30], the cementation of the prosthesis [24] and female gender [3]. However, the precise role of each one of these aspects is still discussed. Several methods have been proposed to reduce blood loss. Intraoperative ones include tourniquet use [14], minimally invasive surgery, computer assisted TKA [7, 13], bipolar sealer versus standard electrocautery [19], autologous platelet gel [11] and various drain placement protocols [29]. Postoperative methods include cryotherapy [12], different knee bandage types [5], and postoperative prolonged limb positioning [10, 17, 23, 29]. Local application of pharmacological agents has also been investigated.
123
Knee Surg Sports Traumatol Arthrosc
However, while local injection of bupivacaine and epinephrine and lavage with norepinephrine [8] were effective in reducing blood loss [2], local injection of the adrenaline [20] or lavage with epinephrine were ineffective [18]. It was hypothesized that the association between intraoperative tranexamic acid administration and prolonged knee flexion in the first 6 h after the operation had positive effects on postoperative blood loss and need for transfusion. In the same way, the possible negative effects of this protocol on wound healing were investigated. Finally, its influence on functional results at day 7 (in terms of range of motion and knee flexion) was evaluated.
Materials and methods One hundred consecutive patients undergoing primary TKA from 2012 to 2013 were randomizely divided into two groups. The study group consisted of 50 patients who laid in supine position with the hip at 45° of flexion and the knee flexed at 90° in a leg holder for the first 6 h after surgery. The control group consisted of 50 patients in whom the knee was completely extended for the first 6 h. All patients were affected by osteoarthritis. Patients with rheumatoid arthritis or previous surgery of the affected knee were excluded from the study. Other exclusion criteria included the following: prior operations performed within a 3-month interval, diabetes, haemostasis defects, history of peripheral vascular disease or thromboembolism, preoperative haemoglobin level less than 10 g/dl and patients in therapy with oral anticoagulants. The study was approved by the local ethic committee and specific and detailed informed consent from all participants was obtained. For randomization, a random number list was generated and each patient was assigned a number. Patients in the two groups were well matched for age, gender, body mass index (BMI), preoperative diagnosis, preoperative haemoglobin (Hb) and haematocrit (Hct) and preoperative circumference at the superior patellar pole (Table 1).
Table 1 Patient characteristics for study and control groups (mean ± SD or number) Study group (n = 50)
Control group (n = 50)
p value
Sex F/M
38/12
40/10
n.s.
Age (years)
69 ± 7
61 ± 6
n.s.
BMI (kg/m2)
28.1 ± 6.0
27.7 ± 5.9
n.s.
Preoperative Hb (g/dl)
13.4 ± 1.09
13.3 ± 1.11
n.s.
Preoperative Ht (%)
41.8 ± 5.5
42.6 ± 4.9
n.s.
Knee circumference (cm)
37.1 ± 1.4
37.0 ± 1.5
n.s.
123
Surgical technique All patients received 4,000 IU of low molecular weight heparin 12 h before surgery. Cefazolin 2 g was administered in all cases 45 min before incision. Spinal anaesthesia was performed in all cases. Blood pressure was evaluated in all cases throughout the operation. All patients received two endovenous doses of Tranexamic Acid (500 mg/5 ml) half an hour before skin incision. The tourniquet was positioned in all cases but not inflated, in order to perform a careful coagulation with a unipolar cautery device during the surgical approach. The operation was performed through a midline skin incision and medial parapatellar approach. A posterior-stabilized fixed TKA (Nexgen LPS, Zimmer, Warsaw, IN, USA). Bony resections were performed with the aid of intramedullary jigs on the femoral side. Extra medullary device was used on the tibia. To reduce blood loss, a bone plug obtained from previous bone cuts was inserted into the femoral canal. The tourniquet was inflated to 350 mmHg, right for the time of cement polymerization, and then deflated again. No intraoperative blood recovery was adopted during the operation. Suction drainage was positioned at the end of surgery and left in place for the first 24 h. All patients received a specific dressing, and a compressive bandage was performed at the end of surgery. Total surgical time as well as intraoperative blood loss was recorded. None of the patients were transfused during the operation. Postoperative management A standard nurse protocol was adopted in all cases. Cefazolin 2 g was administered 12 h after the operation and then stopped. All patients received standard (4,000 IU) low molecular weight heparin daily during the hospitalization. Continuous endovenous infusion of ketorolac and Ranitidin was administered the first 48 h after surgery. Additional pain management was achieved through the administration of three doses of Paracetamol 1 g. The quantity of saline solution infused intravenously within 24 h postoperatively was 2,500–3,000 ml. The drainage was removed 24 h after surgery and blood loss was measured. At this time the bandage was changed. However, the dressing on the skin incision was left in place. This was a special dressing (Aquacel AgÒ) made with hydrocolloid sodium carboxymethylcellulose in fibres and ionic silver; its hydrophilic action converts the exudate into gel. Moreover, silver ions have additional antimicrobial action. Compression stockings were positioned on the healthy leg. The bandage was removed on day 3 after surgery, the status of the dressing was checked again and compression stockings were applied also on the operated leg. Ice bags were placed around the affected knees 3 times a day from
Knee Surg Sports Traumatol Arthrosc
