PERIOPERATIVE MEDICINE
Adductor Canal Block versus Femoral Nerve Block for Total Knee Arthroplasty A Prospective, Randomized, Controlled Trial David H. Kim, M.D., Yi Lin, M.D., Ph.D., Enrique A. Goytizolo, M.D., Richard L. Kahn, M.D., Daniel B. Maalouf, M.D., M.P.H., Asha Manohar, M.D., Minda L. Patt, M.D., Amanda K. Goon, B.A., Yuo-yu Lee, M.S., Yan Ma, Ph.D., Jacques T. YaDeau, M.D., Ph.D. ABSTRACT Background: This prospective double-blinded, randomized controlled trial compared adductor canal block (ACB) with femoral nerve block (FNB) in patients undergoing total knee arthroplasty. The authors hypothesized that ACB, compared with FNB, would exhibit less quadriceps weakness and demonstrate noninferior pain score and opioid consumption at 6 to 8 h postanesthesia. Methods: Patients received an ACB or FNB as a component of a multimodal analgesic. Quadriceps strength, pain score, and opioid consumption were assessed on both legs preoperatively and at 6 to 8, 24, and 48 h postanesthesia administration. In a joint hypothesis test, noninferiority was first evaluated on the primary outcomes of strength, pain score, and opioid consumption at 6 to 8 h; superiority on each outcome at 6 to 8 h was then assessed only if noninferiority was established. Results: Forty-six patients received ACB; 47 patients received FNB. At 6 to 8 h postanesthesia, ACB patients had significantly higher median dynamometer readings versus FNB patients (median [interquartile range], 6.1 kgf [3.5, 10.9] (ACB) vs. 0 kgf [0.0, 3.9] (FNB); P < 0.0001), but was not inferior to FNB with regard to Numeric Rating Scale pain scores (1.0 [0.0, 3.5] ACB vs. 0.0 [0.0, 1.0] FNB; P = 0.019), or to opioid consumption (32.2 [22.4, 47.5] ACB vs. 26.6 [19.6, 49.0]; P = 0.0115). At 24 and 48 h postanesthesia, there was no significant statistical difference in dynamometer results, pain scores, or opioid use between the two groups. Conclusion: At 6 to 8 h postanesthesia, the ACB, compared with the FNB, exhibited early relative sparing of quadriceps strength and was not inferior in both providing analgesia or opioid intake. (Anesthesiology 2014; 120:540-50)
O
PTIMAL pain relief is essential for functional recovery after total knee arthroplasty (TKA).1 Addition of femoral nerve block (FNB) to an analgesic regimen provides superior pain control2,3 and shortens hospital stay,4 in comparison with epidural or intravenous patient-controlled analgesia (PCA) alone.1,5,6 However, prolonged motor blockade from FNB is associated with a small (2%) but clinically important risk of fall.7,8 With FNB there will always be a compromise between the goals of adequate pain relief and muscle strength. An ideal nerve block would provide effective analgesia, minimize opioid use and side effects, and hasten mobilization by preserving motor strength. “Fast-track” total joint replacements are gaining popularity. Motor preservation with adequate analgesia has become the optimal postoperative pain goal in orthopedic surgeries to enable earlier physical therapy, faster recovery, and shorter hospital stays.
What We Already Know about This Topic • Despite the improved analgesia and shortened hospital stays provided by the use of femoral nerve blockade after total knee arthroplasty, these blocks can cause significant motor weakness, delaying mobilization and increasing the risk of falls
What This Article Tells Us That Is New • The results of this randomized, blinded trial suggest that adductor canal block results in less motor impairment after surgery, but provides a comparable level of pain relief
With the advent of ultrasonography, the adductor canal can be easily visualized at the mid-thigh level, allowing performance of adductor canal block (ACB) with a high success rate.9,10 In recent years, ACB has been successfully used for postoperative pain control after knee surgery.9,11 Anatomical study of the
This article is featured in “This Month in Anesthesiology,” page 1A. Corresponding article on page 530. Presented in part at the American Society of Regional Anesthesia Annual Meeting, San Diego, California, March 15–18, 2012. Submitted for publication January 10, 2013. Accepted for publication September 30, 2013. From the Department of Anesthesiology, Hospital for Special Surgery, New York, New York (D.H.K., Y.L., E.A.G., R.L.K., D.B.M., A.K.G., and J.T.Y.); Department of Anesthesiology and Critical Care Medicine, The Johns Hopkins University, Baltimore, Maryland (A.M.); Department of Anesthesiology, Weill-Cornell Medical Center, New York, New York (M.L.P.); and Department of Epidemiology and Biostatistics Core, Hospital for Special Surgery, New York, New York (Y.-y.L. and Y.M.). Presented in part at the American Society of Regional Anesthesia Annual Meeting, San Diego, CA, March 15–18, 2012. Copyright © 2014, the American Society of Anesthesiologists, Inc. Lippincott Williams & Wilkins. Anesthesiology 2014; 120:540-50
Anesthesiology, V 120 • No 3 540
March 2014
PERIOPERATIVE MEDICINE
adductor canal demonstrated that the adductor canal may serve as a conduit for more than just the saphenous nerve, possibly including the vastus medialis nerve, medial femoral cutaneous nerve, articular branches from the obturator nerve, as well as the medial retinacular nerve.10–12 Thus, the sensory changes are not limited to the distribution of the saphenous nerve,13 but includes the medial and anterior aspects of the knee from the superior pole of the patella to the proximal tibia. There has not been a randomized control study comparing ACB with FNB after TKA. This prospective, d ouble-blinded, randomized, controlled study tested the hypothesis that ACB would be associated with less quadriceps motor weakness than FNB and provide analgesia that is not inferior as determined by Numeric Rating Scale (NRS) pain scores and opioid use. Using a joint hypothesis test with three primary outcomes, we hypothesize that the ACB is superior in strength but not inferior in pain score and opioid use at 6 to 8 h postanesthesia.
