Femoral Nerve Block With Selective Tibial Nerve Block Provides Effective Analgesia Without Foot Drop After Total Knee Arthroplasty: A Prospective, Randomized, Observer-Blinded Study

Sanjay K. Sinha, MB, BS,* Jonathan H. Abrams, MD,* Sivasenthil Arumugam, MB, BS,* John D’Alessio, MD,* David G. Freitas, MD,* John T. Barnett, MD,* and Robert S. Weller, MD†

BACKGROUND: Sciatic nerve block when combined with femoral nerve block for total knee arthroplasty may provide superior analgesia but can produce footdrop, which may mask surgically induced peroneal nerve injury. In this prospective, randomized, observer-blinded study, we evaluated whether performing a selective tibial nerve block in the popliteal fossa would avoid complete peroneal motor block. METHODS: Eighty patients scheduled for primary total knee arthroplasty were randomized to receive either a tibial nerve block in the popliteal fossa or a sciatic nerve block proximal to its bifurcation in combination with femoral nerve block as part of a multimodal analgesia regimen. Local anesthetic solution of sufficient volume to encircle the target nerve was administered for the block, up to a maximum of 20 mL. General anesthesia was administered for surgery. After emergence from anesthesia, in the recovery room, the presence or absence of peroneal sensory and motor block was noted. Pain scores and opioid consumption were recorded for 24 hours after surgery. RESULTS: The tibial nerve block and sciatic nerve block were performed 1.7 cm (99% CI, 1.3 to 2.1) and 9.4 cm (99% CI, 8.3 to 10.5) proximal to the popliteal crease, respectively (99% CI for difference between means: 6.4 to 9.0; P Ͻ 0.001). A lower volume of ropivacaine 0.5% was used for the tibial nerve block, 8.7 mL (99% CI, 7.9 to 9.4) versus 15.2 mL (99% CI, 14.9 to 15.5), respectively (99% CI for difference between means, 5.6 to 7.3; P Ͻ 0.001). No patient receiving a tibial nerve block developed complete peroneal motor block compared to 82.5% of patients with sciatic nerve block (P Ͻ 0.001). There were no significant differences in the pain scores and opioid consumption between the groups. CONCLUSIONS: Tibial nerve block performed in the popliteal fossa in close proximity to the popliteal crease avoided complete peroneal motor block and provided similar postoperative analgesia compared to sciatic nerve block when combined with femoral nerve block for patients undergoing total knee arthroplasty. (Anesth Analg 2012;115:202–6)

he addition of a sciatic nerve block to a femoral nerve usually manifests as a postoperative footdrop or peroneal block after total knee arthroplasty (TKA) may pro- paralysis and, if suspected, the treatment involves releasing vide superior analgesia by reducing posterior knee tight dressings, flexing the knee to relieve stretching of the T 5,6 pain, although some investigators have shown no ben- nerve, and, rarely, surgical exploration. efit.1–3 The sciatic nerve block may, however, cause an The sciatic nerve consists of the tibial nerve and CPN inability to dorsiflex the foot due to blockade of the that diverge at a variable distance above the popliteal common peroneal nerve (CPN), and some surgeons request crease. The CPN carries motor fibers resulting in dorsiflex- no preoperative sciatic nerve block be performed for fear of ion and eversion of the ankle and toes as well as sensory masking a surgical injury to the peroneal nerve with a fibers to the anterolateral leg and dorsum of the foot; injury reported incidence of 0.3% to 10%.4 Peroneal nerve injury to this nerve leads to peroneal paralysis and loss of perception in its sensory field. The nociceptive innervation of the posterior aspect of the knee joint is described as From the *Department of Anesthesiology, Saint Francis Hospital and Medical Center, Hartford, CT; and †Department of Anesthesiology, Wake arising from articular branches of both the tibial nerve and Forest University School of Medicine, Winston-Salem, NC. CPN as they cross the knee, but the relative sclerotomal Accepted for publication February 29, 2012. contribution from each is unknown.7 Funding was provided solely from intramural departmental funding. We hypothesized that blocking only the tibial nerve Conflict of Interest: See Disclosures at the end of the article. component of the sciatic nerve would preserve peroneal Reprints will not be available from the authors. nerve function and still provide adequate pain control This report was previously presented, in part, at the 2010 Annual Spring when combined with femoral nerve block in TKA patients. Meeting of the American Society of Regional Anesthesia, April 22–25, Toronto, ON, Canada, which was the subject of an article in Anesthesiology This prospective, randomized, patient and observer- News. blinded study evaluated whether selective tibial nerve Address correspondence to Sanjay K. Sinha, MB, BS, St. Francis Hospital and block in the popliteal fossa would reduce the incidence of Medical Center, Department of Anesthesiology, 114 Woodland Street, Hartford, CT 06105. Address e-mail to [email protected]. postoperative complete peroneal motor block as the pri- Copyright © 2012 International Anesthesia Research Society mary outcome. Secondary outcome measures included DOI: 10.1213/ANE.0b013e3182536193 pain scores and opioid consumption in the postanesthesia