the first to the third postoperative day. Continuous passive motion was adopted from day 1 starting form 40° of flexion and with a daily increase in 10°. Partial weight bearing was permitted since 24 h postoperatively under the supervision of a physiotherapist. Active isometric quadriceps, initiative straight-leg raising and extending-flexing motion were encouraged just after operation. The postoperative levels of haemoglobin and haematocrit were measured on day 1 and 3 postoperatively. The cut off haemoglobin value for blood transfusion was 8.0 g/dl. To avoid bias and inter-observer errors, a single surgeon who did not attend to the operation documented the study data. To ensure that the examining surgeon was blinded to the postoperative limb position, he was not directly involved in patient care in the first 6 h. Outcome assessment Several parameters were analysed as follows: intraoperative blood loss (IBL), hidden blood loss (HBL), calculated total blood loss (CBL), haemoglobin and haematocrit levels measured at the 24th and 72nd postoperative hour, circumference of the knee, the scope of ecchymosis calculated by the palm method and range of active knee flexion on the first, third and seventh postoperative days. Another common parameter such as the postoperative wound blood loss (PWBL) was not taken into account in any case in the present study since the special dressing adopted allowed for minimal leakage from the wound. In fact weighing the soaked dressings and converting this measurement to volume usually approximate it. The specific dressing (Aquacel Ag) adopted in the present study allowed for minimal fluid effusion from the wound thus it was left in place for 15 days. IBL was calculated by adding the amount of the volume in suction bottles after reduction in wound irrigation fluid and the net blood weight of the sponges used during the procedure. CBL represents the total blood loss as the result of IBL and HBL and is calculated from the change in haematocrit using the formula reported by Nadler [21] and Gross [9]. The haematocrit (Hct) value used in the formula was measured preoperatively and at 24 h postoperatively. HBL represents the amount of blood, which is not recognized by the usual practice of assessing intra-operative loss and postoperative drainage. In the present study, it was calculated by subtracting the IBL from the CBL at the 24th postoperative hour (PWBL was not taken into account). The HBL has a great relevance after TKA and it has been estimated to be about 55 % of the calculated blood loss (CBL) [27]. The need for blood transfusion was recorded. Knee circumference (KC) was measured at the level of the superior patellar pole preoperatively and at 3 and 7 days postoperatively. The increase compared with the
preoperative measurement was then calculated. The scope of ecchymosis around the affected knee was calculated by the palm method at 3 days postoperatively. The area of the palm (hand minus digits) was introduced a by the American College of Surgeons and is around 1 % of the body surface area (BSA) and around 75.75 ± 9.93 cm2 [1]. Finally, to evaluate the discomfort of the prolonged knee flexion, the patients were asked to answer the following question: did you feel the knee position you had in the first hours after surgery comfortable, not very comfortable, not comfortable at all? The study was approved by the ethic committee of the Molise University in Campobasso (Prot. N 18800 _II/23, October 10th 2013). Statistical analysis was performed with SPSS 11.0 software for Windows. Continuous data with normal distribution were expressed as means (±SD). Wilcoxon signed-ranks test was performed and differences at a level of p \ 0.05 were considered statistically significant.
Results The two groups were homogeneous for the possible confounding factors age, blood pressure, coagulation parameters and operation time. Overall results are summarized in Table 2. Intraoperative blood loss (IBL) was 285 ± 71 (ml) in the study group and 277 ± 69 (ml) in the control group. CBL was 846 ± 197 (ml) and 1,242 ± 228 (ml), respectively (p \ 0.05). HBL was 451 ± 83 (ml) in the flexion group and 680 ± 90 (ml) in the extension group (p \ 0.05). Drop of Hb and Ht levels at 24 h in the study group and the control group were 1.9 ± 0.8 (g/dl) and 3.0 ± 0.5 (g/dl) (p \ 0.01) and 4.5 ± 0.2 (%) and 6.7 ± 0.3 (%), respectively (p \ 0.05). Drop of Hb levels at 72 h were 3.0 ± 0.7 (g/dl) and 4.5 ± 0.4 (g/dl) in flexion and extension group, respectively (p \ 0.01), whereas drop of Ht levels were 6.6 ± 0.4 (%) and 8.5 ± 0.5 (%), respectively (p \ 0.05). Both knee swelling and area of ecchymosis were significantly greater in the control group. The increase in KC at 3 and 7 days were 3.3 ± 0.75 (%) and 2.0 ± 0.9 (%) in the study group and 5.3 ± 1.8 (%) and 4.4 ± 1.6 (%) in the control group (p \ 0.05). Area of ecchymosis was 0.5 ± 0.3 (%) in the flexion group and 1.1 ± 0.4 (%) in the extension group (p \ 0.05). Blood transfusion was necessary in 5 cases in the control group and was not necessary in any case of the study group. No deep venous thrombosis and local complications were recorded. Thirty-nine patients in the study group and 44 in the control group defined the position of the knee as ‘‘comfortable’’, 9 in flexion group and 6 in extension group
123
Knee Surg Sports Traumatol Arthrosc Table 2 Detail of intra and postoperative changes in study and control groups (mean ± SD or number)
Study group (n = 50)
p value
Intraoperative blood loss (ml) (mean ± SD)
285 ± 71
277 ± 69
Calculated total blood loss (ml) at 24 h (mean ± SD)
846 ± 197
1,242 ± 228
\0.05
Hidden blood loss (ml) at 24 h (mean ± SD)
451 ± 83
680 ± 90
\0.05
Drop of Hb level (g/dl) at 24 h
1.9 ± 0.8
3.0 ± 0.5
\0.01
Drop of Ht level (%) at 24 h
4.5 ± 0.2
6.7 ± 0.3
\0.05
Drop of Hb level (g/dl) at 72 h
3.0 ± 0.7
4.5 ± 0.4
\0.01
Drop of Ht level (%) at 72 h
6.6 ± 0.4
8.5 ± 0.5
\0.05
Increasing rate of knee circumference (%) at day 3
3.3 ± 0.7
5.3 ± 1.8
\0.05
Increasing rate of knee circumference (%) at day 7
2.0 ± 0.9
4.4 ± 1.6
\0.05
Area of ecchymosis (%)
0.5 ± 0.3
1.1 ± 0.4
\0.05
Knee flexion at day 7 Blood transfusion
105° ± 4° 0
98° ± 7° 5
defined it ‘‘not very comfortable’’, whereas 2 in the first group defined it ‘‘not comfortable at all’’. Concerning functional results, average knee flexion at day 7 was 105° ± 4° in the flexion group and 98° ± 7° in the extension group.
