| JOURNAL HOME | CME HOME | THIS MONTH | PAST ISSUES | ETOC | COLLECTIONS |
| AUTHORS | REVIEWERS | EDITORIAL BOARD | FEEDBACK | RSS | HELP |
| ||||||||||||||
| ||||||||||||||
|
|
|
|
||||||||||||
|
||||||||||||||
From the Departments of *Anaesthesia, Physiotherapy, and Orthopaedic Surgery, St. Vincent's Hospital, Melbourne, Australia.
Address correspondence and reprint requests to Michael J. Barrington, FANZCA, Department of Anaesthesia, St. Vincent's Hospital, PO Box 2900 Fitzroy, Victoria 3065, Melbourne, Australia. Address e-mail to Michael.Barrington{at}svhm.org.au.
Abstract
BACKGROUND: Continuous femoral nerve blockade (CFNB) is often used for postoperative analgesia after total knee arthroplasty (TKA). CFNB can be instituted using a variety of techniques. Stimulating catheters (SC) have the advantage of confirming placement of the catheter close to the nerve during advancement.
METHODS: In this randomized, controlled, double-blind trial, we compared a SC with a nonstimulating catheter (NSC) technique for institution of CFNB and its effects on quality of analgesia after TKA performed under general anesthesia in 82 patients. Patients were randomized to have CFNB instituted using either a NSC or a SC technique. Sensory blockade was assessed 10 and 20 min after injection of lidocaine via femoral catheter and on postoperative days 1 (POD 1) and 2 (POD 2). A standardized multimodal analgesic technique, including a single injection sciatic block (preoperative), IV morphine (patient-controlled analgesia), celecoxib, and paracetamol, was administered to all patients. Outcome variables included morphine requirements, pain scores, and markers of early recovery.
RESULTS: The proportion of patients with sensory blockade in the femoral nerve distribution was between 90% and 95% at all measurement times with no difference between groups. In the first 24 h, the NSC group required 19.5 (1–67) [median (10th–90th centiles)] mg morphine compared with the SC Group 18 (2–51) mg (P = 0.69). At 24 h, the 95% confidence interval for difference in morphine consumption between groups was –8 to 5 mg. There was no difference between groups in visual analog scale scores at rest on POD 1 and POD 2, during active and passive physiotherapy; and in markers of early recovery after surgery.
CONCLUSIONS: In this study, blind catheter advancement was as reliable as a SC technique for establishing and maintaining CFNB for postoperative analgesia as a part of multimodal analgesia technique after TKA.
Postoperative pain after total knee arthroplasty (TKA) can be severe. Continuous femoral nerve blockade (CFNB) has the benefits of extended postoperative analgesia compared with single injection femoral nerve blockade1 and has fewer side effects and improved functional recovery compared with either IV morphine or epidural analgesia after TKA.2,3 CFNB is commonly instituted with a peripheral nerve stimulator connected to a stimulating needle to localize the femoral nerve, followed by insertion of the catheter through the needle. Studies using blind advancement of femoral catheters indicate that catheter position in relation to the nerve is unpredictable.4,5 Therefore, even if the initial injection of local anesthetic through the needle produces adequate intraoperative anesthesia/analgesia, subsequent infusion through the catheter may not provide adequate postoperative analgesia. This is termed "secondary analgesic block failure" and has been documented in 10% of patients with continuous peripheral nerve blockade,5,6 but may occur in up to 40%.7 An alternative to injecting a large initial dose of local anesthetic through the needle is to inject only local anesthetic through the catheter, thereby obtaining feedback regarding the location of the catheter in relation to the nerve based on degree of subsequent motor/sensory block.
A further option is to obtain active feedback regarding the proximity of catheter to nerve during catheter advancement. Stimulating catheters provide this feedback during both advancement and after final placement of the catheter. The ability to evoke a motor response with a stimulating catheter (SC) indicates that the catheter remains close to the target nerve.7–9 The primary objective of this study was to compare a SC with nonstimulating catheter (NSC) technique for institution of CFNB and its effects on the quality of analgesia after TKA.
