Anesthesia for total knee arthroplasty

INTRODUCTION — Total knee arthroplasty (TKA) is a very common operation that is increasing in frequency as the population ages [1]. Similar to other surgical procedures, there is a trend towards multidisciplinary optimization of perioperative care to minimize morbidity and mortality while allowing for rapid recovery and early hospital discharge [2]. Perioperative anesthesia care is an important component in an opioid-sparing enhanced recovery pathway designed to minimize pain while maximizing early mobilization.

This topic will discuss the management of anesthesia and analgesia for TKA. The surgical procedure is discussed separately. (See "Total knee arthroplasty" and "Surgical management of end-stage rheumatoid arthritis".)

PREOPERATIVE EVALUATION — Preoperative evaluation should include a medical history and anesthesia-directed physical examination, including airway assessment and examination of any site for regional anesthesia, if warranted. Patients presenting for TKA often have comorbidities, such as osteoarthritis, spinal stenosis, ankylosing spondylitis, severe obesity, rheumatoid arthritis, and those that accompany advanced age and may affect anesthesia care. Preoperative medical evaluation, including medication management and preoperative risk assessment, is discussed more fully separately. (See "Preoperative evaluation for anesthesia for noncardiac surgery".)

The preoperative evaluation process varies by institution and may be standardized as part of a practice pathway. TKA is almost always an elective procedure, and there should be opportunities for prehabilitation and shared decision-making regarding anesthetic and analgesic choices well in advance of surgery. Issues related to chronic pain should be explored, including prior or ongoing opioid use, patient expectations for postoperative pain management, and the psychological aspects of chronic pain. All of these issues may require modification of the postoperative pain control regimen. (See "Management of acute pain in the patient chronically using opioids for non-cancer pain" and 'Plan for perioperative multimodal pain control' below.)

We routinely perform a complete blood count and basic metabolic panel prior to TKA. Further laboratory testing should be based on patient comorbidities and the likelihood that results will affect management. (See "Preoperative medical evaluation of the healthy adult patient", section on 'Laboratory evaluation' and "Preoperative evaluation for anesthesia for noncardiac surgery", section on 'Preoperative testing'.)

The need for blood type and screen prior to TKA depends on comorbidities and expected blood loss. In our experience, blood transfusion is rarely required for primary uncomplicated TKA. We perform type and screen for patients at increased risk for transfusion (ie, complex revision, preoperative anemia) and type and crossmatch blood selectively, based on patient factors and the planned surgery. Blood loss during TKA is discussed below and in other topics. (See 'Surgical issues that affect anesthesia' below and "Total knee arthroplasty", section on 'Minimizing perioperative blood loss' and "Complications of total knee arthroplasty", section on 'Blood loss'.)

Surgical preoperative evaluation for TKA is discussed separately. (See "Total knee arthroplasty", section on 'Preoperative evaluation'.)

PLAN FOR PERIOPERATIVE MULTIMODAL PAIN CONTROL — TKA causes moderate to severe postoperative pain for most patients. The goals for pain control after TKA are to provide excellent analgesia, early mobilization and rehabilitation, and to minimize the use of opioids, including overprescription of opioids after discharge [3].

The evidence comparing specific options for pain control after TKA is mixed and therefore inconclusive, and likely reflects heterogeneous institutional pain management protocols and peripheral nerve block techniques. We use a multimodal approach to pain control that includes one or more peripheral nerve blocks, periarticular infiltration of local anesthetics (LAs), and multiple systemic analgesics. A multimodal approach to pain management allows administration of lower doses of various medications and less dense regional anesthesia techniques than would be required otherwise, and therefore reduced side effects and complications [4,5]. Properly-managed acute pain may prevent persistent postoperative pain after TKA [6], and may reduce the long term use of opioids after TKA [7].

Intravenous (IV) opioid patient-controlled analgesia (PCA) should be avoided if possible for most patients, because of opioid related side effects (ie, nausea, pruritus, sedation, respiratory depression), and to allow early mobilization.

Multimodal, opioid-sparing strategies for postoperative pain control may include regional analgesia techniques (eg, peripheral nerve blocks, LA infiltration, continuous epidural analgesia, neuraxial opioids), in addition to multimodal systemic analgesics. A comprehensive pain control strategy optimally begins preoperatively, with administration of individualized preemptive oral medications (eg, acetaminophen, cyclooxygenase-2 [COX-2] inhibitors, nonsteroidal antiinflammatory drugs, gabapentinoids, oral opioids for those with preoperative opioid dependence), which are then continued throughout the hospitalization and after discharge. The multimodal analgesia strategy used by the authors is shown in a table (table 1). (See "Approach to the management of acute pain in adults".)

Importantly, the plan for both preoperative and postoperative use of anticoagulants and antiplatelet medications for thromboembolism prophylaxis should be determined preoperatively. (See 'Surgical issues that affect anesthesia' below.)

Many institutions have developed their own pathways or protocols for perioperative care that include pain control, prevention of postoperative nausea and vomiting, and other aspects of perioperative care; the optimal components of such pathways for TKA are debated [8]. Even within established pathways, strategies for pain control should be individualized based on patient factors (eg, prior opioid use, comorbidities) and the type of surgery (eg, revision or complicated surgery, unilateral versus bilateral). Our approach to pain control after TKA is as follows:

●Healthy, opioid naïve patients who undergo unicompartmental, primary, or simple revision TKA are often candidates for rapid recovery with the potential for "fast-track" status with a reduced length of stay. For some patients, same-day discharge may be possible. Pain control strategies should include a muscle sparing regional anesthetic technique (adductor canal, interspace between popliteal artery and posterior capsule of the knee [IPACK], periarticular injection [PAI]) to facilitate early physical therapy. An example of such a protocol is shown in a figure (figure 1).

●Patients on chronic opioids or those with high levels of pain-related anxiety will likely be poor candidates for fast-track status and require a tailored regional anesthetic plan (eg, continuous perineural catheter or more comprehensive sensory blockade), higher doses of perioperative opioids, or assistance with coping strategies for psychological conditions (eg, pain catastrophizing behavior). (See 'Preoperative evaluation' above and 'Choice of peripheral nerve block' below.)

●Patients who undergo complex joint reconstruction may require a regional anesthetic technique that blocks sensation from the entire lower extremity (eg, continuous epidural analgesia, combined femoral and sciatic blocks), or long acting neuraxial opioids in addition to the peripheral nerve blocks that are usually used. Such analgesic regimens may prevent aggressive postoperative mobilization and rehabilitation.

REGIONAL ANESTHESIA TECHNIQUES FOR POSTOPERATIVE PAIN CONTROL — Regional analgesia techniques for postoperative pain control after TKA include periarticular LA infiltration, peripheral nerve blocks, and neuraxial analgesia techniques (eg, continuous epidural analgesia, neuraxial opioids).

Periarticular injection (PAI)/local infiltration analgesia (LIA) — Periarticular injection (PAI), also called local infiltration analgesia (LIA), is increasingly used as a component of multimodal postoperative analgesia [9,10]. PAI involves a series of injections performed by the surgeon intraoperatively into areas of the knee that are known to have concentrated pain fibers. The injectate solution and the adjuvants in the solution may vary among institutions. Certain additives appear to improve the analgesic effect (eg, ketorolac [11]); whereas the choice of long-acting LA (bupivacaine versus ropivacaine versus liposomal bupivacaine) added does not [12,13].