202 www.anesthesia-analgesia.org July 2012 • Volume 115 • Number 1 Figure 1. Sonogram showing tibial nerve sur- rounded by local anesthetic solution.

care unit (PACU) and during the 24 hours following nerve block just proximal to the popliteal crease where the discharge from the PACU. nerve could be clearly defined. The block needle was advanced in a medial to anterolateral direction toward the METHODS target nerve using an in-plane approach. Nerve stimulation This study was approved by the IRB of St. Francis Hospital was used only to confirm the target nerve and not for and Medical Center, and written informed consent was needle positioning, which was accomplished using ultra- obtained from all subjects. Eighty patients scheduled for sound and injection of sufficient ropivacaine 0.5% to en- primary TKA under general anesthesia with a multimodal circle the target nerve, up to 20 mL. The block needle analgesic regimen including peripheral nerve block for position was adjusted, if necessary, to achieve circumneural postoperative analgesia were studied at St. Francis Hospital spread of the local anesthetic (Fig. 1). The distance from and Medical Center from March 2009 to December 2010. popliteal crease to needle insertion site and the volume of Patients with preexisting valgus deformity, flexion contrac- local anesthetic used for the block were recorded. Subse- ture, neuropathy, diabetes, or allergy to any medication quently, the femoral nerve catheter was injected incremen- used in the study were excluded. Patients were random- tally with 15 mL of ropivacaine 0.5%. ized using random number tables in blocks of 10, and All patients received general anesthesia with a laryngeal group assignments were placed in serially numbered, mask airway for surgery. Anesthesia was induced with opaque-sealed envelopes to receive either a sciatic nerve propofol and the anesthetic was maintained with sevoflu- block (SN group) or selective tibial nerve block (TN group). rane, air, and oxygen. Fentanyl and hydromorphone were Blocks were performed by anesthesiologists experienced in the only opioids allowed without any specific limitation. ultrasound-guided popliteal sciatic block. All patients were After emergence from anesthesia, in the PACU, tibial and premedicated with oral celecoxib 200 mg and pregabalin 50 peroneal sensorimotor function was recorded by PACU mg. Before block placement, standard monitors were ap- nurses blinded to study group. PACU nurses were trained for plied and oxygen 2 L/min via nasal cannula was adminis- sciatic assessment prestudy, and testing was performed when tered. Patients were sedated with no more than midazolam patients were sufficiently awake to respond appropriately. 2 mg and fentanyl 150 ␮g IV and verbal contact was Tibial motor function (plantar flexion of the foot) was tested maintained with the patient throughout the block proce- with a 3-point scale (0 ϭ normal, 1 ϭ weak, 2 ϭ absent), and dure. An ultrasound-guided femoral nerve catheter was sensation to cold was tested on the plantar aspect of the foot inserted using a Tuohy needle, stimulating catheter assembly using a 3-point scale (0 ϭ normal, 1 ϭ absent cold perception (StimuCathTM Arrow Intl, Teleflex Medical, Research Triangle but touch sensation intact, and 2 ϭ absence of touch sensa- Park, NC). After the femoral nerve catheter was placed, the tion). Cold sensation was tested using an alcohol swab com- patients were positioned in the lateral decubitus position with pared to the nonoperative foot. Common peroneal motor the operative leg nondependent. Using sterile prep and function (dorsiflexion of the foot) was tested with a 3-point sheath, a 13 to 6 MHz linear transducer (HFL38, SonositeTM, scale (0 ϭ normal, 1 ϭ weak, and 2 ϭ absent dorsiflexion), and Bothell, WA) was placed at the popliteal crease to visualize sensation to cold was tested on the dorsum of the foot using the popliteal vessels in short axis, and the tibial nerve was a 3-point scale (0 ϭ normal, 1 ϭ absent cold perception but visualized as a hyperechoic oval structure posterior to the touch sensation intact, and 2 ϭ absence of touch sensation). popliteal vessels. Following the tibial nerve cephalad, the Complete motor block of the peroneal nerve was defined as a convergence of the CPN and tibial nerve was identified. peroneal motor score of 2. A verbal numeric pain rating score Patients randomized to the SN group received a sciatic (NRS) (0–10) was recorded on admission and discharge from nerve block just proximal to the bifurcation of the sciatic the PACU. If patients complained of pain, its location was nerve, whereas patients in the TN group received a tibial ascertained; for anterior knee pain the femoral nerve catheter

