Tips and pearls: Minimizing blood transfusion following primary total knee replacement

Introduction

Anemia is defined by the World Health Organization as hemoglobin (Hgb) levels less than 12 g/dL in women and less than 13 g/dL in men. Following primary knee replacement most patients display some postoperative anemia. This anemia is the resultant from that fact that patients who undergo a single total knee replacement consistently lose two or more units of blood. This includes the measured blood loss during surgery and that which is lost into the soft tissues, from exposed bone and into the joint after wound closure.

Postoperative transfusion rates vary considerably between hospitals and between surgeons in a given institution. For this Orthopedics Today Round Table discussion, I have asked various joint replacement surgeons to share their experience in minimizing the need for postoperative transfusions. Topics we address include: How to prepare patients for knee replacement surgery to minimize the need for transfusion; what steps are taken intraoperatively to minimize the need for transfusion; and the current trigger levels for transfusion.

There are several reasons to reduce and minimize transfusion rates. These include increased infection rates, possible disease transmission, allergic reactions, potential for a cross-match error, lung injuries, slight increased morbidity and the costs involved.

Douglas W. Jackson, MD
Moderator

Round Table Participants Moderator Douglas W. Jackson, MD Chief Medical Editor Orthopedics Today Long Beach, Calif. Clifford W. Colwell Jr., MD Clinical Professor University of California San Diego School of Medicine San Diego, Calif. Louis M. Kwong, MD, FACS Professor and Vice Chairman Department of Orthopaedic Surgery Medical Director, Orthopaedic Clinics Chief, Orthopaedic Arthritis Service Director, Orthopaedic Research Program Program Director, Residency Training Harbor-UCLA Medical Center Torrence, Calif. Craig J. Della Valle, MD Associate Professor of Orthopaedic Surgery Rush University Medical Center Chicago, Ill. Marshall K. Steele, MD Medical Director Anne Arundel Medical Center Joint Center Annapolis, Md.

Douglas W. Jackson, MD: From your experience with hospitals and surgeons around this country, how much do surgeons know about their postoperative allogenic transfusion rates? How much do they vary?

Marshall K. Steele, MD: Nelson Mandala said, “After climbing a great hill, one finds many more hills to climb.” In the last 40 years since joint replacement was introduced, we have climbed over quite a few hills. Appropriate blood management is not one of them. I say this because in the past 4 years I’ve assessed more than 100 hospitals for excellence, which includes blood management practices. I have found it is rare that an orthopedic surgeon or the hospital has a blood management program in place. Few even know their allogenic blood transfusion rate.

When we implement a Destination Joint Center of Superior Performance we begin tracking and benchmarking a variety of clinical, operational and financial measures including the incidence of blood transfusions. The range between hospitals and surgeons is quite wide. For allogenic transfusions after primary joints our registry indicates a range between 6% and 42% with a mean of 25%.

Jackson: What are the costs associated with transfusion?

Steele: Blood transfusions are very expensive. The greatest costs are the complications you mentioned in the introduction. For the blood itself direct cost is $200 for the material, but that is only 19% of the total costs associated with blood transfusion. Once you add in labor, supplies and overhead and the real costs top $1,000. When you consider that patients who need transfusions have a length of stay nearly a one-half to 1 day longer the costs go up. In fact, Materials and Methods Medicare Provider Analysis and Review (MED Par) data from 2008 showed average per-surgery costs were $3,000 higher in knees and $2,000 higher in hips when a blood transfusion was given.

Jackson: What are the key issues that can be influenced to reduce the transfusion rates in patients?

Steele:The three main issues we can influence include: preoperative anemia, intraoperative blood loss and postoperative care. Anemia is fairly common in the elderly patients we schedule for surgery. In 2001, Kulier reported in Anaesthesist a preoperative incidence of 12%, 15% and 20% for patients in their 60s, 70s, and 80s, respectively. Numerous authors have shown the risk for blood transfusion is directly correlated to the presence of preoperative anemia. With an expected Hgb drop of 4 for hips and 3.8 for knees this makes perfect sense.

In Salido’s 2002 Journal of Bone and Joint Surgery (JBJS) study, 69% who patients undergoing total joint replacement were anemic (Hgb less than 13) received transfusions compared with 13% (Hgb greater than 15).

Another obvious issue is reducing intraoperative blood loss during surgery. A swollen bloody knee is not only at greater risk for transfusion, but complications and slowed recovery. Lastly proper management of patients postoperatively makes a difference as well. Understanding the causes of increased postoperative blood loss as well educating yourself and staff in the compensatory mechanisms and current treatment protocols for blood loss anemia is important.

Jackson: What advice do you have for surgeons interested in climbing this next hill and reducing both the risks and costs of transfusions after joint replacement?