Our technique involves injecting the posterior capsule prior to cementing using multiple capsular punctures beginning in the midline and proceeding medially to avoid peroneal nerve blockade, then an injection into the synovium overlying the distal femur medially, laterally, in the suprapatellar pouch, and medial and lateral retinaculum. The final injection site is into the subcutaneous tissue [12]. Our standard PAI solution is shown in a table (table 2).

Use of PAI with peripheral nerve blocks is discussed below. (See 'Nerve blocks plus PAI' below.)

If PAI is performed in addition to peripheral nerve block, the total dose of LA should be calculated to avoid excessive dosing and to minimize the risk of local anesthetic systemic toxicity (LAST). (See "Local anesthetic systemic toxicity".)

Peripheral nerve blocks — Peripheral nerve blocks have been shown to reduce pain scores and opioid consumption after TKA, with or without systemic analgesics [14-17]. Use of peripheral nerve blocks may also reduce some postoperative adverse events. In a meta-analysis performed by the International Consensus on Anesthesia-related Outcomes (ICAROS) group, use of peripheral nerve blocks was associated with reduced odds of cognitive dysfunction, respiratory failure, cardiac complications, surgical site infections, thromboembolism, and blood transfusion [18]. This led the ICAROS group to recommend use of peripheral nerve blocks for TKA. It should be noted that the ICAROS review did not have access to outcomes data related to PAI, intrathecal opioids, or data in enough granular detail to recommend specific types of nerve blocks.

The authors prefer to place peripheral nerve blocks prior to surgery to reduce intraoperative opioid administration and reduce post-anesthesia care unit (PACU) recovery time. We also prefer to place nerve blocks in adults while they have intact sensation in the area of the block (ie, prior to neuraxial anesthesia) to allow patient report of pain or paresthesia during the block. However, the need for intact sensation during regional anesthesia is debated and practice varies [19,20]; in some institutions nerve blocks are placed in the PACU, with neuraxial anesthesia still in effect.

Relevant nerve and muscle anatomy — Comprehensive analgesia for TKA involves blocking some or all of the sensory nerves that innervate the knee joint, depending on patient factors and the planned surgery.

The majority of the innervation to the knee joint arises from the femoral nerve (nerves to the vastus medialis, intermedius, and lateralis, medial and intermediate femoral cutaneous, and saphenous nerves), a smaller but still significant contribution from the sciatic nerve (peroneal and tibial nerves), and yet even smaller contribution from the lateral femoral cutaneous and posterior obturator nerves [21]. The saphenous nerve is the sensory component of the femoral nerve (figure 2).

The femoral nerve block anesthetizes all of the sensory innervation to the knee except for branches of the sciatic and obturator nerves, including sensorimotor nerves to the vastus medialis, vastus intermedius and vastus lateralis, and the sensory saphenous nerve (figure 3).

The ACB involves block of the saphenous nerve (sensory only), likely the nerve to the vastus medialis, and possibly the posterior branch of the obturator nerve [22]. It is classically performed at the level of the mid-thigh, midway between the anterior superior iliac spine and the base of the patella [23]. However, the extent and location of the adductor canal is debated by anatomists, and this point on the thigh may actually overlay the inferior aspect of the femoral triangle. ACB results in less quadriceps weakness than femoral nerve block because the ACB generally spares the nerves to the rectus femoris and vastus intermedius muscles. (See "Adductor canal block procedure guide".)

If a large volume of LA is injected distally in the adductor canal, it can spread into the popliteal fossa, affect the sciatic nerve, and cause foot drop [24]. Conversely, a large volume of LA injected proximally can spread to the femoral nerve and cause quadriceps weakness [25,26].

Motor block and patient falls — The degree to which peripheral nerve blocks and the associated motor weakness increase the risk of patient falls is controversial. Patient falls after orthopedic surgery have been associated with fracture, return to the operating room, major cardiac, pulmonary and thromboembolic events, and mortality [27]. Inpatient falls have been designated preventable "never events" by the Center for Medicare and Medicaid Services in the United States. Whereas peripheral nerve blocks can cause quadriceps weakness, other factors (eg, advanced age, effects of surgery, comorbidities, sedatives) and pain itself may be equally important contributors to fall risk [27-30].

One database study included 191,570 records for TKAs performed in approximately 400 hospitals between 2006 and 2010 [27]. There was no difference in the incidence of inpatient falls in patients who had peripheral nerve block compared with patients who did not have a block (1.58 percent versus 1.6 percent). The analyzed data did not include information on the type of block, dose and concentration of LA, hospital fall prevention protocols, or other potentially confounding factors that may have influenced the results.

ACBs result in less quadriceps weakness than femoral nerve blocks, but have not been shown to reduce the risk of falls [28]. Thus, fall prevention strategies are required for all patients undergoing TKA, as the patient population that undergoes TKA is at risk for falling regardless of peripheral nerve blockade [31,32].

Choice of peripheral nerve block — We include an individualized multimodal analgesic approach with combinations of blocks. We use periarticular PAI plus single injection ACB for uncomplicated primary TKA. We use continuous peripheral nerve blocks selectively (eg, for patients with opioid tolerance, opioid intolerance, pain catastrophizing, allergy to nonopioid analgesics).

Femoral nerve block has historically been used for postoperative analgesia for routine unilateral primary TKA; femoral block is easy to perform and blocks most sensation to the knee. However, the ACB is increasingly preferred because it is as effective as femoral block and is associated with less motor block than femoral block [23,28,33-36] (see 'Relevant nerve and muscle anatomy' above). Our practice is to combine the single injection or continuous ACB with PAI for routine primary TKA, in addition to multimodal analgesics, particularly for patients in a fast-track protocol.

Neither the femoral nerve block nor the ACB provide complete analgesia of the knee, because they do not include the contributions from the sciatic nerve. For patients who undergo complex revision, who are poor candidates for fast-track protocol, or who require complete analgesia of the knee because of opioid tolerance or chronic pain, we perform femoral plus sciatic nerve blocks.

●Adductor canal block – Both single injection and continuous ACBs may be used for analgesia after TKA. We perform a low volume ACB with 10 to 15 mL of bupivacaine 0.25% with 1:200,000 epinephrine.

If a continuous ACB is used, the tourniquet will likely be placed over the catheter insertion site, and the catheter must be taped proximally, under the tourniquet and away from the surgical field. For continuous block, we use an infusion of 0.1% bupivacaine at 6 to 8 mL per hour, with removal of the catheter on postoperative day 2 at 6:00 AM or prior to discharge, whichever is sooner. (See "Adductor canal block procedure guide".)

●Femoral nerve block – Femoral nerve block is simple to perform as either a single injection or catheter technique. The authors often choose femoral block over an ACB for patients who will require a more comprehensive postoperative pain control, and who require denser anatomical regional coverage (eg, patients with chronic pain syndromes, consumption of large quantities of opioids, or who undergo extensive surgical procedures). We prefer to place a perineural catheter for femoral block for such patients because it allows prolonged pain control, potentially for multiple days. We initiate the block with a bolus of 20 to 30 mL of 0.25 or 0.5% bupivacaine with 1:200,000 epinephrine and continue an infusion of 0.1 to 0.2% bupivacaine at 8 to 10 mL/hour until postoperative day 2 at 6:00 AM or discharge, whichever is sooner. (See "Femoral nerve block procedure guide".)