Materials and Methods This study was approved by the Institutional Review Board of Hospital for Special Surgery, New York, New York. The study was registered with ClinicalTrials.gov, Identifier NCT01333943. All patients gave informed written consent. Ninety-four patients scheduled to undergo elective TKA were enrolled in the presurgical area by an anesthesiologist. They were assigned to either ACB or FNB (1:1 allocation, parallel trial design), based on a computer-generated randomization list created by an independent researcher. Group assignment was concealed via opaque envelopes that were opened only after enrollment. Eligibility criteria included elective unilateral TKA, planned combined spinal epidural anesthetic, age 18 to 90 yr, ability to follow study protocol, and American Society of Anesthesiologists class 1 to 3. Exclusion criteria included contraindication for neuraxial anesthetic, chronic opioid use (defined as daily or almost daily use of opioids for >3 months), hypersensitivity and/ or allergies to local anesthetics, intraoperative use of volatile anesthetics, preexisting neuropathy on the operative limb, contraindications to a femoral or ACB, allergy to any of the study medications, aged younger than 18 or older than 90 yr, and American Society of Anesthesiologists class 4 or 5. A research assistant recorded baseline patient demographics and medical history in the presurgical area. Patients were then placed supine with a cushion underneath their knee, resulting in a 45-degree angle at the knee. Quadriceps strength of both legs was assessed by placing the dynamometer on the anterior of the ankle, between the malleoli. Patients were instructed to extend their legs three times each, with a 30-s pause between each attempt (Lafayette Manual Muscle Test System; Lafayette Instrument Company, Lafayette, IN; as described by Maffiuletti14). After each attempt, patients rated their pain using NRS. The patient’s quadriceps were also assessed by a neurologic exam, based on a 12-point scale as described by Bohannon.15 Sensory function along the distribution of the saphenous nerve (medial side of leg above the ankle) was
assessed by pinprick and temperature discrimination using the jagged edges of a broken tongue depressor and an alcohol swab in comparison with the nonoperative side. Patients were randomized to receive either an ACB or FNB. The anesthesiologist performing the block was aware of the treatment, but the patient and the research assistant were blinded to group assignment. An ultrasound-guided ACB (15 cc of 0.5% of bupivacaine with 5 μg/ml epinephrine, via a 21-gauge 4-inch Stimuplex A needle; B. Braun Medical Inc., Melsungen, Germany) was performed at mid-thigh level using a high-frequency linear ultrasound transducer (10–12 Hz; SonoSite Turbo; SonoSite Inc., Bothell, WA), as described by Manickam.10 Ultrasound-guided FNB (30 cc of 0.25% of bupivacaine with 5 μg/ml epinephrine, via a 22-gauge 2-inch Stimuplex A needle; B. Braun Medical Inc.) with nerve stimulator confirmation were performed below the inguinal ligament. The type of motor response (e.g., quadriceps, patellar) and the minimum current needed were recorded. Ultrasound pictures (preinjection and postinjection) were obtained to verify proper local anesthetic placement. All patients received a standardized anesthetic and analgesic. Preoperative oral meloxicam (7.5 or 15 mg based on age; 7.5 mg was given to patients >74 yr) and dexamethasone (6 mg) were given in the holding area. Patients were sedated with intravenously administered midazolam and propofol before performance of the nerve block and epidural placement (opioids and ketamine were not used). Combined spinal epidural anesthesia was administered, with 2.5 cc of 0.5% bupivacaine as the spinal agent. Epidural local anesthetic, if needed, consisted of 2% lidocaine. Ondansetron (4 mg IV) was given during the operation. Intraoperative data included total time to perform the block (starting from needle insertion to exit), surgery time, tourniquet pressure, and total tourniquet time. Oral postoperative pain medications were oxycodone/ acetaminophen (5/325 mg q 4 h as needed) and daily meloxicam (7.5 or 15 mg based on age; 7.5 mg was given to patients >74 yr). Epidural PCA (10 μg/ml hydromorphone, 0.06% bupivacaine) was used for postoperative days (PODs) 0 to 2. Initial settings were 4 ml/h of continuous infusion, 4-ml bolus on demand every 10 min as needed, maximum total of 20 ml/h. At 7 am the following day (POD 1), the continuous infusion was lowered to 2 ml/h, and at 5 pm on POD 1 the continuous infusion was set to 0. At noon on POD 2, the epidural was discontinued. Additional postoperative antiemetics were metoclopramide (10 mg IV every 6 h as needed), and/or ondansetron (4 mg IV every 8 h as needed). At the discretion of the acute pain service, patient’s oral regimens were tailored to address the patient’s pain needs. Quadriceps motor strength, as well as sensory exam (leg pinprick and temperature discrimination) was assessed for both legs at 6 to 8, 24, and 48 h after anesthesia administration. The 6 to 8 h assessment was done in the postanesthesia care unit, whereas the 24- and 48-h assessments were done in the inpatient unit. Block success was verified by testing for pinprick sensation in the saphenous nerve distribution.
Anesthesiology 2014; 120:540-50 541 Kim et al.
