Total knee replacement (TKR) and total hip replacement (THR) surgery offer a multitude of challenges for the anesthesiologist and peri-operative physician. The choice of anesthetic has important implications, not only for the intra-operative course, but also for the post-operative outcome. Intra-operatively, optimal surgical conditions must be balanced with maintaining a stable physiologic state. The choice of anesthetic and analgesic techniques will affect not only the short term success of post-operative analgesia, but also the ability to achieve certain physical therapy milestones. This, in turn, may influence the overall functional outcome.
At Hospital for Special Surgery (HSS), the anesthetic of choice for both TKR and THR is a neuraxial regional technique utilizing an epidural catheter. This type of anesthesia is either an epidural anesthetic alone, or a combined spinal epidural (CSE) technique. Neuraxial regional techniques offer several benefits compared with traditional general anesthesia. Expertly conducted, this technique avoids the need for airway manipulation, eliminates complications of general anesthesia, and allows for continuous communication with your patient throughout the procedure if desired. This latter benefit is important to patients, many of whom place great significance on their ability to choose their level of sedation during the procedure.
Many reports over the past two decades have attempted to compare risks and benefits of regional and general anesthesia in different patient populations. Rodgers et al. (1) performed a meta-analysis on studies comparing neuraxial versus general anesthesia with regard to post-operative mortality and morbidity. These authors concluded that neuraxial blockade reduces major post-operative complications in a wide variety of surgical patients, with the greatest reductions seen in the orthopedic population. These complications included deep venous thrombosis, pulmonary embolism, blood transfusion requirements, pneumonia, and respiratory depression.
In a retrospective review of in-hospital morbidity and mortality at HSS, Sharrock et al. (2) reported a decrease in mortality rate from 0.36% (13 of 3622 patients) during the period of 1981 to 1985, to 0.01% (6 of 5869 patients) during the period of 1987 to 1991. This reduction in mortality coincided with several significant variations in clinical practice. The technique of hypotensive epidural anesthesia (HEA) for THR was initiated to replace general anesthesia. Invasive hemodynamic monitoring was routinely employed for high-risk medical patients. Furthermore, these high-risk patients were monitored and medically managed in the post-anesthesia care unit (PACU) for extended periods following surgery. Finally, the use of epidural analgesia to manage post-operative pain became routine. The individual effect of each of these variables is impossible to determine; however, it is clear that there was a definitive improvement in patient outcomes using these techniques.
The majority of anesthesiologists at HSS utilize the technique of CSE anesthesia for TKR surgery. CSE is performed in the lateral decubitus or sitting position. The patient is sedated to a state of consciousness required to facilitate comfortable placement of the block. The skin is prepared with 10% povidone-iodine or chlorhexidene solution and the L 3/4 or L 4/5 interspace is anesthetized with a subcutaneous infiltration of local anesthetic. A 17g Tuohy needle is advanced via a midline or paramedian approach through the ligamentum flavum until a loss of resistance to air is noted within the epidural space. A 27g pencil-point needle is then placed through the Tuohy needle, and a characteristic “pop” is felt as the spinal needle punctures the dura. After visualization of CSF through the spinal needle, an isobaric local anesthetic solution (bupivacaine 10-15 mg or mepivacaine 45-60 mg) is injected. Following removal of the spinal needle, a 20g multi-orifice catheter is inserted 3-5cm into the epidural space. The Tuohy needle is removed and the catheter is secured to the patient's back with tape. Complete surgical anesthesia will ensue within 5 to 10 minutes.
The hemodynamic effects of spinal or epidural anesthesia may be significant. CSE often results in a fall in systemic blood pressure. The mechanism responsible this effect is a diffuse sympathetic blockade causing a decrease in stroke volume, cardiac output and systemic vascular resistance (3,4). Judicious use of intravenous fluids and vasopressors is essential to maintain optimal cardiac performance (5). Careful monitoring of heart rate and blood pressure by invasive or non-invasive means is essential during induction and maintenance of CSE anesthesia
The CSE technique offers advantages over other neuraxial anesthetics such as epidural or spinal anesthesia. A small dose of local anesthetic is required during induction of anesthesia with a CSE compared to the larger doses and volumes required with epidural anesthesia. Therefore, the potential for local anesthetic toxicity can be avoided when compared with the traditional epidural approach. The smaller dose of local anesthetic also allows the anesthesiologist the option of performing a pre-operative femoral nerve block for a TKR without exceeding limits of local anesthetic dosing guidelines.