Femoral block may be combined with a PAI that focuses on the posterior capsule of the knee; however, total cumulative dose of LA should be monitored.

●Sciatic nerve block – Sciatic nerve block combined with femoral nerve block provides excellent pain control after TKA, but is limited by associated motor weakness (foot drop), which may mask surgery-related peroneal nerve injury. As an alternative, a careful selective block of the tibial nerve (a branch of the sciatic nerve) at the popliteal fossa may provide similar pain control to a complete sciatic block while sparing the peroneal nerve [37]. (See "Sciatic blocks procedure guide".)

Sciatic nerve block is associated with a longer duration of anesthesia than other single injection blocks [38] (>24 hours) and postoperative neuropathy, particularly in patients with small vessel disease [39]. In an effort to minimize these risks, we perform a single-injection block with low concentrations of bupivacaine (0.25%) and epinephrine (1:400,000).

We perform sciatic block for patients undergoing a complex revision or for those who require complete analgesia of the surgical extremity, accepting the risk that this may impede early postoperative rehabilitation. (See 'Plan for perioperative multimodal pain control' above.)

●Lumbar plexus block – Lumbar plexus block is an advanced technique usually reserved for only complex TKA revisions that involve hardware or bony work involving the proximal thigh, precluding use of more distal regional analgesia approaches. It should only be performed by experienced clinicians and is subject to the same restrictions as neuraxial techniques with respect to anticoagulation. Lumbar plexus block is generally not indicated for analgesia for TKA. A three arm study comparing femoral nerve catheter versus lumbar plexus catheter versus femoral and sciatic nerve catheter for TKA reported no difference in functional outcomes and superior pain control with combined femoral and sciatic catheters [40]. (See "Lower extremity nerve blocks: Techniques", section on 'Lumbar plexus (psoas compartment) block' and "Neuraxial anesthesia/analgesia techniques in the patient receiving anticoagulant or antiplatelet medication".)

●Interspace between popliteal artery and posterior capsule of the knee (IPACK) block – The IPACK block is designed to block the small sensory branches of the sciatic nerve that travel in this space without affecting motor components. The IPACK block provides analgesia for the posterior portion of the knee only, and would likely be ineffective alone for postoperative analgesia; thus it is often performed along with an ACB in a multimodal analgesic pathway [41]. The IPACK block may be an alternative to PAI or provide additional coverage for posterior knee pain [42,43]. (See 'Periarticular injection (PAI)/local infiltration analgesia (LIA)' above and 'Nerve blocks plus PAI' below.)

Although this block is performed at several institutions, data are still emerging on its efficacy and is mixed. Several randomized trials have found reduced postoperative pain after TKA when IPACK was added to a multimodal analgesic regimen that included ACB [42,44,45]. However, other randomized trials have not found benefit from adding IPACK to PAI plus continuous adductor canal block [43,46]. (See 'Nerve blocks plus PAI' below.)

●Obturator nerve block – The analgesic benefits of adding an obturator nerve block into a multimodal analgesic pathway are debated, as some studies have shown a modest improvement in postoperative pain control [47,48], while others have shown no added benefit [49]. In our practice, we do not routinely perform this block because of the potential for motor weakness and the required anesthesia resources. (See "Lower extremity nerve blocks: Techniques", section on 'Obturator nerve block'.)

Nerve blocks plus PAI — The literature on the benefits of adding peripheral nerve blocks to PAI is conflicting; results may be affected by block technique, institutional protocols, and other components of multimodal analgesia. Several single institution studies have found improved analgesia or functional outcome with ACB or ACB plus interspace between popliteal artery and posterior capsule of the knee (IPACK) block in addition to PAI, compared with PAI alone [9,44,50,51], whereas one study reported no difference in postoperative pain or opioid consumption with the addition of ACB to PAI [52].

At the authors' institution, IPACK is rarely performed in our fast-track protocol (which includes single-injection ACB and PAI) as our surgeons have had extensive experience with performing PAI and consistent success in covering posterior knee pain. However, adding IPACK to a multimodal fast-track protocol should be considered if surgeons do not perform PAI [43], or if their technique does not consistently cover posterior knee pain. A 2021 meta-analysis of trials that evaluated the impact of adding IPACK to ACB in patients who underwent TKA found that adding IPACK to ACB did not improve analgesia in patients who had PAI (4 studies, 273 patients) [53]. In the absence of PAI, adding IPACK to ACB reduced pain for 24 hours and improved functional recovery (8 studies, 631 patients). (See 'Choice of peripheral nerve block' above.)

Continuous versus single injection nerve blocks — Single-injection or continuous nerve blocks may be placed for analgesia after TKA. Single-injection blocks are less invasive, avoid the need for infusion pumps that may restrict mobility, and may reduce resource utilization and cost. However, the duration of analgesia is limited after single-injection block, and rebound pain may occur when the block wears off. We usually place a perineural catheter for patients with chronic pain or opioid use, or for complex surgical procedures.

The literature comparing efficacy of continuous versus single-injection peripheral nerve block is inconclusive for both the femoral and adductor canal blocks. Studies comparing continuous femoral with continuous ACBs have generally reported improved postoperative mobilization with ACB, but conflicting results with respect to analgesia. Some studies have reported better analgesia with continuous femoral nerve block [54,55], while others have reported no difference [56,57].

●Femoral block – In a meta-analysis of four randomized trials that compared continuous with single-injection femoral block for TKA, continuous block was associated with a small reduction in pain at rest and with movement at 24 hours, and reduced opioid consumption at 24 hours [17]. In contrast, in a randomized trial that compared continuous with single-injection femoral block in patients who received intrathecal morphine for TKA, there was no difference in pain scores or postoperative opioid consumption over 72 hours [58]. Similarly, in a randomized trial in patients who received multimodal analgesia for TKA, there was no difference in postoperative pain scores or opioid consumption in patients who had continuous femoral block versus single-injection block [59].

●Adductor canal block – In a 2022 meta-analysis of randomized trials of patients who underwent TKA, pain scores (4 trials) and opioid consumption (8 trials) in the first 48 hours were similar in patients who had continuous versus single-injection ACB [60]. Conclusions may be limited by varied analgesic regimens aside from the ACB and lack of information on the technical aspects of continuous catheter placement, which could affect catheter migration and analgesic efficacy [61]. (See "Adductor canal block procedure guide", section on 'Catheter placement technique'.)

A 2023 network meta-analysis of 27 randomized trials evaluated the analgesic efficacy of single-injection or continuous ACB, with or without single-injection IPAK block in patients undergoing TKA who also received PAI [62]. Of the various block combinations, continuous ACB with PAI provided better analgesic and functional outcomes between 24 and 48 hours after surgery than single injection ACB or the addition of IPACK. However, the benefits were of questionable clinical significance when compared with PAI alone.

The use of adjuvant medications to prolong the effect of LA for peripheral nerve blocks is discussed separately. (See "Overview of peripheral nerve blocks", section on 'Adjuvants'.)

