Pain following total knee arthroplasty (TKA) is associated with delayed recovery, impaired mobility, increased morbidity, longer hospital stay and greater cost. Adductor canal block has recently been shown to improve the pain control of patients following TKA. It is not known whether a single shot technique or a continuous catheter-based infusion technique provides optimal analgesia. The investigators hypothesize that a continuous technique would provide better analgesia and permit patients to achieve objective measures of recovery following TKA than a single shot technique.

Introduction

Total knee Arthroplasty (TKA) is the standard treatment for knee osteoarthritis, which affects millions of people world wide. In 2008, more than 600,000 TKAs were performed in the United States. This number represents more than double the number performed in 1999 and is estimated to increase by more than 500 % (from the 1999 value) by 2020.

Severe postoperative pain occurs in 60% and moderate pain in 30% of patients undergoing TKA. In addition, the intensity of acute postoperative pain correlates with the risk of developing a persistent post surgical pain which can be chronic and debilitating. Thus, the management of pain following TKA is an important determinant of a patient's long term functional outcome. In addition, recovery from TKA requires extensive and effective postoperative physiotherapy (PT), itself a painful event following TKA. Physiotherapy ensures good range of motion and prevents complications such as tissue retraction, adhesion formation and muscle atrophy, all of which may compromise achievement of long term recovery milestones. Thus, the anesthesiologist caring for patients undergoing TKA attempts to provide a postoperative analgesic regimen which ensures that the patient is as pain free as possible while ensuring that the patient has sufficient motor strength to participate actively in PT.

The need to find non-narcotic alternatives for post surgical pain management has become even more urgent recently due to the recent epidemic of narcotic-related deaths. In the United States the number of opioid related deaths since 1999 has reached almost half a million which represents a four fold increase in incidence, and at least half of opioid related deaths involve prescription drugs. Hence effective non-narcotic approaches to pain management hold great importance not only for clinical care but also for public health on a larger scale.

Previously, intravenous, intrathecal or oral opioids were the mainstay of postoperative analgesia following TKA. Although variably effective, the resulting analgesia is associated with well recognized adverse effects including nausea, vomiting, pruritus and sedation. Each adverse effect is unpleasant and likely to impede recovery and participation in PT. Epidural catheters and femoral nerve blocks have both been used for postoperative analgesia in this setting, but each is associated with lower limb motor weakness which hinders early implementation of PT and may increase the risk of falls.

Recently, adductor canal block (ACB) has been reported as an effective means of providing post-operative analgesia for TKA. The ACB is not a new block, having been first described by van der Wal over 20 years ago. The adductor canal block, in the context of analgesia following TKA, is relatively recent, being first described by Tsai et al in 2010. Infiltration of local anesthetic into the adductor canal blocks primarily the saphenous nerve, a sensory nerve, which supplies the medial and anteromedial aspects of the knee via an infra patellar branch.

Postoperative pain following TKA extends well beyond the 8-12 hour duration of a single bolus injection of local anesthetic. Continuous peripheral nerve blockade (PNB) can facilitate early discharge and better rehabilitation in the early postoperative period. In addition, PNB decreases opioid consumption, its associated side effects and is also associated with greater patient satisfaction .

The investigators hypothesized that patients receiving continuous adductor canal block would achieve a greater degree of active knee flexion on POD 2 following TKA than those who receive a single injection ACB.

METHODS Subjects With Institutional ethical approval from the Clinical Research Ethics Committee of the Cork Teaching Hospitals, 40 consecutive patients scheduled for unilateral Total Knee Arthroplasty who met the study criteria and provided written informed consent, were enrolled in a double blind, randomized clinical trial.

Patients were randomly allocated to receive either an ACB followed by catheter placement (n=20) or an adductor canal block followed by sham catheter placement (n=20). The sham catheter was placed on the surface of the leg and covered by an opaque dressing to conceal the insertion site. (A similar dressing was placed on real catheter insertions to conceal group assignment from both patient and assessor.) All patients had the proximal end of their catheter attached to a Pajunk 350 ml FuserPump (FuserPumps are portable elastomeric infusion pumps) concealed in opaque bags to prevent patients and assessors from determining which pumps were actually functioning. These pumps function without a motor and so make no sound. Computer generated randomization was performed for patient group assignment at the time of enrolment in the study. Randomization assignments were done in blocks of 10 in a 1:1 ratio with assignments placed in sealed numbered opaque envelopes. Only the anesthesiologist responsible for performing the block was aware of the group to which a given patient was assigned. The researchers responsible for data collection were unaware of the group to which each patient was assigned .

Preoperative Preparation Potential patients were provided information on the study during the preoperative physiotherapy assessment several weeks before their scheduled surgery. At this time, baseline values for knee flexion and the TUG test were obtained. On the day of surgery, after obtaining written informed consent, patients were assigned to a study group by attaching a sealed randomization envelope to their charts.

After registration on the day of surgery patients were brought to the pre-anesthetic area where IV cannulation was performed and standard anesthetic monitors applied. Patients received intravenous sedation prior to placement of the spinal anesthetic at the discretion of the responsible consultant anesthesiologist. A spinal anesthetic was performed under standard aseptic conditions in the usual fashion. The choice of local anesthetic (either isobaric bupivacaine 0.5% or hyperbaric bupivacaine 0.5%.) used to perform the spinal anesthetic was at the discretion of the responsible consultant anesthesiologist.