Steele: Like you have to with pain management; be proactive. Develop a total joint blood management program that can be consistently implemented. Read Pierson’s article from JBJS in 2004 in which implementation of his algorithm reduced transfusion rates for knees to 1.4% and for hips to 2.8% without the use of autologous blood or postoperative cell savage. Avoid operating on anemic patients if possible. Set up a system to evaluate every patient for anemia far enough in advance to be able to correct it if possible. This is rarely done in my experience.

Blood work performed 30 days preoperatively will usually suffice to establish a cause and a treatment plan for anemia if present. Use of iron, folic acid and Procrit (epoetin alpha, Centocor Ortho Biotech, Inc.) may be necessary. Be very proactive in reducing intraoperative blood loss. The use of hemodilution, tranexamic acid, modern blood sealing systems, fibrin sprays and avoiding drains have all been shown to reduce blood loss. The use of one or two of these techniques should be considered. When appropriate, avoid anticoagulants that are associated with increased bleeding and higher blood transfusion rates. Don’t rush to transfuse patients who exhibit symptoms of hypotension without assuring proper hydration first. Revising the transfusion trigger to 8 Hbg or below has been shown to be safe in otherwise healthy patients. These are not one-time silver bullet solutions. If you are going to ensure long-term success you must create a physician-led leadership team, decide on the goals and metrics, track results on a quarterly basis, benchmark against others, share this information with your total joint performance improvement team and take action where necessary.

Jackson: The perioperative management of blood transfusions may be impacted by the choice of deep vein thrombosis/pulmonary embolism (DVT/PE) prophylaxes. What are your insights into the different choices available and their potential to increase the postoperative anemia that develops in patients?

Louis M. Kwong, MD, FACS: There are basically two broad classes of VTE prophylaxis strategies: pharmacologic and mechanical. The use of all thromboprophylactic drugs represents a balance between efficacy (clot prevention) and safety (bleeding). Low-molecular-weight heparins (LMWHs), factor Xa inhibitors, and thrombin inhibitors all affect some portion of the coagulation cascade. Aspirin is an inhibitor of platelet aggregation. Because aspirin non-selectively inhibits both cyclo-oxygenase I and II, there is not only a risk of surgical-site bleeding, but also of upper gastrointestinal bleeding.

The only pharmacologic antithrombotics agents currently approved for VTE prophylaxis following TKA in the United States are enoxaparin (Lovenox, Sanofi aventis) and fondaparinux (Arixtra, Glaxo SmithKline) —dalteparin (Fragmin, Esai, Inc.) does not have a TKA indication. The most objective scientific evidence regarding the impact of thromboprophylaxis on postoperative bleeding following TKA comes from randomized, controlled phase 3 pivotal trials. These safety and efficacy studies demonstrate major bleed rates with an average of about 1%. They use active comparators, so there is no comparative information against placebo. However, the last U.S. placebo-controlled thromboprophylaxis trial following TKA completed in the mid-1990s demonstrated a major bleed rate of 4.5% in the placebo group, suggesting that the surgical procedure itself has the greatest impact on bleeding.

Mechanical prophylaxis strategies deal with intermittent pneumatic compression (IPC) and pneumatic plantar compression — foot pumps. These have demonstrated level II scientific evidence supporting their efficacy following TKA in reducing the risk of VTE without an increase in bleeding. They function in not only enhancing venous return, and diminishing peripheral venous pooling by decreasing venous capacitance, but the compressive effects release systemic humeral factors, resulting in an increase in systemic fibrinolysis without incurring an increased bleeding risk. The primary limitation with pneumatic compression is use in the post-discharge setting where compliance may be a concern, as is the case with pharmacologic prophylaxis.

Clifford W. Colwell Jr., MD: Whenever you are using a pharmacologic protocol for the prevention of DVT/PE, be it warfarin, LMWH or fondoparinux, there is a small but meaningful increase in the risk of bleeding. This risk must be measured against the benefit of the anticoagulant. The second iteration of the American Academy of Orthopedic Surgeons guidelines and the ninth of the American College of Chest Physicians are in the process of accessing these risk/benefit ratios as well as dealing with the new oral direct 10A and thrombin inhibitors presently before the FDA. The new modality presently available and reported in the March issue of JBJS is the portable external intermittent compression device with a built-in sensor to coordinate the compressions with your phase of respirations, only compressing during the expiratory phase when your right heart pressure is at the lowest. The device can be used in the hospital in and out of bed and sent home or to a rehabilitation center with the patient.

Jackson: What is your present blood management program in total knee replacement?