Adductor Canal Block versus Femoral Nerve Block for TKA
Physical therapists “dangled” patients (i.e., placed in sitting position with legs on the side of the bed) on POD 0 regardless of block status and patient readiness to ambulate. On POD 1 and afterward, patients were assessed and encouraged to ambulate with assistance. Noninferior analgesia was assessed by measuring both NRS pain scores and opioid consumption, data collected included: (1) NRS pain scores (determined by patient interview, using the standard NRS of 0 to 10, at 6 to 8, 24, and 48 h); and (2) total morphine consumption (converting oral, intravenous, and epidural opioid to morphine equivalent on PODs 0, 1, and 2). To strengthen the claim that the ACB was noninferior in analgesia to the FNB, we decided to conduct a joint hypothesis test using both pain score and opioid consumption as primary outcomes. Additional data collected included: (1) patient satisfaction (patient interviewed, using a scale of 0–10, 10 being the most satisfied, at 6–8 and 24 h); (2) postoperative nausea and vomiting (present or absent, determined by patient interview at 6–8, 24, and 48 h); (3) pruritis, (present or absent, determined by patient interview at 6–8, 24, and 48 h); (4) incidence of complications (if any), including falls, neurologic symptoms, and local anesthetic toxicity; (5) length of hospital stay (days); (6) success of blinding (by asking patients before discharge which treatment they thought they had received. Statistical Analysis Standardized difference was calculated to compare patient demographics and baseline characteristics including age, sex, race, American Society of Anesthesiologists, length of hospital stay, and body mass index between ACB and FNB. An absolute difference greater than 0.2 was considered to be clinically important.16,17 To reduce the chance of confounding, the clinically important variables were adjusted in multiple regression analysis. The primary outcomes included quadriceps muscle strength as measured by dynamometer reading, NRS pain scores, and total opioid consumption. We hypothesized that ACB would be preferred if (1) ACB was noninferior to FNB on all primary outcomes and (2) ACB was superior to FNB at least on quadriceps muscle strength. Therefore we conducted a joint hypothesis testing as described by Mascha and Turan.18 A two-step sequential testing procedure18 was followed for the joint hypothesis testing of (1) and (2), both at 6 to 8 h postanesthesia administration. First, noninferiority was assessed on each individual outcome. Specifically, the following noninferiority was defined for ACB as compared with FNB: (1) the mean dynamometer readings not less than 3 kgf (equivalent to a clinically relevant difference of 20% points as described in Ilfeld et al.19) lower than FNB (2) the mean NRS pain score not more than 1.6 higher than FNB,20 and (3) the mean opioid consumption not more than 50% greater than FNB.2 Second, we evaluated the superiority on each outcome if noninferiority was confirmed on all outcomes. Noninferiority hypotheses were evaluated against a one-sided significance
criterion of 0.025 and superiority hypotheses were evaluated against a one-sided significance criterion of 0.008 (adjusting for the three outcomes, 0.025/3 = 0.008). In addition to the joint hypothesis testing, all primary outcomes were also compared between ACB and FNB at each specific time (baseline, 6–8 h after anesthesia administration, and at PODs 1 and 2). The Holm–Bonferroni stepdown procedure21 was used to control the familywise error rate. The primary outcomes were further studied using multiple regression based on the generalized estimating equation (GEE) method,22,23 adjusting for any clinically important differences that were identified in the baseline variables. In addition, an interaction effect between treatment group and time was also incorporated in the regression analysis. Treatment effect was further assessed at each time point if the interaction effect was significant. For each primary outcome, data collected at baseline (no baseline opioid consumption), 6 to 8, 24, and 48 h postanesthesia were included in the GEE analysis. To reflect the observed correlation structure between repeated measurements, an autoregressive(1) correlation matrix was considered in GEE. The autoregressive correlation structure assumes that measurements closer in time have a higher correlation than those that are further apart.22 The GEE method is able to take into account correlations between repeated measures and does not require a particular distribution for data, leading to robust parameter estimation. Secondary outcomes included side effects (nausea, vomiting) and patient satisfaction. Chi-square test or Fisher exact test was performed to compare the incidence of side effects. Patient satisfaction was analyzed using t test or nonparametric alternative according to the distribution of the data. We powered the study to detect a 50% difference in motor strength as measured by the dynamometer at postanesthesia care unit between the ACB and FNB groups. This was the difference found from an earlier pilot study (unpublished data: The unpublished data were a pilot study done by David Kim, M.D., New York, New York, primarily to estimate the number of patients needed for the study. It was done in August 2010 at the Hospital for Special Surgery. By looking at 10 patients who underwent a total knee replacement under either saphenous nerve block or a FNB (nonrandomized, five patients in each group), motor strength estimates were extrapolated. Quadriceps strength was measured at 6–8 h after the block. From the pilot study, it was estimated that the FNB group would result in at least 50% decrease in motor strength in comparison with the saphenous nerve block.). The mean (61.3 N) and SD (30 N) of dynamometer readings for the FNB group from the pilot study served as reference values for the power analysis. On the basis of a type I error rate of 5% and a power of 80%, and taking into account potential protocol violations and dropouts, we set the target sample size at 47 per group. All analyses used an intention-to-treat approach, in which patients were evaluated in the groups to which they
Anesthesiology 2014; 120:540-50 542 Kim et al.
PERIOPERATIVE MEDICINE
were originally randomly assigned, regardless of the treatment they actually received. SAS version 9.3 (SAS Institute, Cary, NC) was used for all analyses.