July 2012 • Volume 115 • Number 1 www.anesthesia-analgesia.org 203 Selective Tibial Nerve Block For Total Knee Arthroplasty

Table 1. Patient Demographics Tibial nerve Sciatic nerve group group P Number of patients 40 40 Age (years) 66.6 Ϯ 8.0 64.5 Ϯ 10.5 0.329 Weight (kg) 89.2 Ϯ 18.5 87.2 Ϯ 22.7 0.671 Height (cm) 171.1 Ϯ 11.4 167.0 Ϯ 12.0 0.126 Gender (F/M) 18/22 24/16 0.179 Values are expressed as mean Ϯ standard deviation. P values were derived from 2-sided 2-sample tests assuming unequal variances. was initially bolused with 10 mL ropivacaine 0.2% and if the pain continued after the bolus, IV hydromorphone was titrated. For patients complaining of pain in a location other than the anterior knee or poorly localized knee pain, IV hydromorphone was initially administered; if pain persisted after 0.5 mg of hydromorphone, their femoral nerve catheter was also bolused with 10 mL of ropivacaine 0.2%. Patients Figure 2. Percentage of patients with complete, partial, and no continued to receive a regimen of oxycontin 10 mg, celecoxib motor and sensory block in the peroneal nerve distribution, compar- 200 mg, and pregabalin 50 mg every 12 hours, and acetamin- ing the tibial nerve (TN) and the sciatic nerve (SN) groups. The SN ophen 975 mg every 6 hours postoperatively. For break- group had more patients with complete peroneal motor and sensory block than the TN group. *P Ͻ 0.001. through pain, oxycodone tablets or IV hydromorphone was administered. On the orthopedic floor, pain scores were recorded every 6 hours for 24 hours after PACU discharge. All pain scores were recorded by nurses blinded to study groups. Opioid consumption during surgery, in the PACU, and 24 hours after PACU discharge was recorded. Patients were followed postoperatively by surgeons and physical therapists although there was no formal assessment tool for complications. Sample size was calculated based on the assumption that complete peroneal motor block would be at minimum 60% in the SN group and no greater than 20% in the TN group. A sample size of 40 per group (assuming a 90% block success rate) would support the ability to detect this difference with a power of 95% (␤ error ϭ 0.05) and 2-tailed ␣ ϭ 0.05 with unequal variances. Normality assumption of distribution was assessed using the Kolmogorov-Smirnov test with Lilliefor’s correction. Variables that followed a normal distribution included patient age, height, and Figure 3. Percentage of patients with complete, partial, and no weight (Lilliefor’s test P Ͼ 0.05). Significance of differences motor and sensory blocks in the tibial nerve distribution, comparing in categorical variables comparing TN and SN groups was the tibial nerve (TN) and sciatic nerve (SN) groups. (No statistically significant differences between groups.). explored with ␹2 tests. t Tests for unequal variances were used to assess group differences in continuous variables. The Mann-Whitney U was used to evaluate group differ- ences in intraoperative and postoperative analgesics ad- statistically significant) (Fig. 2). Similarly, 72.5% of patients ministered and pain scores, because these were not nor- in the SN group had absence of sensation (sensory score of mally distributed. A Bonferroni ␣ correction was used to 2) in the peroneal distribution compared to 2.5% of patients account for the 6 evaluations of pain scores. We used an ␣ in the TN group (P Ͻ 0.001). Both groups had a similar level of 0.05/6 ϭ 0.008 to assess statistical significance. Data number of patients with a sensory score of 1 in the peroneal analyses were conducted using SPSS version 18.0 (SPSS, distribution (25% in SN versus 32.5% in TN). There were no Chicago, IL). P values Ͻ0.05 were considered statistically differences in the sensory and motor block characteristics significant. between the groups in the tibial nerve distribution (Fig. 3). Tibial nerve block was performed closer to the popliteal RESULTS crease (1.7 cm; 99% CI, 1.3 to 2.1) compared to sciatic nerve There were no differences in the demographic variables block (9.4 cm; 99% CI, 8.3 to 10.5); 99% CI for difference of between groups (Table 1). No patient in the TN group means: 6.4 to 9.0; P Ͻ 0.001. A higher volume of ropiva- developed complete peroneal motor block compared to caine, 15.2 mL (99% CI, 14.8 to 15.4) versus 8.7 mL (99% CI, 82.5% of patients in the SN group (P Ͻ 0.001). In the TN 7.9 to 9.4) was required to achieve complete circumneural group, 22.5% of patients developed partial peroneal motor spread in the SN group; 99% CI for difference of means: 5.6 block, compared to 12.5% of patients in the SN group (not to 7.3; P Ͻ 0.001.