Colwell:Although all of the reasons to reduce blood transfusions that you mentioned are correct, many have reduced rates compared to 10 to 20 years ago. Transfusion triggers have changed dramatically, and presently in our setting the number of transfusions in unilateral knee replacements is close to zero. We would not consider a transfusion for Hgb over 8 g/dL unless the patient had serious heart disease. If the initial preoperative Hgb is over 12 g/dL, we do nothing; between 10 g/dL and 12 g/dL, there are options available of auto blood, but only if you carry out the donation 35 days out from the procedure, EPO or wait and watch. If your initial Hgb is under 10 we use a EPO protocol. Often patient’s preferences play a role as well. In our setting we have only five individuals who carry out total knee replacements and do about 800 per year so the procedure has been streamlined and is very efficient. Yes, anticoagulant protocols all have a higher bleeding risk. With the advent of portable devices that can be taken home this playing field is changing as well.

Jackson: What are the considerations in the administration of DVT prophylaxis and the impact on additional bleeding?

Kwong: The timing of administration of the first dose of antithrombotic agent is important with regard to bleeding. While closer administration to the time of wound closure results in a decrease in thrombus formation, there is a corresponding degradation in safety. Conversely, delaying drug administration results in a decrease in bleeding risk, however, the risk of developing VTE goes up. Vitamin K inhibitors such as warfarin typically are associated with a 48-hour delay from the time of ingestion of the drug before there is a measurable clinical effect. As a result, there is frequently a diminished impact on bleeding during the highest at risk period postoperatively. Injectable antithrombotic agents such as LMWHs or fondaparinux — a indirect inhibitor of factor Xa — reach maximum serum concentration approximately 2 hours after administration. As a result, these drugs, which have been consistently found to have superior efficacy compared to warfarin, are more sensitive to the effects of timing. The first dose of enoxaparin is to be administered 12 to 24 hours postoperative; for fondaparinux, 6 to 8 hours postoperative. “Postoperative” means following wound closure or from the time that primary wound hemostasis has been achieved. This last criteria is an important consideration for thromboprophylaxis following complex cases requiring more soft tissue dissection (eg, revision arthroplasty, or cases with severe deformity requiring extensive releases). An indication that primary hemostasis has not been achieved may be reflected by higher drain output in the recovery room. This finding should be considered by the surgeon in revising the time at which the first dose of drug is administered, and would be a better reference against which to judge the initiation of thromboprophylaxis as compared to the time of wound closure.

In order to eliminate any ambiguity as to when the first dose of antithrombotic should be administered, in particular, to avoid the problem of too early administration of drug, the surgeon should explicitly specify in the postoperative orders the time at which the drug is to be administered.

Jackson: What do you do preoperatively to decrease the risk of postoperative transfusions? Do you have your patients donate blood prior to elective primary total knee arthroplasty?

Craig J. Della Valle, MD: Nearly every study that has examined blood loss and transfusion rates following total joint arthroplasty has identified the preoperative Hgb level as the strongest predictor of the need for postoperative transfusions. Therefore, when we schedule a patient for elective total knee arthroplasty, we obtain a full set of labs at that time and based on the patient’s Hgb, age and their general medical status, we make an individualized recommendation for pre-operative blood management.

Basically, if the patient’s Hgb is less than 11 g/dl, they should probably have an anemia work up preoperatively and we refer them to their internist for evaluation. It may seem onerous to check these labs on every patient, however, it identifies people quickly who have abnormalities that preclude elective surgery so last minute cancellations are decreased.

If the Hgb level is between 11 g/dL and 13 g/dL we discuss with them the use of erythropoietin preoperatively. Although there is cost associated with this option, the literature if pretty convincing that it decreases the risk of transfusion. As an alternative, we offer autologous donation of one unit.

For patients with an Hgb between 13 g/dL and 15 g/dL if they are healthy and less than 65 years old, we don’t recommend any specific intervention preoperatively. However if they are less than 65 years old, or have substantial medical comorbidities, we recommend autologous donation.

For patients with an Hgb greater than 15 g/dL, we don’t recommend any specific intervention.

Jackson: Blood loss during the postoperative period may be minimized by intraoperative approaches. Do you deflate the tourniquet before cementing or closing to directly control bleeding?

Della Valle: I don’t deflate the tourniquet prior to closure. It would seem intuitive that deflation of the tourniquet prior to wound closure would give the surgeon the opportunity to identify larger bleeders, particularly posteriorly and near the meniscal attachments leading to less blood loss. The literature, however, suggests that deflation of the tourniquet prior to wound closure is associated with either no difference in total blood loss or an actual increase in perioperative blood loss. It also has been shown to increase operative time.

Although some reports have shown a higher rate of reoperation for bleeding complications if the tourniquet is not released prior to closure, in my own experience of about 1,800 primary total knees, I can only recall one patient who we had to return to the operating room for a hematoma. Some of this may depend, however, on how vigorous the surgeon anticoagulates the patient postoperatively.