Results Patients were enrolled from March 2011 to November 2011. Patient recruitment and flow through the protocol are described in the CONSORT (Consolidated Standards Of Reporting Trials) diagram. Of the 94 patients enrolled, one patient was excluded for inappropriate enrollment (fig. 1). The patient had a profound preexisting neurological condition (significant quadriceps weakness and numbness at baseline) and should not have been enrolled. Four patients were noted to have failed blocks (i.e., no loss of sensation in the saphenous distribution). Success rates for the ACB and FNB were 93.6 and 97.9%, respectively. Three patients did not have an epidural PCA postoperatively (1 intrathecal catheter, 2 spinals only) but an intravenous PCA. Four patients withdrew from the study on POD 0 or 1, but all available data were included in the intention-to-treat analysis. Three patients were not assessed by the research assistant at either the postanesthesia care unit, PODs 1 and/or 2 time points, but all other available data were included in the intentionto-treat analysis. After the exclusion, 46 patients received ACB; 47 patients received FNB. Baseline values were similar between the two groups (table 1), with the exception of the
age group between 60 to 70 yr, Asian race, and obese class I (body mass index between 30 to 35). A joint hypothesis test was performed using all three outcomes as primary (table 2) at the endpoint of 6 to 8 h postanesthesia. All outcomes were found to be noninferior. Specifically, the lower confidence limit of dynamometer readings was greater than the delta (P < 0.0001); the upper confidence limits of NRS pain scores and opioid use were less than their respective deltas (NRS pain scores, P = 0.019; opioid use, P = 0.012). Therefore the ACB was found to not be weaker than the FNB or have higher pain scores or more opioid use. Next, we conducted a superiority test and found only the dynamometer readings for the ACB to be superior to the readings for the FNB (difference ACB-FNB kgf [98.3% CI], 5.2 [2.7–7.7]; P < 0.0001). Therefore, the joint hypothesis test demonstrates that the ACB is superior to the FNB with regard to strength and not inferior with regard to pain score and opioid consumption at 6 to 8 h postanesthesia. We further compared dynamometer readings, NRS pain score, and opioid use between groups at each specific time. At 6 to 8 h postanesthesia, mean strength during extension of the knee from a starting position of 45-degree flexion was significantly higher for the ACB versus the FNB (table 3, difference ACB-FNB kgf [95% CI], 5.2 [3.1–7.2]; P < 0.0001). At 24 and 48 h, the ACB and FNB groups were not statistically significantly different with a P value of 0.9999 at
Assessed for eligibility (n = 231)
Excluded (n = 137) Did not meet inclusion criteria (n = 82) Declined to participate (n = 55) Randomized (n = 94)
FEMORAL NERVE BLOCK Allocated to intervention (n = 47)
ADDUCTOR CANAL BLOCK Allocated to intervention (n = 47)
Excluded from analysis Known neuropathy (n = 1)
Analyzed (n = 47)
Analyzed (n = 46)
Fig. 1. CONSORT (Consolidated Standards Of Reporting Trials) diagram. Flow of patients through the protocol.
Anesthesiology 2014; 120:540-50
543 Kim et al.
Adductor Canal Block versus Femoral Nerve Block for TKA
Table 1. Demographics
Status, N (%) Included in per-protocol analysis Excluded Age, mean ± SD Age, N (%) Age group 1: age ≤50 Age group 2: 50 < age ≤ 60 Age group 3: 60 < age ≤ 70 Age group 4: 70 < age ≤ 80 Age group 5: 80 < age Sex, N (%) Male Female Race, N (%) Asian Black Hispanic White Other/unknown ASA, N (%) 1 2 3 Hospital stay, mean ± SD BMI, mean ± SD BMI, N (%) Normal: 18.5 < BMI < 25 Overweight: 25 ≤ BMI < 30 Obese class I: 30 ≤ BMI < 35 Obese class II: 35 ≤ BMI < 40 Obese class III: 40 ≤ BMI
ACB
FNB
N = 46
N = 47
Standardized Difference
40 (87.0)
39 (83.0)
0.111 —
6 (13.0) 68.0 ± 9.4
8 (17.0) 67.6 ± 11.3
3 (6.5) 6 (13.0) 20 (43.5) 13 (28.3) 4 (8.7)
4 (8.5) 8 (17.0) 14 (29.8) 16 (34.0) 5 (10.6)
22 (47.8) 24 (52.2)
18 (38.3) 29 (61.7)
0 3 (6.5) 2 (4.3) 40 (87.0) 1 (2.2)
1 (2.1) 5 (10.6) 1 (2.1) 39 (83.0) 1 (2.1)
2 (4.3) 38 (82.6) 6 (13.0) 3.7 ± 0.8 29.9 ± 6.4
3 (6.4) 36 (76.6) 8 (17.0) 3.6 ± 0.8 30.3 ± 5.8
13 (28.3) 17 (37.0) 5 (10.9) 6 (13.0) 5 (10.9)
10 (21.3) 16 (34.0) 9 (19.1) 9 (19.1) 3 (6.4)
— 0.043 — 0.075 0.111 0.287 0.125 0.066 0.193 — — — 0.209 0.147 0.126 0.111 0.003 — 0.09 0.15 0.111 0.07 0.072 — 0.162 0.061 0.233 0.167 0.16
A standardized difference >0.2 is considered to be clinically important. ACB = adductor canal block; ASA = American Society of Anesthesiologists; BMI = body mass index; FNB = femoral nerve block.