METHODS
The St. Vincent's Hospital (Melbourne) Human Research and Ethics Committee approved this study. Written informed consent was obtained from all patients. Patients scheduled for unilateral, primary TKA were included in this prospective, randomized, controlled, double-blind trial comparing two techniques of inserting catheters for CFNB. Exclusion criteria were: an inability to comprehend instructions on using an IV patient-controlled analgesia device (IV-PCA) or a visual analog scale (VAS) for pain, contraindications to femoral nerve blockade (e.g., infection overlying the injection site) or to the use of drugs used in the study, chronic opioid use, a chronic pain syndrome not related to the knee pain, and severe flexion or valgus deformities of the knee. Patients were approached regarding participation at a preadmission clinic at least 1 wk before scheduled surgery. At that time, patients were educated about the use of a VAS and IV-PCA, and this education was repeated on the day before or day of surgery.
Patients were randomized to have CFNB inserted by one of the two techniques according to a computer-generated randomization sequence sealed in opaque envelopes. Group NSC had catheters inserted using a NSC technique with a Contiplex® Tuohy Continuous Nerve Block Set (B. Braun, Melsungen, Germany). Group SC had catheters inserted using a SC technique using StimuCathTM Continuous Nerve Block Set (Arrow International, Reading, PA). Both kits use Tuohy needles. Premedication consisted of controlled-release oxycodone (OxyContin, Mundipharma, Sydney, Australia) 10 mg, temazepam 20 mg, and celecoxib 400 mg orally 1 h before block insertion. In the anesthetic room immediately before block insertion, the sealed envelope was opened revealing group assignment. In both groups, CFNB was then established using a strict aseptic technique with the skin entry point 1–2 cm lateral to the femoral artery, caudad to the inguinal ligament, and cephalad to the groin skin crease. Patients were blinded to their study group allocation. A peripheral nerve stimulator (Stimuplex® HNS11; B Braun) with initial settings of current output 1.0 mA, pulse width 0.3 ms, and 2 Hz was used to locate the femoral nerve with an optimal motor response being quadriceps femoris contraction indicated by cephalad patellar movement. The lowest current output required to elicit a motor response (electrical threshold) via the needle was recorded in both groups and via the catheter in group SC at 0.1 and 0.3 ms. In group NSC after an acceptable motor response, the catheter was advanced 5 cm beyond tip of the needle. In group SC, after an acceptable motor response via the needle, the nerve stimulator was disconnected from the needle and connected to the SC. The catheter was advanced with an initial current output of 0.5 mA and pulse width 0.3 ms. The catheter was advanced beyond the needle tip while the quadriceps contractions (patellar movement) were unchanged or increasing with the goal of securing the catheter 4–6 cm beyond the needle tip. If the quadriceps motor response from catheter stimulation significantly reduced or disappeared before 4 cm, then the catheter was withdrawn into the needle and sequential rotation (clockwise or anticlockwise), angulation (steeper or flatter) and advancement or withdrawal of needle was made in attempt to change direction of catheter advancement and improve motor response via the catheter. The number of catheter passes required (first, two to four, more than four) and the type of motor response were recorded. The process was repeated with a different needle entry site (after reassessment of the landmarks), if the above maneuvers were unsuccessful and the initial motor response via the needle was suboptimal. In both groups, the time taken to perform blocks was recorded and defined as the time from initial insertion to removal of the Tuohy needle. Twenty milliliters of lidocaine 1.5% with epinephrine 5 µg/mL was injected through the catheter. No local anesthetic was injected through the needle. Patients in both groups also received a sciatic nerve block using the subgluteal approach with ropivacaine 150 mg with epinephrine 100 µg in 25 mL.10
A research assistant absent during the procedure, blinded to group allocation and trained in how to perform measurements assessed the success and quality of femoral nerve blockade. Success was defined as clear loss of temperature discrimination to ice over femoral nerve distribution. Patients’ tolerance of transcutaneous electrical stimulation (TES) at 50 Hz in the distribution of the femoral nerve (anterior mid to distal thigh) was measured as previously described.8 This artificial painful stimulus defined the quality of femoral nerve blockade (increasing TES [mA] from baseline indicating denser blockade). Operative and nonoperative limb measurements were made preblock, 10 and 20 min after injection of lidocaine, and on postoperative days 1 (POD 1) and 2 (POD 2), with the day of surgery being considered day 0 (POD 0).