204 www.anesthesia-analgesia.org ANESTHESIA & ANALGESIA Table 2. Perioperative Opioid Consumption Tibial nerve Sciatic nerve 99% CI of median group group differences P Intraop fentanyl (␮g) 50 (50, 0–200) 50 (100, 0–200) 0–50 0.670 Intraop hydromorphone (mg) 0.6 (1.0, 0.0–2.0) 0.4 (1.0, 0.0–4.0) 0–1 0.886 Patients needing narcotics in PACU (%) 25 15 0.264 Patients needing Ropi 0.2% 10 ml bolus 30 17.5 0.189 in PACU (%) PACU hydromorphone (mg) 0 (1.0, 0.0–2.0) 0 (0.0, 0.0–2.0) 0–0 0.575 Floor 24-h oxycodone (mg) 10 (20, 0–40) 10 (15, 0–30) 0–10 0.165 Floor 24-h hydromorphone (mg) 0 (0, 0.0–1.0) 0 (0, 0.0–4.0) 0–0 0.365 Values are expressed as medians with interquartile range, minimum and maximum in parentheses or as percentages. Intraop ϭ intraoperative; Ropi ϭ Ropivacaine; PACU ϭ postanesthesia care unit.

Figure 4. Verbal numeric pain score reported by patients on admission (Adm) and discharge (D/C) from the postanesthesia care unit (PACU) and at 6, 12, 18, and 24 h after discharge from the PACU. Data expressed as mean (columns) and SD (whis- kers). Numeric pain score 0 to 10 (0 ϭ no pain, 10 ϭ worst pain). (No statistically significant differ- ences between groups using a Bonferroni corrected 2-sided ␣ level of 0.008).