Discussion The most important finding of this study was the significantly positive effect of prolonged (for 6 h) knee flexion at 90° with the hip flexed at 45° in association with the infusion of tranexamic acid in reducing total blood loss, HBL and the need for transfusion after TKA. Additionally the reduction in soft tissue swelling has a positive effect on pain perception and makes range of movement recovery quicker and easier. This has a positive impact on the social costs and length of hospitalization of a TKA procedure. Moreover, it is a safe method since no local or general complications were reported, and it is not uncomfortable due to residual spinal anaesthesia and to the effects of painkiller. The reported data are in accordance with those of Speck. He found that 70 degrees of knee flexion for 6 h after the cemented TKA significantly reduced the blood loss measured by means of wound drainage by 30 % [28]. However, considering blood loss from the drainage alone is not a precise method and may lead to underestimation of overall blood loss. Ong et al. [22] compared the effects of three different postoperative positioning on blood loss. Leg extension was compared to knee extension with 35° of hip flexion and to knee flexion at 70° with 35° of hip flexion. In the last two groups, the calculated haemoglobin loss was 25 % less than in the full extension group [22]. Li et al. [15] reported that mild knee flexion (at 30°) with the leg elevated at 30° of hip flexion for 72 h after TKA was effective in decreasing HBL. On the contrary Ma et al. [16]
123
Control group (n = 50)
reported no significant difference in postoperative blood loss between leg extension and knee flexion of 70° with hip flexion of 70° for the first 24 h. Madarevic et al. reported similar results comparing four methods to reduce blood loss, including knee positioning (flexion vs extension). No differences were reported in blood loss according to the considered methods [17]. The discordance in these results has several explanations. First of all, it is of an extreme importance to test the groups for differences in other possibly confounding variables. In the present study, the two groups did not differ according to the age, preoperative levels of haemoglobin and haematocrit and the coagulation capacity. Moreover, blood loss after TKA may be influenced by surgical technique variations and postoperative nurse protocol. In the present study, both aspects have been standardized to avoid confounding factors. The effects of knee flexion on perioperative and postoperative bleeding are multifactorial. Knee flexion causes compression on the popliteal vessels with reduction in local perfusion and finally reduction in oxygen tension. Prolonged deep flexion could be dangerous for tissue healing [23]. For this reason, we left the knee flexed at 90° only for 6 h. However, mild flexion reducing knee swelling has a positive impact on local perfusion and oxygen tension. The reduction in CBL is important since it positively affect the cardio-pulmonary recovery after surgery thus allowing for early and aggressive rehabilitation programme. However, HBL is important too. It is defined as the residual blood into the joint and into the surrounding soft tissues. Clinical indicators of HBL are limb swelling and ecchymosis [21]. As reported by Li, the reduction in HBL is beneficial to the rehabilitation after TKA [15]. Joint effusion causes quadriceps inhibition by reduced afferent activity of intracapsular receptors, resulting in quadriceps weakness [26]. Swelling increases weight of the affected limb, requiring more muscle force for conducting straight-
Knee Surg Sports Traumatol Arthrosc
leg raising activities. Finally, knee swelling increases the tension of soft tissues around the knee, reducing knee flexion motion and possibly inhibiting maximal wound healing because of decreased oxygen tension at the skin edges. Knee flexion at 90° for 6 h is easily reproducible. Moreover, it is a good compromise between degree of flexion and duration of flexion. In the present study the use of two doses of tranexamic acid as well as the absence of tourniquet (except for the cementation) makes postoperative bleeding predictable. In these conditions, knee flexion at 90° for the first hours may reduce local perfusion, bleeding and local effusion. Lower degrees of flexion, even if associated with limb elevation, may be less effective, since they lead to lower compression on blood vessel with lower impact on peripheral blood circulation. On the contrary, prolonging knee flexion for longer periods may help in reducing blood loss but it is hardly accepted by patients. The present study has some limitations. First of all the population is relatively small. In addition, medical history of the patients was not deeply investigated. Thus, other comorbidities that could have influenced the results were completely not taken into account. However, the protocol has shown to be easily reproducible and the results are encouraging. The reduction in blood loss and the need of transfusion were statistically significative. This leads to reduced social costs and length of hospitalization of TKA procedures.
4.
5.
6.
7.
8.
9. 10.
11.
12.
13.
Conclusion 14.
The results of the present study confirm that knee flexion at 90° in the first hours after TKA, associated with the intraoperative use of tranexamic acid, is a safe and effective method to reduce blood loss and the need for blood transfusion. Postoperative knee flexion was well tolerated with minimal discomfort. The routine use of this protocol may help in reducing social costs and the length of hospitalization.
References 1. Agarwal P, Sahu S (2010) Determination of hand and palm area as a ratio of body surface area in Indian population. Indian J Plast Surg 43(1):49–53 2. Anderson LA, Engel GM, Bruckner JD, Stoddard GJ, Peters CL (2009) Reduced blood loss after total knee arthroplasty with local injection of bupivacaine and epinephrine. J Knee Surg 22:130–136 3. Bell TH, Berta D, Ralley F, Macdonald SJ, McCalden RW, Bourne RB, Rorabeck CH, Naudie DD (2009) Factors affecting perioperative blood loss and transfusion rates in primary total
15.
16.
17.
18.
19.
20.
21.