Placement of an indwelling epidural catheter offers several advantages over other neuraxial techniques such as a “single shot” spinal anesthetic. Presence of the catheter allows for extension of anesthesia should the surgery last longer than anticipated. In addition, epidural catheters used in the post-operative period have been shown to provide many benefits. Studies have shown a decreased rate of deep vein thrombosis after TKR and THR in patients using continuous epidural infusions for post-operative pain management (6,7). Pain scores are improved when epidural local anesthetic and narcotic combinations are used to treat post-operative pain when compared to intravenous patient controlled analgesia (PCA) with narcotics alone (8,9). Better pain control improves rehabilitation after TKR by controlling post-operative pain during continuous passive motion. In a prospective randomized study, patients using continuous epidurals infusions were found to have lower pain scores, improved range of motion and shorter rehabilitation times when compared to patients using intravenous patient-controlled morphine (9).
A potentially volatile hemodynamic period during TKR occurs during tourniquet release. Blood pressure usually falls significantly at this time and the anesthesiologists must be prepared to intervene immediately. Kahn et al. (10) noted an approximate 20% reduction in mean arterial pressure within one minute of tourniquet release. The heart rate will often rise at this time due to physiologic compensatory mechanisms, but this response can be variable. Kahn et al. also noted a heart rate decline of greater than 10 bpm in 4% of patients following tourniquet deflation. There were no predictive patterns of heart rate responses in this setting. Careful fluid management during the course of the anesthetic and rapid treatment of hypotension and/or bradycardia can limit the volatile hemodynamic alterations sometimes observed in this population.
Several studies have evaluated the effects of neuraxial anesthesia on tourniquet pain. Compared with general inhalational anesthetics, spinal or epidural anesthesia inhibits blood pressure increases seen during tourniquet inflation (11). This reduction in tourniquet pain may lead to a decrease in hypercoagulability following TKR (12). In a large retrospective review performed at HSS, epidural anesthesia was found to significantly reduce the incidence of proximal thrombosis following TKR surgery using a tourniquet (7).
Not all studies have shown clear benefits of neuraxial anesthesia for TKR. A large randomized study performed at HSS evaluated cognitive function and cardiovascular outcomes following TKR under either general or epidural anesthesia (13) . Evaluating 262 elderly patients (median age = 69), no differences were noted in long-term cognitive function or cardiovascular complications between anesthetic groups. However, post-operative analgesic techniques were not controlled in this protocol, and may have played a role in the outcome.
Considering the evidenced-based risks and benefits of regional anesthesia for TKR, as well as our extensive clinical experience, anesthesiologists at HSS firmly believe that the neuraxial anesthetics are superior techniques.
Anesthesiologists at HSS utilize the technique of hypotensive epidural anesthesia as developed and described by Sharrock (14). One benefits of HEA is reduced blood loss and reduced need for intra-operative and postoperative blood transfusions (14). Recent studies report intra-operative blood loss during THR surgery using normotensive anesthesia ranges from 500-1800mL (15-17). This is significantly greater than the 100-300 mL of blood loss seen intra-operatively at HSS using the HEA technique (14). Approximately 80% of the patients at HSS undergoing THR predonate autologous blood. Within this group, less then 5% require homologous blood transfusions. This fact combined with minimal blood loss seen with HEA has led to a significant decrease in homologous blood transfusions and their associated risks (14).
A reduced amount of blood on the operative field provides optimal conditions that may lead to significantly shorter operative times. The duration of surgery has been shown to affect the incidence of DVT formation (18). Operations less than 70 minutes were found to have a DVT rate of 9%, while operations lasting longer than 70 minutes had DVT rates of 20%. This may be explained by an anatomical event that occurs during THR procedures. Dislocation of the hip joint results in flexion and external rotation of the lower extremity. This in turn causes occlusion and “kinking” of the femoral vein and artery resulting in stasis of blood distal to the “kink” (19). Decreasing the length of time that the vessels are occluded may reduce DVT formation (18).
The patient is initially placed supine, given supplemental oxygen via nasal cannula, and sedated to an appropriate level to facilitate placement of invasive monitors and the neuraxial anesthetic. A radial arterial line and if indicated, a central venous line are inserted. It must be stressed that invasive blood pressure monitoring is needed prior to placement of the epidural or the CSE anesthetic due to the rapid and often profound sympathectomy that occurs with this technique, resulting in rapid hypotension and decreased venous return. Close monitoring of the blood pressure is required to help guide fluid and vasopressor therapy (14).