Intraoperative endotracheal anesthesia consisted of nondepolarizing muscle relaxant and general anesthesia, alfentanil 1 mg, propofol 1–2 mg/kg, and desflurane. An infusion of ropivacaine 0.2% at 12 mL/h was commenced within 10 min of induction of general anesthesia (this was preceded by an initial dose of 20 mL of ropivacaine 0.375%). Operative and tourniquet times were recorded.
Postoperative analgesia consisted of a ropivacaine 0.2% infusion 12–14 mL/h, (10 mL bolus, 60 min lockout, maximum of 2 boluses/4 h period), the sciatic nerve block (inserted preoperatively), IV-PCA morphine (bolus 1 mg, lockout 5 min), paracetamol 1 g 4 times daily, and celecoxib 200 mg twice daily. Local anesthetic infusions and IV-PCA continued until the morning of POD 2. OxyContin 10 mg orally twice daily and oxycodone (Endone, Sigma Pharmaceuticals, Victoria, Australia) 5–10 mg orally 4 times daily prn were added on POD 2 one hour before cessation of the ropivacaine infusion.
Postoperative measurements included morphine dosage at 12 and 24 h postoperatively and at 8:00 am on POD 2. VAS pain scores (range, 0–100 mm) were measured at rest on POD 1 and POD 2, and during active and passive physiotherapy 24 h postoperatively. Physiotherapy and early mobilization were standardized as part of the St. Vincent's Hospital Orthopaedic Clinical Pathway. At 24 h postoperatively, two physiotherapists measured passive range of movement, active range of movement, and active assisted range of movement for flexion of the operative knee using a goniometer with the patient in a supine position. VAS was measured at the maximum amount of knee flexion gained. Other markers of early postoperative recovery were recorded, including the ability to sit out of bed on POD 1 and walk with crutches on POD 3, range of movement on POD 4 and day of discharge. Total local anesthetic infused, side effects, and complications were recorded. Other data collected included: height, weight, age, gender, and medications. Personnel blinded to group allocation collected study data. The primary study end-point was postoperative morphine consumption.
Sample size was determined using an expected morphine consumption of 37 ± 35 mg.11 To detect a clinically significant reduction in morphine consumption of 50% with a power of 0.8 and a P value of 0.05 would require 36 patients in each group. To allow for protocol violations or increased variability in data, 82 patients were recruited. P < 0.05 was accepted statistically significant. Statistical analysis was performed using Stata Version 8.2 software (StataCorp, College Station, TX). Normally distributed data were compared using Student's t-test, with Bonferroni correction for multiple comparisons. Other data were compared using Mann–Whitney U, Fisher's exact, or Wilcoxon's ranked sum test as appropriate. Data are expressed as mean (sd) or median (10th–90th centile) as appropriate. Confidence intervals were calculated using Confidence Interval Analysis software, 2nd ed, BMJ.
RESULTS
Eighty-two patients were recruited to this study between May 2005 and October 2006. There were no significant differences in patient or surgical characteristics between groups (Table 1). There was a significant difference between groups in the time required to insert the femoral catheters, with group SC requiring 10 (6–17) min compared with group NSC 6 (3–12) min (P = 0.0001) (Table 2). In group SC, a motor response was obtained during catheter advancement on the first pass of catheter in 8 (20%) patients, during the second to fourth pass of catheter in 11 (27.5%) patients, and required more than 4 passes in 19 (47.5%) patients. No motor response was obtained via catheter in 2 (5%) patients. The motor responses obtained during catheter advancement were patellar ascension/anterior thigh/sartorius/other or none in 29/4/2/5 patients, respectively. There was no significant difference in the number of patients in each group with a sensory block at 20 min after initial injection and on POD1 and 2 (Table 2) or the quality of the femoral blocks, as measured using TES (Fig. 1). However, patients in both groups had significantly higher thresholds (indicating a denser sensory block) at 24 h compared with those at 10 and 20 min. There was no difference in ropivacaine consumption, VAS scores at rest on POD 1 and POD 2, and during active and passive physiotherapy 24 h postoperatively as well as markers of early recovery (Table 3). Successful sciatic nerve blockade was confirmed in all but three patients (two in group NSC, one in group SC).
|
|
|
|
Two patients had inadvertent arterial puncture, with one patient (group NSC) having a painful hematoma around the catheter insertion site requiring early cessation of the infusion and catheter removal on the evening of POD1. The second patient (group SC) had a minor hematoma develop. There were no permanent neurological complications related to femoral or sciatic nerve blockade. Three patients developed pulmonary emboli. Other complications included one patient with acute pulmonary edema in the postanesthesia care unit and one patient with postoperative acute myocardial infarction requiring coronary artery bypass surgery. One patient died 1 mo postoperatively of a suspected acute myocardial infarction at another hospital.