joint arthroplasty: a prospective analysis of 1642 patients. Can J Surg 52:295–301 Carson JL, Duff A, Poses RM, Berlin JA, Spence RK, Trout R, Noveck H, Strom BL (1996) Effect of anaemia and cardiovascular disease on surgical mortality and morbidity. Lancet 348:1055–1060 Charalambides C, Beer M, Melhuish J, Williams RJ, Cobb AG (2005) Bandaging technique after knee replacement. Acta Orthop 76:89–94 Chen LL, Wang WC, Mao XZ, Yu M, Zhu Q (2007) Evaluation and treatment of hemorrhage after hip and knee arthroplasty in the aged. Zhong Nan Da Xue Xue Bao Yi XueBan 32:316–319 Chotanaphuti T, Ongnamthip P, Teeraleekul K, Kraturerk C (2008) Comparative study between computer assisted-navigation and conventional technique in minimally invasive surgery total knee arthroplasty, prospective control study. J Med Assoc Thai 91:1382–1388 Gasparini G, Papaleo P, Pola P, Cerciello S, Pola E, Fabbriciani C (2006) Local infusion of norepinephrine reduces blood losses and need of transfusion in total knee arthroplasty. Int Orthop 30:253–256 Gross JB (1983) Estimating allowable blood loss: corrected for dilution. Anesthesiology 58:277–280 Hewitt B, Shakespeare D (2001) Flexion vs. extension: a comparison of post-operative total knee arthroplasty mobilisation regimes. Knee 8:305–309 Horstmann WG, Slappendel R, van Hellemondt GG, Wymenga AW, Jack N, Everts PA (2011) Autologous platelet gel in total knee arthroplasty: a prospective randomized study. Knee Surg Sports Traumatol Arthrosc 19:115–121 Kullenberg B, Ylip S, So¨derlund K, Resch S (2006) Postoperative cryotherapy after total knee arthroplasty: a prospective study of 86 patients. J Arthroplasty 21:1175–1179 Lee DH, Choi J, Nha KW, Kim HJ, Han SB (2011) No difference in early functional outcomes for mini-midvastus and limited medial parapatellar approaches in navigation-assisted total knee arthroplasty: a prospective randomized clinical trial. Knee Surg Sports Traumatol Arthrosc 19:66–73 Li B, Wen Y, Wu H, Qian Q, Lin X, Zhao H (2009) The effect of tourniquet use on hidden blood loss in total knee arthroplasty. Int Orthop 33:1263–1268 Li B, Wen Y, Liu D, Tian L (2012) The effect of knee position on blood loss and range of motion following total knee arthroplasty. Knee Surg Sports Traumatol Arthrosc 20:594–599 Ma T, Khan RJ, Carey Smith R, Nivbrant B, Wood DJ (2008) Effect of flexion/extension splintage post total knee arthroplasty on blood loss and range of motion: a randomised controlled trial. Knee 15:15–19 Madarevic T, Tudor A, Sestan B, Santic V, Gulan G, Prpic T, Ruzic L (2009) Postoperative blood loss management in total knee arthroplasty: a comparison of four different methods. Knee Surg Sports Traumatol Arthrosc 19:955–959 Malone KJ, Matuszak S, Mayo D, Greene P (2009) The effect of intra-articular epinephrine lavage on blood loss following total knee arthroplasty. Orthopedics 32:100 Marulanda GA, Krebs VE, Bierbaum BE, Goldberg VM, Ries M, Ulrich SD, Seyler TM, Mont MA (2009) Hemostasis using a bipolar sealer in primary unilateral total knee arthroplasty. Am J Orthop (Belle Mead NJ) 38:E179–E183 Mesa-Ramos F, Mesa-Ramos M, Maqueira-Canosa C, Carpintero P (2008) Predictors for blood transfusion following total knee arthroplasty: a prospective randomised study. Acta Orthop Belg 74:83–89 Nadler SB, Hidalgo JU, Bloch T (1962) Prediction of blood volume in normal human adults. Surgery 51:224–232
123
Knee Surg Sports Traumatol Arthrosc 22. Ong SM, Taylor GJ (2003) Can knee position save blood following total knee replacement? Knee 10:81–85 23. Panni AS, Vasso M, Cerciello S, Salgarello M (2011) Wound complications in total knee arthroplasty. Which flap is to be used? With or without retention of prosthesis? Knee Surg Sports Traumatol Arthrosc 19(7):1060–1068 24. Porteous AJ, Bartlett RJ (2003) Post-operative drainage after cemented, hybrid and uncemented total knee replacement. Knee 10(4):371–374 25. Prasad N, Padmanabhan V, Mullaji A (2007) Blood loss in total knee arthroplasty: an analysis of risk factors. Int Orthop 31:39–44 26. Schro¨der D, Pa¨ssler HH (1994) Combination of cold and compression after knee surgery. Knee Surg Sports Traumatol Arthrosc 2:158–165 27. Sehat KR, Evans RL, Newman JH (2004) Hidden blood loss following hip and knee arthroplasty. Correct management of
123
blood loss should take sudden loss into account. J Bone Joint Surg Br 86:561–565 28. Speck M, Jakob R, Brinkmann KE (1999) Blood loss after total knee arthroplasty in relation to positioning: 70° flexion vs extension. J Bone Joint Surg [Br] 81(Suppl 11):245 29. Stucinskas J, Tarasevicius S, Cebatorius A, Robertsson O, Smailys A, Wingstrand H (2009) Conventional drainage versus four hour clamping drainage after total knee arthroplasty in severe osteoarthritis: a prospective, randomised trial. Int Orthop 33:1275–1278 30. Tai TW, Jou IM, Chang CW, Lai KA, Lin CJ, Yang CY (2010) Non-drainage is better than 4-hour clamping drainage in total knee arthroplasty. Orthopedics 10:156–160
KNEE
Knee flexion after total knee arthroplasty reduces blood loss Alfredo Schiavone Panni • Simone Cerciello Michele Vasso • Chiara Del Regno
•
Received: 30 May 2013 / Accepted: 1 April 2014 Ó Springer-Verlag Berlin Heidelberg 2014
Abstract Purpose Extensive blood loss after total knee arthroplasty (TKA) may be a potential problem since it leads to anaemia, increased need for transfusion and prolonged hospitalization. Aim of this study was to investigate the effects of postoperative knee flexion after TKA on blood loss and the need for transfusion. Methods One hundred consecutive patients undergoing primary TKA from 2012 to 2013 were randomizely divided into two groups. In one group, the knee was extended for the first 6 h after surgery, whereas in the other was flexed at 90° for the same time. Two doses of endovenous tranexamic acid were administered in all subjects. Patients were homogeneous for all the possible confounding factors. Results Calculated blood loss was 846 ± 197 (ml) in the flexion group and 1,242 ± 228 (ml) in the extension group (p \ 0.05). Drop of haemoglobin levels at 24 h in the study group and the control group was 1.9 ± 0.8 (g/dl) and 3.0 ± 0.5 (g/dl), respectively (p \ 0.01). Drop of haematocrit at 24 h was 4.5 ± 0.2 (%) in the flexion group and 6.7 ± 0.3 (%) in the extension group (p \ 0.05). Blood transfusion was necessary in 5 patients in the control group and was not necessary in any patient of the study group. Average knee flexion at day 7 was 105° ± 4° in the flexion group and 98° ± 7° in the extension group. Conclusion Knee flexion at 90° after TKA, associated with the intraoperative use of tranexamic, acid is an