Emerging techniques for pain control after TKA — Each of the following techniques has been suggested for analgesia after TKA. Further study is required before recommending any of them for routine use.

●Radiofrequency ablation – Radiofrequency ablation (RFA) of sensory nerves has been reported for treatment of some chronic pain conditions, and preoperative RFA has also been described for reducing pain after TKA, with unclear benefit [63]. Further study is required before recommending the use of this technique for postoperative analgesia for TKA.

●Peripheral nerve stimulation – Peripheral nerve stimulation (PNS) is an emerging therapy for postoperative analgesia after TKA; PNS has been used for a variety of chronic pain conditions. Percutaneous PNS delivers electrical stimulation to peripheral nerve fibers through leads placed near the femoral and sciatic nerves, and connected to an external pulse generator [64]. Purported advantages of PNS after TKA are the possibility of opioid avoidance, and lack of motor block.

●Genicular nerve and popliteal plexus blocks – Both of these peripheral nerve blocks are described in case reports and small case series for analgesia after TKA [65,66]. In a randomized trial of 40 patients who underwent TKA, addition of genicular nerve block to IPACK plus continuous ACB reduced opioid consumption at 24 hours compared with sham block (mean 23 ± 20 versus 58 ± 35 morphine milligram equivalents) [67]. Further study is required before recommending routine use of such a protocol, which involves three peripheral nerve blocks.

●Cryotherapy/Cryoneurolysis – Cryotherapy refers to the use of cooling devices to lower the temperature of body tissues. It can be used to manage pain and inflammation following injury and surgery. Examples of cryotherapy devices include ice packs and circulating water cuffs, with or without compression. Analgesic benefits of these types of cryotherapy are unclear. A 2023 systematic review of the literature found 6 trials that evaluated the use of cryotherapy after TKA [68]. Cryotherapy reduced opioid consumption in the first postoperative week (3 trials), without clear reductions in pain scores. Conclusions are limited by significant heterogeneity among the studies and lack of data on other methods of pain control.

Cryoneurolysis is a method of cooling peripheral sensory nerves to produce reversible nerve damage. It has been used to treat knee pain from osteoarthritis and has been studied as preoperative treatment before TKA. In a single institution unblinded industry-sponsored randomized trial of 124 patients undergoing TKA, preoperative cryoneurolysis of the superficial genicular nerves decreased opioid consumption, improved function, and reduced pain scores at 6 weeks compared to standard multimodal analgesic care [69]. Further study independent of industry is required before recommending widespread use of cryoneurolysis prior to TKA.

Neuraxial analgesic techniques

Epidural analgesia — For most patients who undergo unilateral primary TKA, epidural analgesia provides no additional analgesic benefit compared with peripheral nerve blocks or PAI [70-75]. Epidural analgesia is associated with hypotension and unnecessary bilateral block, which may delay mobilization and hospital discharge.

Continuous epidural analgesia may be beneficial for patients who undergo complex revision or bilateral TKA. (See "Continuous epidural analgesia for postoperative pain: Technique and management".)

Neuraxial opioids — Long-acting neuraxial opioids (eg morphine, hydromorphone) may be part of multimodal analgesia after TKA, and provide modest reduction in postoperative opioid requirements. In one trial, approximately 190 patients were randomly assigned to PAI, PAI with ACB, or PAI, ACB, and intrathecal morphine, for analgesia after TKA, along with oral multimodal analgesics [76]. Patients who received intrathecal morphine had lower pain scores at 12 hours postoperatively and required lower opioid doses up to 48 hours than patients in the other two groups; patients who received ACB with PAI had intermediate opioid requirements and pain scores.

Despite a modest analgesic benefit, we avoid long acting neuraxial opioids for routine TKA because of the risk of delayed respiratory depression and the need for enhanced respiratory monitoring for patients who receive these drugs [77]. (See "Continuous epidural analgesia for postoperative pain: Technique and management", section on 'Monitoring during epidural analgesia' and "Post-cesarean delivery analgesia", section on 'Side effects and complications'.)

Neuraxial opioids may be added to a multimodal analgesic regimen for patients who undergo complex TKA or are expected to have challenging pain control.

SURGICAL ISSUES THAT AFFECT ANESTHESIA — The patient is positioned supine for TKA and a midline incision is performed. The diseased articular surface is resected, followed by insertion of femoral, tibial, and sometimes patellar components. A mid-thigh tourniquet is usually used to create a bloodless field for cementing the prosthesis. Muscle relaxation is not required for TKA. (See "Total knee arthroplasty", section on 'Total knee arthroplasty procedure'.)

Thromboprophylaxis — The use of antithrombotic and/or antiplatelet medications may affect the use of neuraxial anesthesia and analgesia and performance of deep peripheral nerve blocks because of the risk of spinal epidural hematoma and other bleeding [78]. Postoperative removal of an epidural catheter requires coordination with the surgical service if anticoagulants are administered, including clear lines of responsibility for early investigation of neurologic findings that might indicate spinal epidural hematoma [79]. (See "Neuraxial anesthesia/analgesia techniques in the patient receiving anticoagulant or antiplatelet medication" and "Prevention of venous thromboembolism in adults undergoing hip fracture repair or hip or knee replacement" and "Perioperative management of patients receiving anticoagulants".)

Nerve injury — The potential for nerve damage related to surgery or use of a tourniquet should be discussed by the surgeon with the patient prior to selection for regional anesthesia techniques, to avoid confusion should nerve injury occur. Nerve injury is much more likely to be related to surgery or the use of a tourniquet than a nerve block. Nonetheless, nerve injury after orthopedic procedures is a significant malpractice concern for anesthesia clinicians. A study of malpractice claims from 2000 to 2013 reported that claims arising from nerve injury related to the use of regional anesthesia for non-spine orthopedic procedures were more frequent than claims for other areas of anesthesia [80].

The most common neurologic complication after TKA is peroneal nerve palsy, particularly in patients with severe valgus alignment in combination with a flexion deformity [81,82]. It may be wise to avoid sciatic nerve block in patients at high risk of peroneal nerve injury, since the peroneal nerve is a branch of the sciatic nerve. Nerve injury associated with TKA is discussed separately. (See "Overview of peripheral nerve blocks", section on 'Nerve injury' and "Complications of total knee arthroplasty", section on 'Peroneal nerve palsy'.)

Blood loss — Reported perioperative blood loss for TKA varies widely, from approximately 650 to 1400 mL [83]. The reported rates of transfusion also vary widely, and depend on the patient population and transfusion trigger. Historic national averages report a transfusion rate of 18.3 percent for a primary TKA [84]; however, with contemporary transfusion guidelines and perioperative protocols to reduce bleeding (ie, tranexamic acid [TXA] and local infiltration analgesia) this rate can be reduced to almost zero [85]. Significant patient risk factors for transfusion include: age, preoperative anemia, female sex, body mass index (BMI) <30 kg/m², and American Society of Anesthesiologists (ASA) status >2 [84]. Intravenous (IV), topical, oral, and combined administration routes of TXA are increasingly used to reduce perioperative blood loss and the need for transfusion [86]. (See 'Antifibrinolytics' below.)