DISCUSSION
In this study, the use of a SC for CFNB after TKA did not improve postoperative analgesia, as indicated by no differences in morphine requirements, pain, TES scores, or early mobilization. There was an increase in procedural time required for insertion of a SC compared with a NSC (10 and 6 min, respectively); however, this is of debatable clinical significance. Evidence regarding SCs is inconsistent. Salinas et al.8 demonstrated improved quality of sensory and motor blockade with SCs in volunteers not undergoing surgery. Their technique used to define quality of sensory blockade was used in this study and there was no difference in TES between study groups, although there was a significant difference between blocked and nonblocked legs in both groups, indicating functional femoral catheters. Morin et al.12 reported similar results to ours: no difference in the quality of analgesia using SC or NSC techniques for CFNB after TKA. However, patients in their study did not receive sciatic nerve blockade, which may have influenced morphine requirements due to posterior knee pain. Morin et al. reported that multiple catheter manipulations were required and in our study more than four passes were required in 53% of patients in group SC, which has a theoretical risk of damaging or even shearing a catheter. There were no such events in our study.
A significant limitation of our study is the use of different regional anesthesia equipment for the study groups. There are differences in catheter gauge and flexibility and, possibly, ease of use. Therefore, this study is potentially not just a comparison of techniques (SC vs. NSC) but also a study comparing two catheter types. The studies by Morin et al. and Salinas et al. used the same SC for both groups, whereas the catheter we used for group NSC was a non-SC. A further limitation of our study is a relatively high local anesthetic infusion rate of 12–14 mL/h, as this may have masked the benefits of superior catheter placement. Reducing the infusion rate also allows a longer duration of infusion in the outpatient setting and reduces the theoretical risk of myotoxicity. In addition, there was a relatively large local anesthetic initial dose (0.375% ropivacaine) preoperatively, and the residual effects of the primary block may have influenced postoperative analgesia. Two other factors potentially influencing postoperative analgesia and interpretation of the results are the use of a multimodal technique of analgesia and lack of placebo group.
SCs have been compared with NSCs for regional blocks other than CFNB. After sciatic blocks for hallux valgus surgery, Rodriguez et al.13 demonstrated reduced pain scores using an infusion rate of 3 mL/h of 0.125% levobupivacaine, while Casati et al.14 documented reduced block onset time and local anesthetic requirements. However, the surgical insult in TKA is extensive compared with hallux valgus surgery. After shoulder surgery, Stevens et al.15 compared SCs with NSCs for interscalene block and demonstrated no difference in postoperative pain; but they did demonstrate faster onset and improved functional recovery after 6 wk.
Ultrasound imaging also is an advancement in continuous peripheral nerve blockade with the potential to image nerves, fascial planes, injectate, and catheter. For femoral nerve or fascia iliaca blockade, once the anesthesiologist is confident that the needle tip is deep to fascia iliaca and in close proximity to the nerve (perhaps further evidenced by patellar movement), the relative importance of SCs or NSCs may not be as significant.
In summary, randomized controlled trials in the clinical environment12,16 have yielded limited evidence (consistent with this study) to justify use of SCs for CFNB after TKA. The increased cost and need for additional catheter adjustments compared with NSCs also make it hard to justify their use. This study indicates that a NSC is as reliable as a SC for establishing and maintaining femoral nerve blockade postoperatively as part of a multimodal analgesia technique after TKA.
ACKNOWLEDGMENTS
The authors acknowledge the assistance of the Anaesthesia nurses in the Operating room, the nursing staff of the Orthopaedic Unit, and Mr. Michael Dooley at St. Vincent's Hospital, Melbourne. Abbott Australasia Pty Ltd. provided the infusion pumps for this study.
Footnotes
Accepted for publication December 6, 2007.
Supported, in part, by Mayo Health Care Pty Ltd.
Presented at the Australian Society of Anaesthetists, Annual Scientific Meeting, October 2006, Coolum, Australia.
REFERENCES
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