Table 2. Joint Hypothesis Testing for Outcomes at Postanesthesia 6–8 h Noninferiority One-tailed Test Outcome at Postanesthesia 6–8 h Dynamometer readings NRS pain scores at rest Opioids (oral opioids + PCA)
ACB
FNB
Difference
N = 46
N = 47
ACB-FNB (95% CI)*
7.3 ± 5.4 6.1 [3.5, 10.9] 1.7 ± 1.9 1.0 [0.0, 3.5] 36.6 ± 17.9 32.2 [22.4, 47.5]
2.2 ± 3.8 0.0 [0.0, 3.9] 0.9 ± 1.8 0.0 [0.0, 1.0] 35.8 ± 20.7 26.6 [19.6, 49.0]
Superiority One-tailed Test Difference ACB-FNB (98.3% CI)†
P Value (Holm–Bonferroni)
Delta
P Value
−3
Adductor Canal Block versus Femoral Nerve Block for Total Knee Arthroplasty A Prospective, Randomized, Controlled Trial David H. Kim, M.D., Yi Lin, M.D., Ph.D., Enrique A. Goytizolo, M.D., Richard L. Kahn, M.D., Daniel B. Maalouf, M.D., M.P.H., Asha Manohar, M.D., Minda L. Patt, M.D., Amanda K. Goon, B.A., Yuo-yu Lee, M.S., Yan Ma, Ph.D., Jacques T. YaDeau, M.D., Ph.D. ABSTRACT Background: This prospective double-blinded, randomized controlled trial compared adductor canal block (ACB) with femoral nerve block (FNB) in patients undergoing total knee arthroplasty. The authors hypothesized that ACB, compared with FNB, would exhibit less quadriceps weakness and demonstrate noninferior pain score and opioid consumption at 6 to 8 h postanesthesia. Methods: Patients received an ACB or FNB as a component of a multimodal analgesic. Quadriceps strength, pain score, and opioid consumption were assessed on both legs preoperatively and at 6 to 8, 24, and 48 h postanesthesia administration. In a joint hypothesis test, noninferiority was first evaluated on the primary outcomes of strength, pain score, and opioid consumption at 6 to 8 h; superiority on each outcome at 6 to 8 h was then assessed only if noninferiority was established. Results: Forty-six patients received ACB; 47 patients received FNB. At 6 to 8 h postanesthesia, ACB patients had significantly higher median dynamometer readings versus FNB patients (median [interquartile range], 6.1 kgf [3.5, 10.9] (ACB) vs. 0 kgf [0.0, 3.9] (FNB); P < 0.0001), but was not inferior to FNB with regard to Numeric Rating Scale pain scores (1.0 [0.0, 3.5] ACB vs. 0.0 [0.0, 1.0] FNB; P = 0.019), or to opioid consumption (32.2 [22.4, 47.5] ACB vs. 26.6 [19.6, 49.0]; P = 0.0115). At 24 and 48 h postanesthesia, there was no significant statistical difference in dynamometer results, pain scores, or opioid use between the two groups. Conclusion: At 6 to 8 h postanesthesia, the ACB, compared with the FNB, exhibited early relative sparing of quadriceps strength and was not inferior in both providing analgesia or opioid intake. (Anesthesiology 2014; 120:540-50)
O
PTIMAL pain relief is essential for functional recovery after total knee arthroplasty (TKA).1 Addition of femoral nerve block (FNB) to an analgesic regimen provides superior pain control2,3 and shortens hospital stay,4 in comparison with epidural or intravenous patient-controlled analgesia (PCA) alone.1,5,6 However, prolonged motor blockade from FNB is associated with a small (2%) but clinically important risk of fall.7,8 With FNB there will always be a compromise between the goals of adequate pain relief and muscle strength. An ideal nerve block would provide effective analgesia, minimize opioid use and side effects, and hasten mobilization by preserving motor strength. “Fast-track” total joint replacements are gaining popularity. Motor preservation with adequate analgesia has become the optimal postoperative pain goal in orthopedic surgeries to enable earlier physical therapy, faster recovery, and shorter hospital stays.
What We Already Know about This Topic • Despite the improved analgesia and shortened hospital stays provided by the use of femoral nerve blockade after total knee arthroplasty, these blocks can cause significant motor weakness, delaying mobilization and increasing the risk of falls
What This Article Tells Us That Is New • The results of this randomized, blinded trial suggest that adductor canal block results in less motor impairment after surgery, but provides a comparable level of pain relief
With the advent of ultrasonography, the adductor canal can be easily visualized at the mid-thigh level, allowing performance of adductor canal block (ACB) with a high success rate.9,10 In recent years, ACB has been successfully used for postoperative pain control after knee surgery.9,11 Anatomical study of the
This article is featured in “This Month in Anesthesiology,” page 1A. Corresponding article on page 530. Presented in part at the American Society of Regional Anesthesia Annual Meeting, San Diego, California, March 15–18, 2012. Submitted for publication January 10, 2013. Accepted for publication September 30, 2013. From the Department of Anesthesiology, Hospital for Special Surgery, New York, New York (D.H.K., Y.L., E.A.G., R.L.K., D.B.M., A.K.G., and J.T.Y.); Department of Anesthesiology and Critical Care Medicine, The Johns Hopkins University, Baltimore, Maryland (A.M.); Department of Anesthesiology, Weill-Cornell Medical Center, New York, New York (M.L.P.); and Department of Epidemiology and Biostatistics Core, Hospital for Special Surgery, New York, New York (Y.-y.L. and Y.M.). Presented in part at the American Society of Regional Anesthesia Annual Meeting, San Diego, CA, March 15–18, 2012. Copyright © 2014, the American Society of Anesthesiologists, Inc. Lippincott Williams & Wilkins. Anesthesiology 2014; 120:540-50
Anesthesiology, V 120 • No 3 540
March 2014
PERIOPERATIVE MEDICINE
adductor canal demonstrated that the adductor canal may serve as a conduit for more than just the saphenous nerve, possibly including the vastus medialis nerve, medial femoral cutaneous nerve, articular branches from the obturator nerve, as well as the medial retinacular nerve.10–12 Thus, the sensory changes are not limited to the distribution of the saphenous nerve,13 but includes the medial and anterior aspects of the knee from the superior pole of the patella to the proximal tibia. There has not been a randomized control study comparing ACB with FNB after TKA. This prospective, d ouble-blinded, randomized, controlled study tested the hypothesis that ACB would be associated with less quadriceps motor weakness than FNB and provide analgesia that is not inferior as determined by Numeric Rating Scale (NRS) pain scores and opioid use. Using a joint hypothesis test with three primary outcomes, we hypothesize that the ACB is superior in strength but not inferior in pain score and opioid use at 6 to 8 h postanesthesia.