A. S. Panni S. Cerciello M. Vasso C. Del Regno Department of Medicine and Science for Health, Molise University, Via De Sanctis 1, Campobasso, Italy S. Cerciello (&) Via R. Zandonai 11, 00135 Rome, Italy e-mail: [email protected]
effective method to reduce blood loss and the need for blood transfusion. The routine use of the present protocol is effective in reducing social costs and length of hospitalization of TKA procedures. Level of evidence Prospective comparative study, Level I. Keywords Total knee arthroplasty Knee flexion Blood loss Blood transfusion Tranexamic acid
Introduction Blood loss after the total knee arthroplasty (TKA) is a common occurrence and may result in anaemia and prolonged postoperative recovery [4]. The total amount of blood loss is the result of evident and hidden loss. It can be estimated by doubling the measured wound drainage loss and may reach up to 1.500 ml in average [6, 27]. Thus, it represents an important health issue and also an economical concern. Several aspects contribute to the overall blood loss; the use of tourniquet and the surgery time [3, 25], the use or the type of drainage (suction or not) [29, 30], the cementation of the prosthesis [24] and female gender [3]. However, the precise role of each one of these aspects is still discussed. Several methods have been proposed to reduce blood loss. Intraoperative ones include tourniquet use [14], minimally invasive surgery, computer assisted TKA [7, 13], bipolar sealer versus standard electrocautery [19], autologous platelet gel [11] and various drain placement protocols [29]. Postoperative methods include cryotherapy [12], different knee bandage types [5], and postoperative prolonged limb positioning [10, 17, 23, 29]. Local application of pharmacological agents has also been investigated.
123
Knee Surg Sports Traumatol Arthrosc
However, while local injection of bupivacaine and epinephrine and lavage with norepinephrine [8] were effective in reducing blood loss [2], local injection of the adrenaline [20] or lavage with epinephrine were ineffective [18]. It was hypothesized that the association between intraoperative tranexamic acid administration and prolonged knee flexion in the first 6 h after the operation had positive effects on postoperative blood loss and need for transfusion. In the same way, the possible negative effects of this protocol on wound healing were investigated. Finally, its influence on functional results at day 7 (in terms of range of motion and knee flexion) was evaluated.
Materials and methods One hundred consecutive patients undergoing primary TKA from 2012 to 2013 were randomizely divided into two groups. The study group consisted of 50 patients who laid in supine position with the hip at 45° of flexion and the knee flexed at 90° in a leg holder for the first 6 h after surgery. The control group consisted of 50 patients in whom the knee was completely extended for the first 6 h. All patients were affected by osteoarthritis. Patients with rheumatoid arthritis or previous surgery of the affected knee were excluded from the study. Other exclusion criteria included the following: prior operations performed within a 3-month interval, diabetes, haemostasis defects, history of peripheral vascular disease or thromboembolism, preoperative haemoglobin level less than 10 g/dl and patients in therapy with oral anticoagulants. The study was approved by the local ethic committee and specific and detailed informed consent from all participants was obtained. For randomization, a random number list was generated and each patient was assigned a number. Patients in the two groups were well matched for age, gender, body mass index (BMI), preoperative diagnosis, preoperative haemoglobin (Hb) and haematocrit (Hct) and preoperative circumference at the superior patellar pole (Table 1).
Table 1 Patient characteristics for study and control groups (mean ± SD or number) Study group (n = 50)
Control group (n = 50)
p value
Sex F/M
38/12
40/10
n.s.
Age (years)
69 ± 7
61 ± 6
n.s.
BMI (kg/m2)
28.1 ± 6.0
27.7 ± 5.9
n.s.
Preoperative Hb (g/dl)
13.4 ± 1.09
13.3 ± 1.11
n.s.
Preoperative Ht (%)
41.8 ± 5.5
42.6 ± 4.9
n.s.
Knee circumference (cm)
37.1 ± 1.4
37.0 ± 1.5
n.s.
123
Surgical technique All patients received 4,000 IU of low molecular weight heparin 12 h before surgery. Cefazolin 2 g was administered in all cases 45 min before incision. Spinal anaesthesia was performed in all cases. Blood pressure was evaluated in all cases throughout the operation. All patients received two endovenous doses of Tranexamic Acid (500 mg/5 ml) half an hour before skin incision. The tourniquet was positioned in all cases but not inflated, in order to perform a careful coagulation with a unipolar cautery device during the surgical approach. The operation was performed through a midline skin incision and medial parapatellar approach. A posterior-stabilized fixed TKA (Nexgen LPS, Zimmer, Warsaw, IN, USA). Bony resections were performed with the aid of intramedullary jigs on the femoral side. Extra medullary device was used on the tibia. To reduce blood loss, a bone plug obtained from previous bone cuts was inserted into the femoral canal. The tourniquet was inflated to 350 mmHg, right for the time of cement polymerization, and then deflated again. No intraoperative blood recovery was adopted during the operation. Suction drainage was positioned at the end of surgery and left in place for the first 24 h. All patients received a specific dressing, and a compressive bandage was performed at the end of surgery. Total surgical time as well as intraoperative blood loss was recorded. None of the patients were transfused during the operation. Postoperative management A standard nurse protocol was adopted in all cases. Cefazolin 2 g was administered 12 h after the operation and then stopped. All patients received standard (4,000 IU) low molecular weight heparin daily during the hospitalization. Continuous endovenous infusion of ketorolac and Ranitidin was administered the first 48 h after surgery. Additional pain management was achieved through the administration of three doses of Paracetamol 1 g. The quantity of saline solution infused intravenously within 24 h postoperatively was 2,500–3,000 ml. The drainage was removed 24 h after surgery and blood loss was measured. At this time the bandage was changed. However, the dressing on the skin incision was left in place. This was a special dressing (Aquacel AgÒ) made with hydrocolloid sodium carboxymethylcellulose in fibres and ionic silver; its hydrophilic action converts the exudate into gel. Moreover, silver ions have additional antimicrobial action. Compression stockings were positioned on the healthy leg. The bandage was removed on day 3 after surgery, the status of the dressing was checked again and compression stockings were applied also on the operated leg. Ice bags were placed around the affected knees 3 times a day from