Patients are positioned in the sitting or lateral decubitus position to facilitate epidural or CSE placement. The skin is prepared with 10% povidone-iodine or chlorhexidene and local anesthetic is infiltrated subcutaneously at the appropriate interspace. A 17g Tuohy needle is advanced using the paramedian or midline approach until a loss of resistance to air is encountered in the epidural space. If an epidural anesthetic is planned, a total of 15-25 mL of isobaric local anesthetic (0.75% bupivacaine or 2% lidocaine) is injected. An epidural catheter is placed approximately 3-5cm into the epidural space and can be used to augment the anesthesia intra-operatively as well as provide analgesia in the post-operative period.
Another technique commonly used to achieve deliberate hypotension for THR at HSS employs the CSE in place of the epidural alone. The CSE is performed using the technique previously described for TKR. If the MAP does not fall in a timely fashion following the initial spinal dose, additional local anesthetic can be given through the epidural catheter to increase the level of sympathetic blockade. After negative aspiration for blood or CSF, local anesthetic (2% lidocaine or 0.5% bupivacaine) is given through the epidural catheter in 5mL increments.
The goal of either technique is to achieve a sensory and sympathetic block to the T1-T4 range. The extensive sympathectomy will cause a rapid decrease in mean arterial blood pressure (MAP). It is important that an infusion of epinephrine is prepared and connected to the patient either via central or peripheral intravenous catheter prior to, or immediately after induction of epidural or CSE anesthesia. An epinephrine infusion rate of 2 to 5mcg/minute is often needed to maintain blood pressure and cardiac output.
An air-filled roll is placed under the axilla to decrease the pressure on the dependant shoulder. The arms are placed in a neutral position and padded to minimize the risk of brachial plexus injury. The patient is sedated to allow tolerance of the lateral decubitus position for the length of surgery. It is important that the level of sedation is titrated so that the patient is easily arousable in order to assess cognitive function periodically during the procedure. In a study of 235 older patients, cognitive testing was performed before and after HEA for primary THR (13). Inclusion criteria included an age greater then 70, or younger patients with at least one of the following co-morbidities: hypertension, cardiac disease, or diabetes mellitus. No change in cognitive function was noted.
The hemodynamic goal of this technique is to maintain a MAP of 50 to 60mmHg for the duration of the procedure. If the MAP remains high, the dose of epinephrine infusion can be reduced or small volumes (no more than 5ml increments) of local anesthetic can be carefully titrated via the epidural catheter. If the MAP falls too rapidly, the epinephrine infusion can be increased and a bolus of 300-500mL of intravenous fluid is given. Strict fluid management should be guided by heart rate (greater then 80 beats per minute using this technique may reflected volume depletion) or, when available, central venous pressures. Using HEA technique, the total blood loss for a primary unilateral hip is typically 100-300 mL and most patients only require 1000 to 1500 mL of intravenous fluid (14).
The use of low-dose epinephrine infusion is a vital part of this technique. The extensive sympathetic blockade will decrease systemic vascular resistance (SVR) as well as provide blockade of the cardiac accelerator fibers. If untreated, this will lead to severe hypotension, bradycardia and a significant decrease in cardiac output (4,20). A MAP of 50-60mmHg can be achieved while using vasopressor infusions such as norepinephrine or phenylephrine. However, heart rate, central venous pressure, and cardiac output will also decrease (21-23). Beta agonists, such as isoproterenol or dobutamine, will lead to a decrease in central venous pressure and an increase in heart rate (24,25). In many patients this will lead to a decrease in cardiac output and a MAP that is difficult to stabilize. However, Sharrock et al. (20) showed an increase in stroke volume (SV) and cardiac index (CI), with little change in heart rate, pulmonary artery diastolic pressure and central venous pressure (CVP) when a low-dose epinephrine infusion (approximately 2-4 mcg/minute) was used. The result is hypotension to a MAP of 50-60mmHg with preservation of HR, filling pressures (CVP) and cardiac output (21-23). These favorable hemodynamic conditions may be the reason why the technique using low-dose epinephrine infusions has been proven safe in the majority of patients, including the elderly (greater than 70 years old), patients with controlled hypertension, and patients with a low cardiac index (20,26).
The technique of HEA at HSS also includes a single intra-operative dose of heparin given prior to implantation of the femoral components. Studies measuring the markers of thrombin generation and fibrin formation (fibrinopeptide A, thrombin-antithrombin III complexes, and prothrombin fragments F1+2) during THR surgery show that thrombogenesis occurs during preparation of the femur and implantation of the femoral components, especially when cement is used (2). Low-dose heparin given after implantation on the acetabular cup was found to suppress fibrin formation that occurs during insertion of the femoral component. In a series of 212 patients undergoing THR using the technique of HEA as described, 15 units of heparin per kilogram of body weight were given to the patients prior to femoral preparation. The rate of DVT as determined by ultrasound was found to be only 6% (27). No evidence of excessive bleeding or epidural hematoma was noted. Furthermore, no pulmonary emboli were diagnosed.