Materials and Methods This study was approved by the Institutional Review Board of Hospital for Special Surgery, New York, New York. The study was registered with ClinicalTrials.gov, Identifier NCT01333943. All patients gave informed written consent. Ninety-four patients scheduled to undergo elective TKA were enrolled in the presurgical area by an anesthesiologist. They were assigned to either ACB or FNB (1:1 allocation, parallel trial design), based on a computer-generated randomization list created by an independent researcher. Group assignment was concealed via opaque envelopes that were opened only after enrollment. Eligibility criteria included elective unilateral TKA, planned combined spinal epidural anesthetic, age 18 to 90 yr, ability to follow study protocol, and American Society of Anesthesiologists class 1 to 3. Exclusion criteria included contraindication for neuraxial anesthetic, chronic opioid use (defined as daily or almost daily use of opioids for >3 months), hypersensitivity and/ or allergies to local anesthetics, intraoperative use of volatile anesthetics, preexisting neuropathy on the operative limb, contraindications to a femoral or ACB, allergy to any of the study medications, aged younger than 18 or older than 90 yr, and American Society of Anesthesiologists class 4 or 5. A research assistant recorded baseline patient demographics and medical history in the presurgical area. Patients were then placed supine with a cushion underneath their knee, resulting in a 45-degree angle at the knee. Quadriceps strength of both legs was assessed by placing the dynamometer on the anterior of the ankle, between the malleoli. Patients were instructed to extend their legs three times each, with a 30-s pause between each attempt (Lafayette Manual Muscle Test System; Lafayette Instrument Company, Lafayette, IN; as described by Maffiuletti14). After each attempt, patients rated their pain using NRS. The patient’s quadriceps were also assessed by a neurologic exam, based on a 12-point scale as described by Bohannon.15 Sensory function along the distribution of the saphenous nerve (medial side of leg above the ankle) was
assessed by pinprick and temperature discrimination using the jagged edges of a broken tongue depressor and an alcohol swab in comparison with the nonoperative side. Patients were randomized to receive either an ACB or FNB. The anesthesiologist performing the block was aware of the treatment, but the patient and the research assistant were blinded to group assignment. An ultrasound-guided ACB (15 cc of 0.5% of bupivacaine with 5 μg/ml epinephrine, via a 21-gauge 4-inch Stimuplex A needle; B. Braun Medical Inc., Melsungen, Germany) was performed at mid-thigh level using a high-frequency linear ultrasound transducer (10–12 Hz; SonoSite Turbo; SonoSite Inc., Bothell, WA), as described by Manickam.10 Ultrasound-guided FNB (30 cc of 0.25% of bupivacaine with 5 μg/ml epinephrine, via a 22-gauge 2-inch Stimuplex A needle; B. Braun Medical Inc.) with nerve stimulator confirmation were performed below the inguinal ligament. The type of motor response (e.g., quadriceps, patellar) and the minimum current needed were recorded. Ultrasound pictures (preinjection and postinjection) were obtained to verify proper local anesthetic placement. All patients received a standardized anesthetic and analgesic. Preoperative oral meloxicam (7.5 or 15 mg based on age; 7.5 mg was given to patients >74 yr) and dexamethasone (6 mg) were given in the holding area. Patients were sedated with intravenously administered midazolam and propofol before performance of the nerve block and epidural placement (opioids and ketamine were not used). Combined spinal epidural anesthesia was administered, with 2.5 cc of 0.5% bupivacaine as the spinal agent. Epidural local anesthetic, if needed, consisted of 2% lidocaine. Ondansetron (4 mg IV) was given during the operation. Intraoperative data included total time to perform the block (starting from needle insertion to exit), surgery time, tourniquet pressure, and total tourniquet time. Oral postoperative pain medications were oxycodone/ acetaminophen (5/325 mg q 4 h as needed) and daily meloxicam (7.5 or 15 mg based on age; 7.5 mg was given to patients >74 yr). Epidural PCA (10 μg/ml hydromorphone, 0.06% bupivacaine) was used for postoperative days (PODs) 0 to 2. Initial settings were 4 ml/h of continuous infusion, 4-ml bolus on demand every 10 min as needed, maximum total of 20 ml/h. At 7 am the following day (POD 1), the continuous infusion was lowered to 2 ml/h, and at 5 pm on POD 1 the continuous infusion was set to 0. At noon on POD 2, the epidural was discontinued. Additional postoperative antiemetics were metoclopramide (10 mg IV every 6 h as needed), and/or ondansetron (4 mg IV every 8 h as needed). At the discretion of the acute pain service, patient’s oral regimens were tailored to address the patient’s pain needs. Quadriceps motor strength, as well as sensory exam (leg pinprick and temperature discrimination) was assessed for both legs at 6 to 8, 24, and 48 h after anesthesia administration. The 6 to 8 h assessment was done in the postanesthesia care unit, whereas the 24- and 48-h assessments were done in the inpatient unit. Block success was verified by testing for pinprick sensation in the saphenous nerve distribution.
Anesthesiology 2014; 120:540-50 541 Kim et al.