Intraoperative blood loss is generally low, particularly if the surgeon leaves the tourniquet inflated until the wound is closed and a compressive dressing is applied. Postoperative blood loss into surgical drains is often at least twice the intraoperative blood loss, even if the tourniquet is released intraoperatively [87,88]. (See "Complications of total knee arthroplasty", section on 'Blood loss'.)

Antifibrinolytics — Antifibrinolytic agents (eg, TXA, epsilon-aminocaproic acid [EACA]) are widely used for reducing blood loss during orthopedic surgery. TXA is the most studied and most commonly used antifibrinolytic agent for TKA, due to excellent serum-to-joint space diffusion pharmacokinetics [89]. We recommend administration of TXA for patients without a baseline high risk of thromboembolic events (eg, prior venous thromboembolism [VTE], any prior arterial thromboembolic event [ATE] such as myocardial infarction, cerebrovascular disease, or vascular stent) who undergo TKA. If TXA is unavailable, EACA is a reasonable alternative, though the data supporting its use for TKA is limited [90,91]. We agree with the 2018 guidelines issued jointly by orthopedic and regional anesthesia societies, strongly supporting the use of TXA in patients without risk factors for VTE who undergo primary total joint arthroplasty, including TKA [92]. For patients with a high baseline risk of VTE, the decision to administer TXA should be individualized, based on assessed risks of thrombosis with TXA versus the risks of transfusion without it.

Efficacy of TXA — Multiple trials have reported reductions in blood loss and transfusion with IV, topical, or oral administration of TXA [93-103]. A 2011 meta-analysis of 14 randomized placebo controlled trials (total 824 patients) of TXA administration for TKA found that TXA administration was associated with a reduction in the proportion of patients who required blood transfusion (23.5 versus 47.1 percent, risk ratio [RR] 2.6, 95% CI 2.1-3.1) [97]. Based on nine trials, TXA was also associated with reduced total blood loss (mean difference 591 mL, 95% CI 536-647 mL). However, there was substantial heterogeneity among the trials, and most trials were small. A 2015 meta-analysis of 13 randomized trials found that administration of TXA reduced the rate of transfusion of at least one unit of red blood cells by 50 percent [104].

Dosing regimen — The optimal route, timing, and dose for administration of TXA has not been determined, and practice varies. For simplicity, assurance of adherence, and standardization of practice, the authors administer TXA 1 g IV before incision and 1 g IV during closure, for all patients who are candidates for TXA. Other contributors administer 10 mg/kg IV (maximum dose 1 g IV) before incision, and repeat that dose during closure. Doses should be reduced for patients with renal dysfunction, lower body weight, or malnutrition.

The majority of studies on the use of TXA have involved IV administration. Available studies that have compared different routes of administration have reported clinically insignificant or non-inferior differences in efficacy. Examples of such studies that have evaluated different routes of administration include the following:

●A network meta-analysis of 67 studies of efficacy of TXA for TKA found that IV, topical, and oral routes of administration effectively reduced blood loss, without superiority of a particular route, dose, or number of doses [105]. This study found that preincision IV administration may improve efficacy, but requires further study.

●In a single institution randomized trial of 387 patients who underwent TKA or THA, IV (1000 mg) and oral (1950 mg) administration of TXA were equally effective at preventing blood loss and transfusions, without a difference in adverse effects [106]. Importantly, oral TXA was administered in the hospital two hours prior to surgery, which ensured that patients received the dose. Switching from IV to oral TXA for total joint arthroplasty could be a significant cost-saving initiative, estimated at $60 to180 million in the US annually by 2030. However, implementation may be problematic. In this trial, three patients withdrew consent due to the inability to take the large-sized oral TXA tablets prior to surgery. Whether complete compliance with preoperative oral dosing can be relied upon in other settings is unclear.

●In a large randomized trial of 640 TKA patients, the risk of receiving a blood transfusion was higher with topical TXA relative to intravenous TXA, after controlling for demographic factors on multivariate analysis (odds ratio [OR] 2.2, 95% CI 1.83-2.67) [107]. IV TXA was associated with less calculated blood loss and lower drain output. However, only seven patients received transfusions in this trial, underscoring that either method of TXA administration results in low transfusion rates, and the differences between them may not be clinically important.

●Topical and IV administration of TXA were compared in a randomized trial of 76 patients who underwent TKA [108]. Patients received a single dose of topical TXA in the wound prior to closure, or two doses of IV TXA, one prior to tourniquet inflation, and another three hours later. Thrombogenic potential, as measured by blood levels of a thrombogenic marker, were similar between groups. Fibrinolytic activity was similar between groups after the first IV dose, but higher in the IV group after the second dose. Calculated blood loss was lower in the IV group, but there was no difference in drain output or in blood transfusion.

Risks of TXA

Thrombotic complications — Patients who undergo TKA are at high risk of perioperative VTE; patients with low risk of major or significant bleeding typically receive pharmacologic DVT prophylaxis postoperatively (see "Prevention of venous thromboembolism in adults undergoing hip fracture repair or hip or knee replacement"). Several meta-analyses of randomized trials have reported no increase in the risk of deep vein thrombosis or pulmonary embolism after administration of TXA for TKA in healthy patients [94,104,109]. However, the safety of TXA for patients at high baseline risk of thromboembolic events is less certain. Randomized trials have generally excluded patients with a prior history of thromboembolic events, but retrospective studies of adverse events including patients with higher risk (eg, history of venous or arterial thromboembolism) have not reported an increase in thromboembolic events or mortality after receiving TXA [86,95,110-114]. As an example, in a large retrospective case-control study of 38,220 patients who underwent TKA or total hip arthroplasty, of which 8,877 were deemed to be at high risk of thrombotic events, the odds of postoperative thrombotic complications, mortality, and readmission were similar in patients who received TXA and those who did not (OR 1.00, 95% CI 0.85-1.18) [115]. Similarly, in a large United States database study of over 760,000 patients who had TKA or total hip arthroplasty (THA), TXA administration was not associated with increased odds of a composite outcome of venous thromboembolism (VTE), myocardial infarction (MI), seizures, and ischemic stroke/transient ischemic attack (TIA), including in high-risk patients [113]. Three high-risk groups were defined, those with prior VTE, MI, seizures, or ischemic stroke/TIA (28,000 patients), those with renal disease (45,000 patients), and those with atrial fibrillation (46,000 patients). TXA was administered to approximately one-half of the high-risk patients. The composite outcome was similar in high-risk patients who received TXA and those who did not (1.0 versus 1.6 percent). Conclusions from this study are limited by its retrospective nature, and by the relatively low incidence of complications.

Patients with comorbidities may benefit the most from avoiding transfusion, and the available data suggest that patients with comorbidities, defined as ASA status ≥3, are not at increased risk of thromboembolic events with administration of TXA. A 2018 meta-analysis of 78 randomized trials found lack of evidence of harm related to the use of TXA overall in patients who underwent TKA, and moderate quality evidence to support the safety of TXA in patients with ASA score ≥3 [94].

Seizures — Clinicians should recognize that there is a dose-related risk of postoperative seizures after TXA administration, and that doses should be reduced for patients with renal dysfunction, lower body weight, or malnutrition. Seizure related to TXA administration is discussed in detail separately. (See "Blood management and anticoagulation for cardiopulmonary bypass", section on 'Antifibrinolytic administration'.)