Intraoperative Period Patients received intraoperative sedation at the discretion of the consultant anesthesiologist using fentanyl, midazolam and propofol as deemed clinically appropriate. At the conclusion of the procedure, the patients received 30 ml 0.5% bupivacaine diluted with 70 ml 0.9% saline local infiltration analgesia (LIA) injected into the layers of soft tissue around the knee joint by the surgeon. .

Postoperative Period All blocks were performed in the Post Anaesthesia Care Unit immediately after surgery with standard anesthetic monitoring and an IV cannula in place. No subcutaneous local anesthetic or sedation were required as the spinal anesthetic was still effective. All patients received an adductor canal block under ultrasound guidance using a high frequency probe at the mid thigh level. After appropriate asepsis local anesthesia to the skin was placed at the level of the mid femur on the medial aspect of the thigh. Using ultrasound guidance (General Electric Venue 40, 3050 Lake Drive, Citywest Business Campus, Dublin 24, Ireland) with a high frequency L12 linear transducer the femoral artery was visualized in cross section and the ultrasound needle (B Braun Contiplex D ultrasound needle and catheter set, B Braun Melsungen AG, 34209 Melsungen, Germany) placed using in-plane visualization below the Sartorius muscle and just lateral to the femoral artery and medial to the Vastus Medialis. All patients received an injection of bupivacaine 0.5% 10 ml. For patients allocated to receive a continuous infusion postoperatively, a catheter was inserted immediately after injection of the initial bolus of local anesthetic. Patients allocated to the single injection group received the injection of local anesthetic followed by placement of a sham catheter on the skin, covered with a large opaque dressing. (Figure 3) In both groups, patients' catheters were attached to the concealed in opaque bags. The sham pumps were filled with normal saline. The pumps containing local anesthetic contained 0.15625% bupivacaine 350 ml were set to run at 8 ml/ hour (the equivalent of 0.125 % bupivacaine running at 10 ml/hour) All blocks were placed by Consultant anesthesiologists with fellowship training in Regional Anaesthesia.

All patients received multimodal oral analgesia which included Paracetamol 1g po every 6 hours, oxycodone (oxycontin) po twice a day (10 mg/dose for patients >70 years and 20 mg/dose for those <70 years) , pregabalin 50 mg po 3 times a day, and oxycodone (oxynorm) 5-10mg 4 hourly po as required for breakthrough pain (Pain score > 4/10)

Recordings and Measurements Research personnel unaware of patient randomization performed all clinical assessments. General Demographic data collected included: age, gender, height, weight, American Society of Anesthesiologists (ASA) Status and BMI.

Assessment of block function was made by checking for sensory changes to pain and temperature (using a blunt needle and ethyl chloride spray) in the distribution of the saphenous nerve at 20:00 on POD 0. A functioning adductor canal block with bupivacaine will last at least 8-12 hours. Thus all blocks,( i.e. both with and without a continuous infusion) were anticipated to still be effective at 20:00H on the day of surgery, as no blocks were performed before 10 AM. Decreased sensation in the extremity receiving the block compared with the unblocked one at the level of the medial malleolus was accepted as evidence of a functioning block.

Adverse Events Each day the assessor inquired about complications that could be attributable to the study, such as patient falls or symptoms or signs of local anesthetic toxicity. Any such events were documented. The common adverse effects of opioids (i.e. nausea, vomiting) were not recorded. However, there were no episodes of respiratory depression that required resuscitation.

Statistical Methodology Sample size was determined based on an unpublished retrospective review of knee flexion in patients who had undergone TKA at the investigators institution having received a single injection ACB. From this review, a mean maximum knee flexion of 82 (SD = 2) degrees on the second day postoperatively was assumed. A 20% difference in knee flexion was considered to be clinically significant. To achieve an 80% power and to detect a 20% difference in knee flexion with a 2-sided Type I error rate of 0.05, the calculated minimum sample was 9 patients per group. To prevent loss of power due to early withdrawal of patients, incomplete data, non-functional adductor canal blocks and absence of statistical information on a comparable continuous infusion group, 20 patients in each were included group. General demographic data in both groups were compared using unpaired, two tailed t-tests for independent samples for continuous variables, and the Chi squared test (or Fisher's Exact test as appropriate) for categorical variables. Basic statistical tools were employed to describe the data: mean, median and standard deviation. Cumulative oxycodone consumption was compared between groups using analysis of covariance (ANCOVA) adjusted for age, gender and ASA status.

Inclusion Criteria:

• Scheduled unilateral total knee arthroplasty,
• ASA physical status I to III,
• age > 18 years.

Exclusion Criteria:

• Body Mass Index >35,
• pre-existing lower extremity neurologic abnormalities,
• infection, history of chronic opioid use,
• pregnancy, contraindications to peripheral nerve block or central neuraxial blockade,
• allergy or contraindications to local anesthetics or drugs that would be used for multimodal analgesia,
• inadequate command of English
• refusal of spinal anaesthetic