Knee Surg Sports Traumatol Arthrosc
the first to the third postoperative day. Continuous passive motion was adopted from day 1 starting form 40° of flexion and with a daily increase in 10°. Partial weight bearing was permitted since 24 h postoperatively under the supervision of a physiotherapist. Active isometric quadriceps, initiative straight-leg raising and extending-flexing motion were encouraged just after operation. The postoperative levels of haemoglobin and haematocrit were measured on day 1 and 3 postoperatively. The cut off haemoglobin value for blood transfusion was 8.0 g/dl. To avoid bias and inter-observer errors, a single surgeon who did not attend to the operation documented the study data. To ensure that the examining surgeon was blinded to the postoperative limb position, he was not directly involved in patient care in the first 6 h. Outcome assessment Several parameters were analysed as follows: intraoperative blood loss (IBL), hidden blood loss (HBL), calculated total blood loss (CBL), haemoglobin and haematocrit levels measured at the 24th and 72nd postoperative hour, circumference of the knee, the scope of ecchymosis calculated by the palm method and range of active knee flexion on the first, third and seventh postoperative days. Another common parameter such as the postoperative wound blood loss (PWBL) was not taken into account in any case in the present study since the special dressing adopted allowed for minimal leakage from the wound. In fact weighing the soaked dressings and converting this measurement to volume usually approximate it. The specific dressing (Aquacel Ag) adopted in the present study allowed for minimal fluid effusion from the wound thus it was left in place for 15 days. IBL was calculated by adding the amount of the volume in suction bottles after reduction in wound irrigation fluid and the net blood weight of the sponges used during the procedure. CBL represents the total blood loss as the result of IBL and HBL and is calculated from the change in haematocrit using the formula reported by Nadler [21] and Gross [9]. The haematocrit (Hct) value used in the formula was measured preoperatively and at 24 h postoperatively. HBL represents the amount of blood, which is not recognized by the usual practice of assessing intra-operative loss and postoperative drainage. In the present study, it was calculated by subtracting the IBL from the CBL at the 24th postoperative hour (PWBL was not taken into account). The HBL has a great relevance after TKA and it has been estimated to be about 55 % of the calculated blood loss (CBL) [27]. The need for blood transfusion was recorded. Knee circumference (KC) was measured at the level of the superior patellar pole preoperatively and at 3 and 7 days postoperatively. The increase compared with the
preoperative measurement was then calculated. The scope of ecchymosis around the affected knee was calculated by the palm method at 3 days postoperatively. The area of the palm (hand minus digits) was introduced a by the American College of Surgeons and is around 1 % of the body surface area (BSA) and around 75.75 ± 9.93 cm2 [1]. Finally, to evaluate the discomfort of the prolonged knee flexion, the patients were asked to answer the following question: did you feel the knee position you had in the first hours after surgery comfortable, not very comfortable, not comfortable at all? The study was approved by the ethic committee of the Molise University in Campobasso (Prot. N 18800 _II/23, October 10th 2013). Statistical analysis was performed with SPSS 11.0 software for Windows. Continuous data with normal distribution were expressed as means (±SD). Wilcoxon signed-ranks test was performed and differences at a level of p \ 0.05 were considered statistically significant.
Results The two groups were homogeneous for the possible confounding factors age, blood pressure, coagulation parameters and operation time. Overall results are summarized in Table 2. Intraoperative blood loss (IBL) was 285 ± 71 (ml) in the study group and 277 ± 69 (ml) in the control group. CBL was 846 ± 197 (ml) and 1,242 ± 228 (ml), respectively (p \ 0.05). HBL was 451 ± 83 (ml) in the flexion group and 680 ± 90 (ml) in the extension group (p \ 0.05). Drop of Hb and Ht levels at 24 h in the study group and the control group were 1.9 ± 0.8 (g/dl) and 3.0 ± 0.5 (g/dl) (p \ 0.01) and 4.5 ± 0.2 (%) and 6.7 ± 0.3 (%), respectively (p \ 0.05). Drop of Hb levels at 72 h were 3.0 ± 0.7 (g/dl) and 4.5 ± 0.4 (g/dl) in flexion and extension group, respectively (p \ 0.01), whereas drop of Ht levels were 6.6 ± 0.4 (%) and 8.5 ± 0.5 (%), respectively (p \ 0.05). Both knee swelling and area of ecchymosis were significantly greater in the control group. The increase in KC at 3 and 7 days were 3.3 ± 0.75 (%) and 2.0 ± 0.9 (%) in the study group and 5.3 ± 1.8 (%) and 4.4 ± 1.6 (%) in the control group (p \ 0.05). Area of ecchymosis was 0.5 ± 0.3 (%) in the flexion group and 1.1 ± 0.4 (%) in the extension group (p \ 0.05). Blood transfusion was necessary in 5 cases in the control group and was not necessary in any case of the study group. No deep venous thrombosis and local complications were recorded. Thirty-nine patients in the study group and 44 in the control group defined the position of the knee as ‘‘comfortable’’, 9 in flexion group and 6 in extension group
123
Knee Surg Sports Traumatol Arthrosc Table 2 Detail of intra and postoperative changes in study and control groups (mean ± SD or number)
Study group (n = 50)
p value
Intraoperative blood loss (ml) (mean ± SD)
285 ± 71
277 ± 69
Calculated total blood loss (ml) at 24 h (mean ± SD)
846 ± 197
1,242 ± 228
\0.05
Hidden blood loss (ml) at 24 h (mean ± SD)
451 ± 83
680 ± 90
\0.05
Drop of Hb level (g/dl) at 24 h
1.9 ± 0.8
3.0 ± 0.5
\0.01
Drop of Ht level (%) at 24 h
4.5 ± 0.2
6.7 ± 0.3
\0.05
Drop of Hb level (g/dl) at 72 h
3.0 ± 0.7
4.5 ± 0.4
\0.01
Drop of Ht level (%) at 72 h
6.6 ± 0.4
8.5 ± 0.5
\0.05
Increasing rate of knee circumference (%) at day 3
3.3 ± 0.7
5.3 ± 1.8
\0.05
Increasing rate of knee circumference (%) at day 7
2.0 ± 0.9
4.4 ± 1.6
\0.05
Area of ecchymosis (%)
0.5 ± 0.3
1.1 ± 0.4
\0.05
Knee flexion at day 7 Blood transfusion
105° ± 4° 0
98° ± 7° 5
defined it ‘‘not very comfortable’’, whereas 2 in the first group defined it ‘‘not comfortable at all’’. Concerning functional results, average knee flexion at day 7 was 105° ± 4° in the flexion group and 98° ± 7° in the extension group.