Inadvertent intrathecal administration — There have been multiple reported cases of inadvertent intrathecal administration of TXA, due in most cases to the similar appearances of ampoules of TXA and local anesthetic, particularly hyperbaric bupivacaine 0.5% [116-118]. Reported signs and symptoms were back, buttock, and leg pain; myoclonus; seizures; tachycardia and hypertension; and in some cases ventricular fibrillation. Death occurred in approximately one-half of the reported cases. Vigilance is required when administering all drugs during anesthesia, and is particularly important when neuraxial medications are used. (See "Prevention of perioperative medication errors".)

Cementing the prosthesis — Methyl methacrylate cement is most often used to secure the TKA prosthesis to bone. The use of bone cement during arthroplasty can rarely cause bone cement implantation syndrome (BCIS), which is a poorly understood syndrome that includes hypoxia and hypotension, and may ultimately cause cardiac arrest. The primary pathophysiologic mechanism for BCIS is debated, and probably involves pulmonary embolism of cement or fat, with resultant increased pulmonary vascular resistance and right heart compromise or failure [119,120].

Risk factors for BCIS include patient factors (ie, preexisting pulmonary hypertension, significant cardiac disease, osteoporosis, poor ASA functional status), and surgical factors (ie pathologic fracture, intertrochanteric fracture, and long-stem arthroplasty) [121].

Tourniquet — Although controversial, a thigh tourniquet is typically used during TKA to create a bloodless surgical field. Tourniquet times greater than 100 minutes may be associated with postoperative complications, and efforts should be made to minimize tourniquet time [122]. Release of the tourniquet and reperfusion of the ischemic extremity typically cause mild and brief hemodynamic and metabolic derangements, which can be significant for vulnerable patients (eg, patients with cardiac disease).

The literature on the use of a tourniquet during TKA is conflicting. Several trials have suggested that tourniquet use may be associated with increased perioperative pain and/or opioid consumption, compared with tourniquet-less TKA [123-125]. However, one randomized controlled trial reported no difference in early postoperative pain or deleterious effects on function following tourniquet use during TKA [126]. (See "Total knee arthroplasty", section on 'Tourniquet use'.)

Bilateral TKA — Bilateral simultaneous TKA may be performed in selected patients, usually younger and healthier patients, who understand and accept the increased cardiopulmonary risks associated with bilateral procedures. Selection of patients for bilateral TKA is discussed separately. (See "Total knee arthroplasty", section on 'Timing of surgery for bilateral disease'.)

Anesthetic considerations for bilateral TKA include the following:

●Single-injection spinal anesthesia may not last long enough for bilateral TKA; continuous neuraxial anesthesia (combined spinal-epidural [CSE], epidural, or rarely continuous spinal) allows extension of the block as long as necessary. (See 'Neuraxial anesthesia' below.)

●The timing and dose of TXA should be discussed with the surgeon. Some centers administer a higher or extra dose of TXA during bilateral TKA [127].

●As additions to nonopioid and opioid medications for postoperative pain control, options for regional anesthesia techniques include bilateral single injection or continuous peripheral nerve blocks (ie, femoral nerve block, adductor canal block [ACB], interspace between popliteal artery and posterior capsule of the knee [IPACK] block), periarticular infiltration, and neuraxial opioids. The authors usually utilize periarticular injection (PAI) with or without single injection ACBs for bilateral TKA. (See 'Peripheral nerve blocks' above.)

●Importantly, total local anesthetic dose for all injections should be carefully calculated and discussed with the surgeon to avoid administration of potentially toxic doses. Local anesthetic blood levels were measured for 24 hours after TKA in a pharmacokinetic modeling study of patients who underwent TKA (13 unilateral and 15 bilateral) with high volume PAI [128]. Infiltration was performed with ropivacaine 0.2%, 200 mL in each knee, for a total of 400 mg ropivacaine for unilateral and 800 mg for bilateral TKA. Free ropivacaine plasma levels stayed below proposed toxic thresholds. Further study is required before recommending the use of these high doses for PAI. (See "Local anesthetic systemic toxicity", section on 'Local anesthetic dose'.)

●Patients who undergo bilateral TKA may require more intensive thromboprophylaxis than patients who undergo unilateral TKA, due to both the duration of surgery and more difficult postoperative mobilization. (See 'Thromboprophylaxis' above.)

●Perioperative blood loss may be greater after bilateral TKA compared with unilateral TKA. Thus, patients with a low preoperative hemoglobin may not be appropriate candidates for bilateral procedures.

ANESTHETIC MANAGEMENT

Monitoring, venous access, and positioning — Standard American Society of Anesthesiologists (ASA) monitors are applied (electrocardiogram, blood pressure cuff, pulse oximetry). Advanced monitoring is added only as indicated by patient comorbidities. In our practice, we typically place an 18 or 20 gauge intravenous (IV) catheter for a unicompartmental, primary, or simple revision, and a 16 gauge or two smaller bore peripheral IV catheters for a complex revision.

Patients are positioned supine for TKA, which allows for easy transition to a general anesthesia with a laryngeal mask airway or endotracheal tube if the patient is experiencing pain or unable to tolerate lying still.

Choice of anesthetic technique — TKA may be performed with general or neuraxial anesthesia. The choice of anesthetic technique should be based on patient comorbidities and patient choice; for patients who undergo unilateral TKA in whom either general anesthesia or neuraxial anesthesia would be appropriate, we suggest neuraxial anesthesia, and prefer spinal. (See "Overview of neuraxial anesthesia", section on 'General versus neuraxial anesthesia' and "Overview of neuraxial anesthesia", section on 'Preoperative evaluation'.)

General versus regional anesthesia — The literature comparing outcomes after general anesthesia versus neuraxial anesthesia for joint replacement has long been debated. However, the global evidence is becoming more compelling in favor of neuraxial anesthesia. Multiple perioperative factors, some of which are institution-specific, including significant changes in anesthesia and surgical practice over the last decade, may be influencing important perioperative outcomes and confounding data on this issue. As an example, some studies that compared rates of thromboembolism after neuraxial versus general anesthesia were completed prior to the routine use of thromboprophylaxis; those prior studies are no longer relevant. In addition, single institution studies may not be generalizable to other institutions, and large population-based studies may not be generalizable to individual practices. Large multi-institutional randomized trials are required to determine whether choice of spinal anesthesia is directly linked to improved outcomes.

Large population-based observational studies [129-134] and systematic reviews [135-137] of perioperative outcomes after TKA have found no disadvantages to neuraxial anesthesia, and some evidence for improved morbidity and mortality after neuraxial anesthesia.

As part of creation of 2019 consensus recommendations, the multidisciplinary International Consensus on Anaesthesia-Related Outcomes after Surgery (ICAROS) group performed a systematic review and meta-analysis of 27 randomized and observational studies on the choice of anesthetic technique for TKA [135]. Compared with general anesthesia, the use of neuraxial anesthesia was associated with reduced odds of pulmonary and renal failure, infections including pneumonia, thromboembolic events, blood transfusion, and readmission, and reduced length of stay. There were no significant differences in the odds of mortality or stroke between the use of neuraxial and general anesthesia. ICAROS made a weak recommendation based on overall low quality evidence quality to use neuraxial anesthesia for TKA in patients without contraindications. The ICAROS group believed that the desirable effects of neuraxial anesthesia (both morbidity reduction and beneficial resource utilization) outweigh the undesirable effects of general anesthesia, which was not associated with superiority in any outcome.