One could also make the argument that the release of the tourniquet gives the surgeon a better chance to assess patellar tracking, however I don’t think that it has really been proven that the tourniquet affects excursion of the extensor mechanism in any meaningful way.

Jackson: Do you use any sprays or other applications to minimize postoperative oozing? Do you feel there is any evidence that these reduce transfusion rates in total knee replacements enough to justify their costs?

Della Valle: I have tried several of the topical hemostatic agents, but I personally did not see enough of a difference to justify their cost. Furthermore, I have not yet seen compelling peer-reviewed literature in orthopedic patients with an end point that I felt was relevant to support their use. I have seen data presented orally at meetings regarding the use of these agents, and most of the data I have seen as of yet, other than one study, have not supported a clinically relevant difference in outcomes.

Jackson: Do you do anything else prior to closure that you feel reduces postoperative anemia in your patients?

Della Valle: I use Marcaine combined with epinephrine in most of my primary total knee arthroplasties. In addition to the analgesic effects, there are at least two studies that I am aware of that suggest a beneficial effect of Marcaine with epinephrine in terms of Hgb decreases post-operatively.

Epinephrine is an inexpensive medication that is easily administered and pretty widely available, but surgeons should recognize that Marcaine can have systemic effects including cardiac toxicity so surgeons should discuss with their pharmacy and anesthesia staff what is the largest volume that can be used safely. We typically limit the volume to 60 cc of 0.5% Marcaine with 1:200,000 of epinephrine.

Jackson: What historical information and laboratory studies do you obtain from the patient to minimize postoperative anemia?

Kwong: The preoperative history and assessment of the patient play prominent roles in reducing the risk of postoperative anemia. With regard to medical comorbidities, a history of peptic ulcer disease, liver disease, or renal dysfunction all increase the risk of postoperative anemia due to potential enteric blood loss, auto-anticoagulation, and diminished renal clearance of antithrombotics (leading to drug accumulation in the case of LMWHs and fondaparinux) respectively.

Drug history is critical, as the use of antiplatelet drugs is very common in the general population. Similarly, the use of traditional NSAIDs also leads to prolongation of bleeding time due to nonreversible inhibition of platelet aggregation.

In the case of warfarin use in the atrial fibrillation patient, cessation of warfarin must be integrated with preoperative bridging with enoxaparin or heparin. The discontinuation of these drugs in advance of surgery needs to be coordinated with the primary care physician. A careful drug inventory must also include specific questioning regarding the use of herbal supplements. Popular supplements such as gingko — and even ginger and garlic — can increase the risk of postoperative bleeding, and should be stopped at least 2 weeks in advance of surgery. Essential preoperative laboratory information must include not only a CBC and coagulation parameters (PT/INR, PTT, platelet count), but also a basic metabolic panel to assess renal function, as well as liver function tests if warranted based on the clinical history.

We must keep in mind that, by the very nature of surgery, we can only reduce the risk of bleeding leading to postoperative anemia and cannot reasonably expect to reduce the bleeding risk to zero.

References:

• Colwell CW, Friomson MI, Mont MA, et al. Thrombosis prevention after total hip arthroplasty. A prospective, randomized trial comparing a mobile compression device with low-molecular-weight heparin. J Bone Joint Surg (Am). 2010;92(3):527-535.
• Kulier A, Gombotz H. Perioperative anemia. Anaesthesist. 2001;50(2):73-86.
• Peirson JL, Hannon TJ, Earles DR. A blood-conservation algorithm to reduce blood transfusions after total hip and knee arthroplasty. J Bone Joint Surg (Am). 2004;86(7):1512-1518.
• Salido JA, Marín LA, Gómez LA, et al. Preoperative hemoglobin levels and the need for transfusion after prosthetic hip and knee surgery: analysis of predictive factors. J Bone Joint Surg (Am). 2002;84(2):216-220.

• Clifford W. Colwell Jr., MD, can be reached at N Torrey Pines Road, MS 116, La Jolla, CA 92037; 858- 554-7993; e-mail: cliffcolwellmd@yahoo.com.
• Craig J. Della Valle, MD, can be reached at the Department of Orthopaedic Surgery, Rush University Medical Center, 1611 West Harrison, Suite 300, Chicago, IL 60612; 312-432-2350; e-mail: craigdv@yahoo.com.
• Douglas W. Jackson, MD, is the chief medical editor of Orthopedics Today. He can be reached at ot@wyanokegroup.com.
• Louis M. Kwong, MD, FACS, can be reached at Department of Orthopedic Surgery, Harbor/UCLA Medical Center, 1000 W Carson St, Bin 422, Torrance, CA 90509; e-mail: lmkwongmd@hotmail.com.
• Marshall K. Steele, MD, can be reached at 410-271-1785; or e-mail: msteele@marshallsteele.com.

Follow OrthoSuperSite.com on Twitter