Adductor Canal Block versus Femoral Nerve Block for TKA
Physical therapists “dangled” patients (i.e., placed in sitting position with legs on the side of the bed) on POD 0 regardless of block status and patient readiness to ambulate. On POD 1 and afterward, patients were assessed and encouraged to ambulate with assistance. Noninferior analgesia was assessed by measuring both NRS pain scores and opioid consumption, data collected included: (1) NRS pain scores (determined by patient interview, using the standard NRS of 0 to 10, at 6 to 8, 24, and 48 h); and (2) total morphine consumption (converting oral, intravenous, and epidural opioid to morphine equivalent on PODs 0, 1, and 2). To strengthen the claim that the ACB was noninferior in analgesia to the FNB, we decided to conduct a joint hypothesis test using both pain score and opioid consumption as primary outcomes. Additional data collected included: (1) patient satisfaction (patient interviewed, using a scale of 0–10, 10 being the most satisfied, at 6–8 and 24 h); (2) postoperative nausea and vomiting (present or absent, determined by patient interview at 6–8, 24, and 48 h); (3) pruritis, (present or absent, determined by patient interview at 6–8, 24, and 48 h); (4) incidence of complications (if any), including falls, neurologic symptoms, and local anesthetic toxicity; (5) length of hospital stay (days); (6) success of blinding (by asking patients before discharge which treatment they thought they had received. Statistical Analysis Standardized difference was calculated to compare patient demographics and baseline characteristics including age, sex, race, American Society of Anesthesiologists, length of hospital stay, and body mass index between ACB and FNB. An absolute difference greater than 0.2 was considered to be clinically important.16,17 To reduce the chance of confounding, the clinically important variables were adjusted in multiple regression analysis. The primary outcomes included quadriceps muscle strength as measured by dynamometer reading, NRS pain scores, and total opioid consumption. We hypothesized that ACB would be preferred if (1) ACB was noninferior to FNB on all primary outcomes and (2) ACB was superior to FNB at least on quadriceps muscle strength. Therefore we conducted a joint hypothesis testing as described by Mascha and Turan.18 A two-step sequential testing procedure18 was followed for the joint hypothesis testing of (1) and (2), both at 6 to 8 h postanesthesia administration. First, noninferiority was assessed on each individual outcome. Specifically, the following noninferiority was defined for ACB as compared with FNB: (1) the mean dynamometer readings not less than 3 kgf (equivalent to a clinically relevant difference of 20% points as described in Ilfeld et al.19) lower than FNB (2) the mean NRS pain score not more than 1.6 higher than FNB,20 and (3) the mean opioid consumption not more than 50% greater than FNB.2 Second, we evaluated the superiority on each outcome if noninferiority was confirmed on all outcomes. Noninferiority hypotheses were evaluated against a one-sided significance
criterion of 0.025 and superiority hypotheses were evaluated against a one-sided significance criterion of 0.008 (adjusting for the three outcomes, 0.025/3 = 0.008). In addition to the joint hypothesis testing, all primary outcomes were also compared between ACB and FNB at each specific time (baseline, 6–8 h after anesthesia administration, and at PODs 1 and 2). The Holm–Bonferroni stepdown procedure21 was used to control the familywise error rate. The primary outcomes were further studied using multiple regression based on the generalized estimating equation (GEE) method,22,23 adjusting for any clinically important differences that were identified in the baseline variables. In addition, an interaction effect between treatment group and time was also incorporated in the regression analysis. Treatment effect was further assessed at each time point if the interaction effect was significant. For each primary outcome, data collected at baseline (no baseline opioid consumption), 6 to 8, 24, and 48 h postanesthesia were included in the GEE analysis. To reflect the observed correlation structure between repeated measurements, an autoregressive(1) correlation matrix was considered in GEE. The autoregressive correlation structure assumes that measurements closer in time have a higher correlation than those that are further apart.22 The GEE method is able to take into account correlations between repeated measures and does not require a particular distribution for data, leading to robust parameter estimation. Secondary outcomes included side effects (nausea, vomiting) and patient satisfaction. Chi-square test or Fisher exact test was performed to compare the incidence of side effects. Patient satisfaction was analyzed using t test or nonparametric alternative according to the distribution of the data. We powered the study to detect a 50% difference in motor strength as measured by the dynamometer at postanesthesia care unit between the ACB and FNB groups. This was the difference found from an earlier pilot study (unpublished data: The unpublished data were a pilot study done by David Kim, M.D., New York, New York, primarily to estimate the number of patients needed for the study. It was done in August 2010 at the Hospital for Special Surgery. By looking at 10 patients who underwent a total knee replacement under either saphenous nerve block or a FNB (nonrandomized, five patients in each group), motor strength estimates were extrapolated. Quadriceps strength was measured at 6–8 h after the block. From the pilot study, it was estimated that the FNB group would result in at least 50% decrease in motor strength in comparison with the saphenous nerve block.). The mean (61.3 N) and SD (30 N) of dynamometer readings for the FNB group from the pilot study served as reference values for the power analysis. On the basis of a type I error rate of 5% and a power of 80%, and taking into account potential protocol violations and dropouts, we set the target sample size at 47 per group. All analyses used an intention-to-treat approach, in which patients were evaluated in the groups to which they
Anesthesiology 2014; 120:540-50 542 Kim et al.
PERIOPERATIVE MEDICINE
were originally randomly assigned, regardless of the treatment they actually received. SAS version 9.3 (SAS Institute, Cary, NC) was used for all analyses.