Discussion The most important finding of this study was the significantly positive effect of prolonged (for 6 h) knee flexion at 90° with the hip flexed at 45° in association with the infusion of tranexamic acid in reducing total blood loss, HBL and the need for transfusion after TKA. Additionally the reduction in soft tissue swelling has a positive effect on pain perception and makes range of movement recovery quicker and easier. This has a positive impact on the social costs and length of hospitalization of a TKA procedure. Moreover, it is a safe method since no local or general complications were reported, and it is not uncomfortable due to residual spinal anaesthesia and to the effects of painkiller. The reported data are in accordance with those of Speck. He found that 70 degrees of knee flexion for 6 h after the cemented TKA significantly reduced the blood loss measured by means of wound drainage by 30 % [28]. However, considering blood loss from the drainage alone is not a precise method and may lead to underestimation of overall blood loss. Ong et al. [22] compared the effects of three different postoperative positioning on blood loss. Leg extension was compared to knee extension with 35° of hip flexion and to knee flexion at 70° with 35° of hip flexion. In the last two groups, the calculated haemoglobin loss was 25 % less than in the full extension group [22]. Li et al. [15] reported that mild knee flexion (at 30°) with the leg elevated at 30° of hip flexion for 72 h after TKA was effective in decreasing HBL. On the contrary Ma et al. [16]
123
Control group (n = 50)
reported no significant difference in postoperative blood loss between leg extension and knee flexion of 70° with hip flexion of 70° for the first 24 h. Madarevic et al. reported similar results comparing four methods to reduce blood loss, including knee positioning (flexion vs extension). No differences were reported in blood loss according to the considered methods [17]. The discordance in these results has several explanations. First of all, it is of an extreme importance to test the groups for differences in other possibly confounding variables. In the present study, the two groups did not differ according to the age, preoperative levels of haemoglobin and haematocrit and the coagulation capacity. Moreover, blood loss after TKA may be influenced by surgical technique variations and postoperative nurse protocol. In the present study, both aspects have been standardized to avoid confounding factors. The effects of knee flexion on perioperative and postoperative bleeding are multifactorial. Knee flexion causes compression on the popliteal vessels with reduction in local perfusion and finally reduction in oxygen tension. Prolonged deep flexion could be dangerous for tissue healing [23]. For this reason, we left the knee flexed at 90° only for 6 h. However, mild flexion reducing knee swelling has a positive impact on local perfusion and oxygen tension. The reduction in CBL is important since it positively affect the cardio-pulmonary recovery after surgery thus allowing for early and aggressive rehabilitation programme. However, HBL is important too. It is defined as the residual blood into the joint and into the surrounding soft tissues. Clinical indicators of HBL are limb swelling and ecchymosis [21]. As reported by Li, the reduction in HBL is beneficial to the rehabilitation after TKA [15]. Joint effusion causes quadriceps inhibition by reduced afferent activity of intracapsular receptors, resulting in quadriceps weakness [26]. Swelling increases weight of the affected limb, requiring more muscle force for conducting straight-
Knee Surg Sports Traumatol Arthrosc
leg raising activities. Finally, knee swelling increases the tension of soft tissues around the knee, reducing knee flexion motion and possibly inhibiting maximal wound healing because of decreased oxygen tension at the skin edges. Knee flexion at 90° for 6 h is easily reproducible. Moreover, it is a good compromise between degree of flexion and duration of flexion. In the present study the use of two doses of tranexamic acid as well as the absence of tourniquet (except for the cementation) makes postoperative bleeding predictable. In these conditions, knee flexion at 90° for the first hours may reduce local perfusion, bleeding and local effusion. Lower degrees of flexion, even if associated with limb elevation, may be less effective, since they lead to lower compression on blood vessel with lower impact on peripheral blood circulation. On the contrary, prolonging knee flexion for longer periods may help in reducing blood loss but it is hardly accepted by patients. The present study has some limitations. First of all the population is relatively small. In addition, medical history of the patients was not deeply investigated. Thus, other comorbidities that could have influenced the results were completely not taken into account. However, the protocol has shown to be easily reproducible and the results are encouraging. The reduction in blood loss and the need of transfusion were statistically significative. This leads to reduced social costs and length of hospitalization of TKA procedures.
4.
5.
6.
7.
8.
9. 10.
11.
12.
13.
Conclusion 14.
The results of the present study confirm that knee flexion at 90° in the first hours after TKA, associated with the intraoperative use of tranexamic acid, is a safe and effective method to reduce blood loss and the need for blood transfusion. Postoperative knee flexion was well tolerated with minimal discomfort. The routine use of this protocol may help in reducing social costs and the length of hospitalization.
References 1. Agarwal P, Sahu S (2010) Determination of hand and palm area as a ratio of body surface area in Indian population. Indian J Plast Surg 43(1):49–53 2. Anderson LA, Engel GM, Bruckner JD, Stoddard GJ, Peters CL (2009) Reduced blood loss after total knee arthroplasty with local injection of bupivacaine and epinephrine. J Knee Surg 22:130–136 3. Bell TH, Berta D, Ralley F, Macdonald SJ, McCalden RW, Bourne RB, Rorabeck CH, Naudie DD (2009) Factors affecting perioperative blood loss and transfusion rates in primary total
15.