The relationship between the choice of anesthetic technique for TKA and long-term cognitive dysfunction is an area of active research. Most studies have reported no difference in the incidence of long term cognitive dysfunction in patients who received general versus regional anesthesia for TKA [138-140], though there may be an increase in delirium and cognitive decline in the early postoperative period after general anesthesia [140,141]. The use of opioid-sparing analgesic techniques, depth of periprocedural sedation, and sedative selection (eg, propofol rather than benzodiazepines) may be more important for decreasing early postoperative decline than the choice of primary anesthetic technique for TKA. (See "Anesthesia for the older adult", section on 'Anesthetic techniques' and "Anesthesia for the older adult", section on 'Postoperative pain management'.)

Special consideration should be given to patients with anticipated difficulty with breathing while sedated during regional anesthesia (eg, patients with severe obesity or obstructive sleep apnea). Choice of anesthetic technique and intraoperative management of these patients are discussed separately. (See "Intraoperative management of adults with obstructive sleep apnea" and "Anesthesia for the patient with obesity".)

Tourniquet pain is generally prevented by spinal anesthesia. In contrast, during general anesthesia treatment may be required (eg, deeper anesthesia, opioids) after 60 to 75 minutes of tourniquet time. In one study, tourniquet-related hypertension during general anesthesia for ankle arthroplasty was prevented by injecting local anesthetic around the femoral artery, in an attempt to block the adventitial plexus of nerves that surround the femoral artery [142]. This technique can cause quadriceps weakness due to femoral block if the local anesthetic is not carefully placed, as the femoral nerve lies adjacent to femoral artery deep to the fascia iliaca.

Neuraxial anesthesia — Choice of neuraxial anesthesia (ie, spinal, epidural, or combined spinal-epidural [CSE]) technique varies by institution and clinician. We prefer spinal anesthesia for TKA.

●Spinal anesthesia – Spinal anesthesia provides consistent, dense, bilateral anesthesia. A single injection spinal can reliably provide a duration of block adequate for primary unilateral TKA. The duration of spinal anesthesia is determined by the local anesthetic (LA) used and the dose. Bupivacaine is the most commonly used LA for spinal anesthesia in the United States. A typical dose is 12 to 15 mg of isobaric bupivacaine, with or without opioids or epinephrine added to improve the quality of the block and/or extend the duration (table 3). (See "Spinal anesthesia: Technique", section on 'Choice of spinal drugs'.)

Fast-track protocols in institutions with short surgical times may include the use of isobaric mepivacaine [143,144] or chloroprocaine [145] for spinal anesthesia (table 3). A 2020 study comparing intrathecal mepivacaine versus isobaric bupivacaine or hyperbaric bupivacaine for spinal anesthesia for total hip arthroplasty found earlier time to ambulation and increased likelihood of discharge on the day of surgery in patients who received mepivacaine [146].

The use of isobaric mepivacaine for spinal anesthesia has been associated with transient neurologic symptoms (TNS), though the true incidence of TNS with mepivacaine is unclear [147-150], and in our experience it is very rare. A possible explanation for a low incidence of TNS in some institutions may be the use of antiinflammatory drugs as part of multimodal analgesia. If the authors use mepivacaine for spinal anesthesia, we administer dexamethasone (0.1 mg/kg IV) and postoperative nonsteroidal antiinflammatory drugs (NSAIDs) in an attempt to reduce TNS, though this practice is not evidence-based.

•In a network meta-analysis of 24 trials that compared the incidence of TNS after spinal anesthesia with various local anesthetics, the incidence of TNS with mepivacaine was similar to lidocaine, at approximately 7 percent [151]. The quality of evidence for this comparison was very low.

•In a single institution randomized trial that compared recovery from spinal anesthesia in 154 patients who underwent TKA or THA with either mepivacaine (70 mg, 3.5 mL of 2%) or low dose bupivacaine (10 mg, 2 mL of 0.5%), no patient experienced TNS in either group [144].

Continuous spinal anesthesia is an option that is rarely used and only in carefully selected patients. (See "Spinal anesthesia: Technique", section on 'Continuous spinal'.)

●Epidural anesthesia – Epidural anesthesia may be used for prolonged procedures (eg, bilateral or revision TKA) in which the duration of surgery is expected to outlast spinal anesthesia or when postoperative epidural analgesia is desired. (See 'Epidural analgesia' above and "Epidural and combined spinal-epidural anesthesia: Techniques".)

●Combined spinal-epidural anesthesia (CSE) – CSE anesthesia provides the dense symmetric block from a spinal injection with the option to extend the anesthetic and allow postoperative pain control with the epidural catheter. (See "Epidural and combined spinal-epidural anesthesia: Techniques", section on 'Combined spinal-epidural anesthesia'.)

General anesthesia — General anesthesia is used more commonly than neuraxial anesthesia for TKA in the United States [27]. Induction and maintenance of general anesthesia are discussed separately. (See "Induction of general anesthesia: Overview" and "Maintenance of general anesthesia: Overview".)

We prefer to place a supraglottic airway (SGA) for airway management for TKA, rather than performing endotracheal intubation. Because the patient is positioned supine for TKA, the anesthesia clinician has access to the airway throughout the procedure, and can adjust the airway device as necessary. In addition, muscle relaxation is not required for TKA, so spontaneous or assisted ventilation may be used with the SGA. (See "Airway management for induction of general anesthesia", section on 'Choice of airway device'.)

POST-ANESTHESIA CARE — Most patients are transferred to the post-anesthesia care unit (PACU) for recovery from anesthesia, with monitoring and discharge criteria similar to patients who have other types of surgery. These issues are discussed separately. (See "Overview of post-anesthetic care for adult patients".)

If a peripheral nerve block is placed or replaced in the PACU, patients should be monitored for signs and symptoms of local anesthetic systemic toxicity (LAST). Presentation of LAST may be delayed beyond the period immediately after block placement. (See "Local anesthetic systemic toxicity", section on 'Clinical presentation of toxicity'.)

In our practice, we monitor patients in the PACU based on the dose and type of local anesthetic (LA) used for the block (if performed postoperatively), even if other PACU discharge criteria have been met, as follows:

●For LA bolus of ≤150 mg bupivacaine or ≤450 mg mepivacaine we monitor for 30 minutes

●For LA bolus of >150 mg bupivacaine or >450 mg mepivacaine we monitor for 60 minutes

OUTPATIENT TOTAL KNEE ARTHROPLASTY (SAME DAY DISCHARGE) — TKA is increasingly performed on an outpatient basis. The trend towards reduced length of stay and even same day discharge may be due to improvements in surgical and anesthesia techniques and the desire for cost savings, possibly accelerated by the need to conserve hospital resources during the COVID-19 pandemic [152]. There is not a standard definition of outpatient or ambulatory TKA, though this most commonly refers to hospital or surgery center discharge on the day of surgery.