Results Patients were enrolled from March 2011 to November 2011. Patient recruitment and flow through the protocol are described in the CONSORT (Consolidated Standards Of Reporting Trials) diagram. Of the 94 patients enrolled, one patient was excluded for inappropriate enrollment (fig. 1). The patient had a profound preexisting neurological condition (significant quadriceps weakness and numbness at baseline) and should not have been enrolled. Four patients were noted to have failed blocks (i.e., no loss of sensation in the saphenous distribution). Success rates for the ACB and FNB were 93.6 and 97.9%, respectively. Three patients did not have an epidural PCA postoperatively (1 intrathecal catheter, 2 spinals only) but an intravenous PCA. Four patients withdrew from the study on POD 0 or 1, but all available data were included in the intention-to-treat analysis. Three patients were not assessed by the research assistant at either the postanesthesia care unit, PODs 1 and/or 2 time points, but all other available data were included in the intentionto-treat analysis. After the exclusion, 46 patients received ACB; 47 patients received FNB. Baseline values were similar between the two groups (table 1), with the exception of the
age group between 60 to 70 yr, Asian race, and obese class I (body mass index between 30 to 35). A joint hypothesis test was performed using all three outcomes as primary (table 2) at the endpoint of 6 to 8 h postanesthesia. All outcomes were found to be noninferior. Specifically, the lower confidence limit of dynamometer readings was greater than the delta (P < 0.0001); the upper confidence limits of NRS pain scores and opioid use were less than their respective deltas (NRS pain scores, P = 0.019; opioid use, P = 0.012). Therefore the ACB was found to not be weaker than the FNB or have higher pain scores or more opioid use. Next, we conducted a superiority test and found only the dynamometer readings for the ACB to be superior to the readings for the FNB (difference ACB-FNB kgf [98.3% CI], 5.2 [2.7–7.7]; P < 0.0001). Therefore, the joint hypothesis test demonstrates that the ACB is superior to the FNB with regard to strength and not inferior with regard to pain score and opioid consumption at 6 to 8 h postanesthesia. We further compared dynamometer readings, NRS pain score, and opioid use between groups at each specific time. At 6 to 8 h postanesthesia, mean strength during extension of the knee from a starting position of 45-degree flexion was significantly higher for the ACB versus the FNB (table 3, difference ACB-FNB kgf [95% CI], 5.2 [3.1–7.2]; P < 0.0001). At 24 and 48 h, the ACB and FNB groups were not statistically significantly different with a P value of 0.9999 at
Assessed for eligibility (n = 231)
Excluded (n = 137) Did not meet inclusion criteria (n = 82) Declined to participate (n = 55) Randomized (n = 94)
FEMORAL NERVE BLOCK Allocated to intervention (n = 47)
ADDUCTOR CANAL BLOCK Allocated to intervention (n = 47)
Excluded from analysis Known neuropathy (n = 1)
Analyzed (n = 47)
Analyzed (n = 46)
Fig. 1. CONSORT (Consolidated Standards Of Reporting Trials) diagram. Flow of patients through the protocol.
Anesthesiology 2014; 120:540-50
543 Kim et al.
Adductor Canal Block versus Femoral Nerve Block for TKA
Table 1. Demographics
Status, N (%) Included in per-protocol analysis Excluded Age, mean ± SD Age, N (%) Age group 1: age ≤50 Age group 2: 50 < age ≤ 60 Age group 3: 60 < age ≤ 70 Age group 4: 70 < age ≤ 80 Age group 5: 80 < age Sex, N (%) Male Female Race, N (%) Asian Black Hispanic White Other/unknown ASA, N (%) 1 2 3 Hospital stay, mean ± SD BMI, mean ± SD BMI, N (%) Normal: 18.5 < BMI < 25 Overweight: 25 ≤ BMI < 30 Obese class I: 30 ≤ BMI < 35 Obese class II: 35 ≤ BMI < 40 Obese class III: 40 ≤ BMI
ACB
FNB
N = 46
N = 47
Standardized Difference
40 (87.0)
39 (83.0)
0.111 —
6 (13.0) 68.0 ± 9.4
8 (17.0) 67.6 ± 11.3
3 (6.5) 6 (13.0) 20 (43.5) 13 (28.3) 4 (8.7)
4 (8.5) 8 (17.0) 14 (29.8) 16 (34.0) 5 (10.6)
22 (47.8) 24 (52.2)
18 (38.3) 29 (61.7)
0 3 (6.5) 2 (4.3) 40 (87.0) 1 (2.2)
1 (2.1) 5 (10.6) 1 (2.1) 39 (83.0) 1 (2.1)
2 (4.3) 38 (82.6) 6 (13.0) 3.7 ± 0.8 29.9 ± 6.4
3 (6.4) 36 (76.6) 8 (17.0) 3.6 ± 0.8 30.3 ± 5.8
13 (28.3) 17 (37.0) 5 (10.9) 6 (13.0) 5 (10.9)
10 (21.3) 16 (34.0) 9 (19.1) 9 (19.1) 3 (6.4)
— 0.043 — 0.075 0.111 0.287 0.125 0.066 0.193 — — — 0.209 0.147 0.126 0.111 0.003 — 0.09 0.15 0.111 0.07 0.072 — 0.162 0.061 0.233 0.167 0.16
A standardized difference >0.2 is considered to be clinically important. ACB = adductor canal block; ASA = American Society of Anesthesiologists; BMI = body mass index; FNB = femoral nerve block.
Table 2. Joint Hypothesis Testing for Outcomes at Postanesthesia 6–8 h Noninferiority One-tailed Test Outcome at Postanesthesia 6–8 h Dynamometer readings NRS pain scores at rest Opioids (oral opioids + PCA)
ACB
FNB
Difference
N = 46
N = 47
ACB-FNB (95% CI)*
7.3 ± 5.4 6.1 [3.5, 10.9] 1.7 ± 1.9 1.0 [0.0, 3.5] 36.6 ± 17.9 32.2 [22.4, 47.5]
2.2 ± 3.8 0.0 [0.0, 3.9] 0.9 ± 1.8 0.0 [0.0, 1.0] 35.8 ± 20.7 26.6 [19.6, 49.0]
Superiority One-tailed Test Difference ACB-FNB (98.3% CI)†
P Value (Holm–Bonferroni)
Delta
P Value
−3