16.
17.
18.
19.
20.
21.
joint arthroplasty: a prospective analysis of 1642 patients. Can J Surg 52:295–301 Carson JL, Duff A, Poses RM, Berlin JA, Spence RK, Trout R, Noveck H, Strom BL (1996) Effect of anaemia and cardiovascular disease on surgical mortality and morbidity. Lancet 348:1055–1060 Charalambides C, Beer M, Melhuish J, Williams RJ, Cobb AG (2005) Bandaging technique after knee replacement. Acta Orthop 76:89–94 Chen LL, Wang WC, Mao XZ, Yu M, Zhu Q (2007) Evaluation and treatment of hemorrhage after hip and knee arthroplasty in the aged. Zhong Nan Da Xue Xue Bao Yi XueBan 32:316–319 Chotanaphuti T, Ongnamthip P, Teeraleekul K, Kraturerk C (2008) Comparative study between computer assisted-navigation and conventional technique in minimally invasive surgery total knee arthroplasty, prospective control study. J Med Assoc Thai 91:1382–1388 Gasparini G, Papaleo P, Pola P, Cerciello S, Pola E, Fabbriciani C (2006) Local infusion of norepinephrine reduces blood losses and need of transfusion in total knee arthroplasty. Int Orthop 30:253–256 Gross JB (1983) Estimating allowable blood loss: corrected for dilution. Anesthesiology 58:277–280 Hewitt B, Shakespeare D (2001) Flexion vs. extension: a comparison of post-operative total knee arthroplasty mobilisation regimes. Knee 8:305–309 Horstmann WG, Slappendel R, van Hellemondt GG, Wymenga AW, Jack N, Everts PA (2011) Autologous platelet gel in total knee arthroplasty: a prospective randomized study. Knee Surg Sports Traumatol Arthrosc 19:115–121 Kullenberg B, Ylip S, So¨derlund K, Resch S (2006) Postoperative cryotherapy after total knee arthroplasty: a prospective study of 86 patients. J Arthroplasty 21:1175–1179 Lee DH, Choi J, Nha KW, Kim HJ, Han SB (2011) No difference in early functional outcomes for mini-midvastus and limited medial parapatellar approaches in navigation-assisted total knee arthroplasty: a prospective randomized clinical trial. Knee Surg Sports Traumatol Arthrosc 19:66–73 Li B, Wen Y, Wu H, Qian Q, Lin X, Zhao H (2009) The effect of tourniquet use on hidden blood loss in total knee arthroplasty. Int Orthop 33:1263–1268 Li B, Wen Y, Liu D, Tian L (2012) The effect of knee position on blood loss and range of motion following total knee arthroplasty. Knee Surg Sports Traumatol Arthrosc 20:594–599 Ma T, Khan RJ, Carey Smith R, Nivbrant B, Wood DJ (2008) Effect of flexion/extension splintage post total knee arthroplasty on blood loss and range of motion: a randomised controlled trial. Knee 15:15–19 Madarevic T, Tudor A, Sestan B, Santic V, Gulan G, Prpic T, Ruzic L (2009) Postoperative blood loss management in total knee arthroplasty: a comparison of four different methods. Knee Surg Sports Traumatol Arthrosc 19:955–959 Malone KJ, Matuszak S, Mayo D, Greene P (2009) The effect of intra-articular epinephrine lavage on blood loss following total knee arthroplasty. Orthopedics 32:100 Marulanda GA, Krebs VE, Bierbaum BE, Goldberg VM, Ries M, Ulrich SD, Seyler TM, Mont MA (2009) Hemostasis using a bipolar sealer in primary unilateral total knee arthroplasty. Am J Orthop (Belle Mead NJ) 38:E179–E183 Mesa-Ramos F, Mesa-Ramos M, Maqueira-Canosa C, Carpintero P (2008) Predictors for blood transfusion following total knee arthroplasty: a prospective randomised study. Acta Orthop Belg 74:83–89 Nadler SB, Hidalgo JU, Bloch T (1962) Prediction of blood volume in normal human adults. Surgery 51:224–232
123
Knee Surg Sports Traumatol Arthrosc 22. Ong SM, Taylor GJ (2003) Can knee position save blood following total knee replacement? Knee 10:81–85 23. Panni AS, Vasso M, Cerciello S, Salgarello M (2011) Wound complications in total knee arthroplasty. Which flap is to be used? With or without retention of prosthesis? Knee Surg Sports Traumatol Arthrosc 19(7):1060–1068 24. Porteous AJ, Bartlett RJ (2003) Post-operative drainage after cemented, hybrid and uncemented total knee replacement. Knee 10(4):371–374 25. Prasad N, Padmanabhan V, Mullaji A (2007) Blood loss in total knee arthroplasty: an analysis of risk factors. Int Orthop 31:39–44 26. Schro¨der D, Pa¨ssler HH (1994) Combination of cold and compression after knee surgery. Knee Surg Sports Traumatol Arthrosc 2:158–165 27. Sehat KR, Evans RL, Newman JH (2004) Hidden blood loss following hip and knee arthroplasty. Correct management of
123
blood loss should take sudden loss into account. J Bone Joint Surg Br 86:561–565 28. Speck M, Jakob R, Brinkmann KE (1999) Blood loss after total knee arthroplasty in relation to positioning: 70° flexion vs extension. J Bone Joint Surg [Br] 81(Suppl 11):245 29. Stucinskas J, Tarasevicius S, Cebatorius A, Robertsson O, Smailys A, Wingstrand H (2009) Conventional drainage versus four hour clamping drainage after total knee arthroplasty in severe osteoarthritis: a prospective, randomised trial. Int Orthop 33:1275–1278 30. Tai TW, Jou IM, Chang CW, Lai KA, Lin CJ, Yang CY (2010) Non-drainage is better than 4-hour clamping drainage in total knee arthroplasty. Orthopedics 10:156–160