In a retrospective case control study assessing radiographic appearances of cemented acetabular components, improved fixation was noted when HEA was used (28). This suggests that HEA provides optimal cement penetration into bone, which improves the fixation of cemented components. The improved fixation may lead to a higher degree of functionality and longer lasting prosthesis.
The most common complication of neuraxial anesthesia is benign, short-term backache. A second complication, post dural puncture headache (PDPH), is nearly non-existent following performance of a CSE using a 27g pencil-point spinal needle in the TKR or THR population of patients. In this population, the incidence of PDPH remains low, even in the setting of dural puncture with the 17g Tuohy needle.
Nerve injury from needle trauma is very rare as well. Since persistent neurologic deficits have been reported following paresthesias during needle placement (29), it is important to limit the level of pre-block sedation. Maintaining some degree of patient contact will allow reporting of paresthesias or pain on injection of local anesthetic. It is also extremely important to carefully and accurately assess the spinal level prior to CSE placement. Visual inspection of Tuffier's line does not provide an accurate assessment of the actual spinal level in several studies (30,31). Damage to the spinal cord or conus medularis by needle trauma is a rare, but devastating complication of CSE performed above the L 2/3 interspace.
Serious complications following neuraxial anesthesia are rare; however, as with any anesthetic, significant complications may occur. The incidence of serious or permanent complications (such as cardiac arrest, neurological injury, or cauda equina syndrome) following neuraxial anesthesia is reported to be less than 0.1% (29). Other complications, such as infections and anterior spinal artery syndrome, are exceedingly rare if proper technique and management of the anesthetic are observed.
Patients presenting for TKR and THR are often chronically taking medications that affect the coagulation system. Furthermore, post–operative anti-thrombotic therapy may affect the choice of anesthesia and analgesia following TKR and THR. Medications of concern include oral anticoagulants (warfarin), standard and low molecular weight heparin, and anti-platelet medications (platelet aggregation inhibitors, and glycoprotein IIb/IIIa receptor antagonists). At Hospital for Special Surgery, we follow the American Society of Regional Anesthesia and Pain Medicine consensus guidelines on neuraxial anesthesia and anticoagulation using medications as noted above (http://www.asra.com/).
The widespread use of non-steroidal anti-inflammatory drugs and herbal therapies (garlic, ginkgo, and ginseng) poses some concern for anesthesiologists performing neuraxial anesthetics for TKR and THR. At HSS, we believe that of themselves, these drugs add no significant risk of neuraxial bleeding; however, when used in combination with other medications that affect coagulation, the risks of some regional techniques may be increased.
Surgical pain following TKR or THR can be extremely intense during the immediate post -operative period. Aggressive analgesic therapy is essential to maximize patient satisfaction and potentially minimize complications. Adequate analgesia also may affect the acute and long-term rehabilitation of the patient. The goals of post-operative analgesia include maximizing pain relief, minimizing side effects, and maintaining aggressive rehabilitation regimens. It is exceedingly important to maintain active dialogue between surgeons, internists, and the acute pain team regarding issues of post-operative anticoagulation. Many of these analgesic techniques must be modified depending on the anticoagulation regimen employed.
The analgesic regimen of choice at Hospital for Special Surgery is the use of patient controlled epidural anesthesia (PCEA). Patients are given an epidural infusion of bupivacaine (0.06%) and hydromorphone (10 mcg/ml) at an infusion rate of 4-8 ml/hour. This can be supplemented with patient controlled boluses of 4-6 ml every 10 to 20 minutes. Based on the patient's response and pain tolerance, this infusion is tapered and transitioned to oral pain medication by the first or second post-operative day.
Complications from acute pain management are rare. Based on HSS internal data on over 7400 patients admitted to the acute pain service in 2003, the major complication rate was 0.1%. Major complications were defined as respiratory depression, epidural hematoma, infection, medication error, neurological injury, or death. Common side effects of PCEA using this regimen include pruritis, nausea, and vomiting. Again, based on internal data at HSS, 24% of patients were treated with an anti-emetic for nausea and vomiting while on the acute pain service. Eighteen percent of patients were treated for pruritis. Management of these side effects, including reducing or discontinuing the PCEA, must be balanced with the ability to provide effective analgesia.