Many centers that perform outpatient TKA have developed multidisciplinary protocols for such procedures [153]. Most protocols are consistent with fast track (figure 1) or enhanced recovery protocols, with added components for specific patient selection criteria, preoperative patient education, same day physical therapy, and specific functional discharge criteria. An example of a perioperative protocol for ambulatory TKA is shown in a table (table 4).

●Patient selection for outpatient TKA – Outpatient total knee arthroplasty is considered safe in select healthy patients [154,155]. However, optimal patient selection criteria have not been determined and institutional protocols vary. Patients with significant medical comorbidities, opioid dependency, having complex surgery, and/or who cannot have a friend or family member present after discharge are typically excluded from outpatient TKA. The authors use the same patient selection criteria as for a fast-track protocol (figure 1).

Retrospective studies have attempted to identify risk factors for failure of discharge or readmission after outpatient TKA [156]. In a database study of approximately 118,000 TKAs (of whom 3015 had same day discharge), dependent functional status was the strongest patient risk factor associated with readmission within 30 days after same day discharge [157]. Conclusions are limited by the small number of patients with dependent functional status (11), and lack of information on selection criteria for patients who underwent outpatient surgery.

●Anesthetic management for outpatient TKA – Principles for anesthetic management for these patients are similar to those that are used for fast track protocols (figure 1), including the following (table 4) [158]:

•Opioid sparing multimodal analgesia

•Muscle-sparing nerve blocks

•Short/intermediate acting spinal

•Prophylaxis for postoperative nausea and vomiting ([PONV] ie, antiemetics and/or total intravenous anesthesia if general anesthesia [GA] is used),

•Avoiding drugs that potentiate urinary retention (eg, scopolamine, anticholinesterases)

Long acting neuraxial opioids should be avoided, since they require respiratory monitoring for 24 hours after administration. (See 'Neuraxial opioids' above.)

An important goal for anesthetic management is to facilitate same day physical therapy by avoiding persistent numb extremities, symptomatic orthostatic hypotension, nausea/dizziness, grogginess/over-sedation, and limiting pain.

Similar to patients who are to be admitted to the hospital overnight, for patients in whom either general anesthesia or neuraxial anesthesia would be appropriate, we suggest neuraxial anesthesia and prefer spinal (see 'General versus regional anesthesia' above). In a multicenter retrospective study of over 11,500 outpatient knee and hip arthroplasties, use of neuraxial anesthesia was associated with lower postoperative pain scores and opioid consumption, less PONV, and higher successful same day discharge, compared with GA [159]. GA was associated with shorter PACU stay. There were no differences in minor or major adverse events within 30 days of surgery.

●Outcomes after outpatient TKA – The vast majority of patients who are selected for outpatient TKA are successfully discharged on the same day, with low incidences of readmission and/or complications.

•In a 2018 systematic review of the literature including 10 retrospective reports of series of outpatient TKAs and THAs (1000 patients), 94.7 percent of patients were successfully discharged the same day [153]. The most common reasons for failure of same day discharge were pain, hypotension, and nausea. Return to the emergency department occurred in 1.2 percent of patients, and readmission within 90 days occurred in 0.9 percent. There was one major complication and there were no deaths.

•In the database study described above including 3015 TKAs, the incidence of 30 day readmission was 2.6 percent [157]. The most common reason for readmission was a wound complication, followed by thromboembolic complications. The average time of readmission was approximately 11 days after surgery.

SOCIETY GUIDELINE LINKS — Links to society and government-sponsored guidelines from selected countries and regions around the world are provided separately. (See "Society guideline links: Total knee arthroplasty" and "Society guideline links: Enhanced recovery after surgery".)

SUMMARY AND RECOMMENDATIONS

●Preanesthesia evaluation

•Patients who undergo total knee arthroplasty (TKA) often have comorbidities that affect anesthesia care, including osteoarthritis and rheumatoid arthritis, along with comorbidities that accompany advanced age. (See 'Preoperative evaluation' above.)

•The plan for thromboembolism prophylaxis should be discussed with the surgeon before finalizing plans for regional analgesia techniques. (See 'Plan for perioperative multimodal pain control' above and 'Surgical issues that affect anesthesia' above.)

●Plan for postoperative analgesia

•Institutional pathways or protocols for perioperative care typically include pain control, prevention of postoperative nausea and vomiting, and other aspects of perioperative care (figure 1).

•For all TKA patients, we use a multimodal, opioid-sparing strategy for perioperative pain control and postoperative management, which often includes a single injection or continuous peripheral nerve block, periarticular injection (PAI) of local anesthetics (LAs), non-opioid analgesics, prophylaxis for postoperative nausea and vomiting, and early mobilization (table 1). (See 'Plan for perioperative multimodal pain control' above.)

•For an uncomplicated primary TKA, we prefer a single injection or continuous adductor canal block (ACB) combined with a surgeon administered PAI. If PAI is not used, it is reasonable to add interspace between popliteal artery and posterior capsule of the knee (IPACK) block for postoperative pain control (table 1). The ACB is as effective as femoral block and is associated with less motor block. (See 'Regional anesthesia techniques for postoperative pain control' above.)

•Patients who undergo complicated or revision TKA, have multiple comorbidities, known opioid dependence, or pain catastrophizing syndrome may require comprehensive analgesia with femoral and sciatic nerve blocks, lumbar plexus blocks, epidural analgesia, and/or neuraxial opioids. (See 'Regional anesthesia techniques for postoperative pain control' above.)

●Choice of anesthetic technique — Choice of neuraxial versus general anesthesia should be based on patient comorbidities and patient choice. For patients who undergo unilateral TKA in whom either type of anesthesia would be appropriate, we suggest neuraxial anesthesia (Grade 2B), and prefer spinal. (See 'Choice of anesthetic technique' above.)

●Use of tranexamic acid

•For patients without a baseline high risk of thromboembolic events (ie, history of venous or arterial thromboembolism, including myocardial infarction, transient ischemic attack, stroke, or vascular stent) we recommend administration of tranexamic acid (TXA) at the time of TKA to reduce perioperative blood loss and blood transfusion (Grade 1B). For patients with risk factors for thromboembolism, the decision to administer TXA must be individualized, balancing the risk of thromboembolism versus the benefits of reduction in blood loss and transfusion. (See 'Antifibrinolytics' above.)

•The optimal dose, timing, and route of administration of TXA is unclear. We suggest intravenous (IV) administration rather than oral or topical administration (Grade 2C). (See 'Antifibrinolytics' above.)

●Immediate postoperative care — Most patients recover in the post-anesthesia care unit (PACU) uneventfully after TKA. If a peripheral nerve block is placed or replaced in the PACU, the patient should be monitored in the PACU for signs and symptoms of local anesthetic systemic toxicity (LAST) even if other criteria for PACU discharge have been met. (See 'Post-anesthesia care' above.)

●Ambulatory TKA – Principles for perioperative anesthetic management for patients who will be discharged the day of surgery include the use of opioid sparing multimodal analgesia, facilitating early physical therapy and mobilization with use of short acting spinal anesthesia and motor sparing peripheral nerve blocks, prophylaxis for postoperative nausea and vomiting, and avoiding urinary retention (table 4). (See 'Outpatient total knee arthroplasty (same day discharge)' above.)