Chapter 140 Treatment of Hematoma and Hemarthrosis Following Total Knee Arthroplasty
Definition of a Bleed
Bleeding is inherent following total knee arthroplasty because the procedure involves numerous bone cuts, soft tissue releases, and intramedullary violation. Blood loss is well described in other chapters of this text, but some bleeding is expected following total knee arthroplasty (TKA). In blood loss can be predicted, with an average postoperative drop in the hemoglobin of approximately 3 g/dL reported.2 Blood loss can also be reported as amount received in postoperative drainage devices. Numerous studies have shown an estimate of 300 to 1000 mL of blood recovered postoperatively in the salvage drainage devices.11,14
Wound drainage can occur once the hematoma is established. It may present as a knee that continues to have bloody or serous drainage that continues long after the first or second dressing change. It is this continued wound drainage that is most worrisome, with increased wound infection rates when prolonged drainage is allowed to persist.19
Treatment
Preventive Measures
Clamping the drain in the postoperative period has also been evaluated. Tai and colleagues18 have performed a meta-analysis of randomized controlled studies and found that clamping for at least 4 hours may decrease blood loss; no decrease in transfusions, range of motion, or wound complications was reported. Another potential method of decreasing hemarthrosis would be to deflate the tourniquet prior to closure, obtain hemostasis, and then reinflate the tourniquet only if needed. Recently, Rama and associates13 performed a meta-analysis on the timing of tourniquet release during knee arthroplasty. Less blood loss was seen when the tourniquet was deflated at the end of the procedure, but at the expense of more complications. The rate of reoperation in this group was 3.1% for those deflated at the close of the case versus 0.3% who underwent intraoperative tourniquet deflation. Well-conducted large studies are needed to clarify this issue further.
Another method to help reduce hemarthrosis would be to control the hematoma better with a watertight closure. We have recently evaluated a novel wound closure system (Quill Knotless Tissue-Closure Device, Angiotech, Vancouver, Canada) that uses a bidirectional barb with no surgical knots. In the laboratory setting, a more watertight closure was obtained using this barb suture design compared with a traditional intraoperative closure.10 Khanuja8 has recently reported their results with this device, which showed increased strength at the repair site using this type of surgical design. Not only was strength superior to interrupted sutures, but the barbed design was able to withstand suture breakage without loss of the arthrotomy repair. A more watertight closure should be able to contain the hemarthrosis, obtain earlier hemostasis, and avoid the development of a hematoma. Further studies are needed to document the benefits.
In regard to controlling postoperative hematoma, the most important factor would be closure and surgical technique that limit the incidence of lateral release. Johnson and Eastwood6 have described increased swelling and wound drainage when a lateral patellar release is performed. If a lateral release is being considered, the tourniquet should be deflated to confirm the need for lateral release. An effort should also be made to maintain the surrounding soft tissue envelope and try to contain the hemarthrosis as much as possible. If there is direct connection to the joint, the use of subcutaneous drainage should be considered to help decrease blood loss and decrease the dead space noted. Drainage may also play a role in the occurrence of bleeding hematomas. Holt and associates4 have demonstrated increased bruising and wound drainage following total knee arthroplasty when a drain was not used. Thus, the clinical appearance of the wound was less acceptable in patients in whom a drain was not used.
Patel and coworkers12 have evaluated over 1200 total knee arthroplasties and collected data on body mass, intraoperative blood loss, surgical time, type of DVT prophylaxis, and length of hospital stay. They also evaluated the association and duration of postoperative wound drainage in connection with the factors described. They concluded that morbid obesity is a predictor for prolonged wound drainage in total hip arthroplasty patients but no relationship was seen in the total knee arthroplasty patients. The use of low-molecular-weight heparin resulted in an increase of wound drainage in the total hip arthroplasty patients, but this was not seen in the total knee arthroplasty patients. Greater drain output was the only factor and seemed to predict prolonged hospital stay following total knee arthroplasty. Every 100-mL increase of postoperative drainage resulted in an additional 0.03 day of wound drainage compared with an increase of 0.20 day in the total hip arthroplasty patient. Given that each day of prolonged wound drainage increases the rate of infection by 42%, the importance of minimizing this drainage cannot be overstated. Perhaps an effort to control early postoperative bleeding by clamping the drain and placing the knee in a flexed position should be performed to help prevent drainage on subsequent postoperative days.
The relationship of wound drainage to rate of infection is not a new concept. Weiss and Krackow19 have described increased infection rates when wound drainage is allowed to occur and is not surgically addressed. Again, the best treatment is prevention; perhaps high-draining knees should be identified early in the recovery room and addressed with an immediate drain clamp and flexed knee positioning to limit the bleed and prevent the late development of wound drainage and infection.