Over the years, we have continually modified our analgesic techniques to meet the changing needs of our orthopedic surgeons and rehabilitation specialists. In order to maintain the ability to aggressively rehabilitate patients on post-operative day one, we have decreased the bupivacaine concentration from 0.25% to 0.06% over the past decade. While this does allow us to achieve rehabilitation goals, analgesia has suffered. We, therefore, have supplemented our PCEA regimen with single shot peripheral nerve or plexus blocks in many cases.
For TKR, femoral nerve blocks provide excellent supplemental analgesia. These blocks are usually performed in the operating room prior to placement of the CSE. Following a sterile 10% povidone-iodine or chlorhexidine preparation, the femoral pulse is palpated in the inguinal crease. A 22g, 50 mm insulated needle (Braun, Stimuplex) is inserted 1 cm lateral to the femoral pulse. When a quadriceps twitch is elicited, the current is reduced to less than 0.5 mA while maintaining the observed motor response. Twenty to 30 ml of bupivacaine 0.25 to 0.5% or ropivacaine 0.3 to 0.7% (both with 1:200,000 epinephrine) is injected. It is important to elicit the quadriceps twitch (i.e. “patella jerk”) and not the sartorius muscle (medial aspect of the thigh) to ensure a high success rate for the block. The analgesia provided by this block will last from 16 to 24 hours and will markedly reduce the consumption of bupivacaine and hydromorphone via PCEA (32). This, in turn, can minimize side effects and improve patient satisfaction. The addition of a sciatic nerve block to a femoral block for post-operative analgesia following TKR has been shown not to provide significant additional analgesic benefits (33).
Aggressive post-operative analgesia following TKR may also improve achievement of rehabilitation milestones. Pati et al. (34) noted expedited joint mobility in patients treated with epidural anesthesia compared with conventional intravenous therapy. Furthermore, Yadeau, et al. (32) found the combination of femoral nerve block and epidural PCEA significantly improved analgesia, and did not delay physical therapy milestones.
Pain after THR is significantly less intense than that experienced following a TKR. The PCEA, as described previously, offers adequate pain control for the majority of patients following THR surgery. However, several anesthesiologists at HSS perform psoas (lumbar plexus) blocks on patients undergoing THR surgery to facilitate earlier rehabilitation and mobility. The psoas block can be placed in the operating room with the patient in the lateral decubitus position either before or after the operative procedure.
The technique for performing a psoas block begins by drawing a line connecting the iliac crests to help locate the L3 spinous process. A mark is made 5 cm lateral to and 4 cm caudal to the L3 spinous process. This represents the location of the L4 transverse process. The skin is prepared with 10% povidone-iodine or chlorhexidene and infiltrated with local anesthetic using a 25g needle. A nerve stimulator set at 1-2 mA is attached to the patient and a 21g 100 mm insulated needle (Braun, Stimuplex) is inserted perpendicular to the skin. The needle is angled slightly medially until the transverse process of L4 is contacted and the depth is noted. If the bone is not contacted within 5-6 cm, the needle insertion point should be reassessed. Once the transverse process has been identified the needle is pulled back to just under the skin and redirected 15 degrees caudally. The needle is advanced until a quadriceps twitch is elicited. The current is reduced until the quadriceps twitch is seen at a current less than 0.5 mA. Aspiration of the needle should be negative for blood or CSF prior to injecting the local anesthetic. The same local anesthetic used for a femoral nerve block is also employed for the psoas block. The duration or analgesia is 16-24 hours.
In addition to the complications seen with all nerve blocks (intravascular injections, intraneural injections, local anesthetic toxicity, etc.), the complications of a psoas block include epidural or subarachnoid injections, as well as, peritoneal, kidney, and ureter puncture. Because of these increased risks and lower pain scores noted following THR, routine use of psoas blocks is not recommended. In experienced hands, this block can be done with relative ease in the recovery room in those patients with unusually elevated levels of pain.
In conclusion, the anesthetic and analgesic management of patients during and following TKR and THR may have a significant impact on medical and surgical outcome, as well as patient satisfaction. The anesthetic experience at HSS over the past two decades has included thousands of TKR and THR procedures. It is important to continually adapt and improve anesthetic and analgesic techniques to ensure the highest quality in medical care, minimize side effects and complications, and maintain a high degree of efficiency and patient satisfaction. In considering all of the evidence, anesthesiologists at HSS overwhelmingly agree that neuraxial anesthesia supplemented by analgesic nerve and plexus blocks have both real and potential advantages over general anesthesia and intravenous narcotics for TKR and THR.