Intraoperatively, there was no significant difference in dorsiflexion and decreased sensation on the dorsal aspect of the amount of fentanyl and hydromorphone administered, the foot seen in some of our patients, although we did not or in NRS pain score on admission to the PACU with 25% evaluate the spread of local anesthetic with ultrasound after of patients in the TN group and 15% of patients in the SN injection. In addition, the variability of the level of bifurcation of group requiring hydromorphone (P ϭ 0.264) (Table 2). In the sciatic nerve may have made patients with more distal addition, 30% of TN patients and 17.5% of SN patients divergence more vulnerable to partial blockade of the CPN needed rebolusing of their femoral nerve catheter (P ϭ when selective tibial nerve block was performed. Further reduc- 0.189). The time course of pain assessments throughout the tion in volume or concentration of local anesthetic solution at the study is illustrated in Figure 4. There were no significant selective TN site may reduce the degree of common peroneal differences at all points and in the average cumulative pain block, but may also negatively impact the success and duration scores for the 24-hour pain scores on the orthopedic floor of the desired analgesia; this was not tested in this study. scores postoperatively (P ϭ 0.714). Opioid consumption on In our study, 65% of patients in the SN group developed the orthopedic floor was similar between groups (Table 2). complete sensorimotor block in both the tibial and common peroneal distribution. This is comparable to a published DISCUSSION report of complete sensorimotor block in 68% of patients with This study demonstrates that selectively blocking the tibial 20 mL of ropivacaine 0.75%.8 In the TN group, 55% of patients nerve in the popliteal fossa as described with 0.5% ropiva- had complete sensorimotor block in the tibial nerve distribu- caine did not result in complete motor block of the peroneal tion which is lower than the success rate reported by Prasad et nerve in patients having TKA. Some degree of blockade of the al.9 where the tibial nerve and CPN were blocked separately peroneal nerve occurred in the TN group, with 22.5% of those and all patients developed a complete block in 50 minutes in patients showing partial motor block and 32.5% showing both tibial and common peroneal distributions. However, partial sensory block. Although partial peroneal motor block larger volumes of a more highly concentrated local anesthetic might be undesirable in patients and prompt surgeons to (30 mL of 2% lidocaine/0.5% bupivacaine with 1:200,000 request no block, weak dorsiflexion may still allow for serial epinephrine) were used and may explain the more complete monitoring in the postoperative period to detect progression nerve block. The block success for the selective TN group in of peroneal dysfunction and intervention, or recovery. Schin- our study is challenging to compare to other studies using sky et al.5 in their series of patients with nerve injury after different local anesthetics, definition of success, or time to TKA reported that all patients had sensory deficits and assessment. In this study, the rate of complete tibial sensory peroneal paralysis as their initial presentation in the immedi- block was high (72.5% in TN group and 75% SN group), ate postoperative period. Cephalad spread of local anesthetic which provided adequate analgesia postoperatively (Fig. 3). solution from the site of tibial nerve block leading to partial The analgesic benefit from blockade of only the tibial block of the CPN was likely responsible for the weakness of nerve was similar to that of the more proximal sciatic nerve

July 2012 • Volume 115 • Number 1 www.anesthesia-analgesia.org 205 Selective Tibial Nerve Block For Total Knee Arthroplasty

block with equivalent NRS pain scores and opioid con- Name: Sivasenthil Arumugam, MB, BS. sumption. Although articular branches to the knee joint Contribution: This author helped design the study and con- arise from both the tibial nerve and CPN, the results of this duct the study. study suggest that the sensory contribution from the CPN Attestation: Sivasenthil Arumugam approved the final may be clinically less important than from the tibial nerve. manuscript. This study has a number of limitations. Patients were Conflicts of Interest: The author has no conflicts of interest to excluded for ethical reasons if they were at increased risk of declare. peroneal dysfunction, because the ability to avoid peroneal Name: John D’Alessio, MD. paralysis with the selective tibial nerve block could not be Contribution: This author helped design the study and con- assured before study. It remains to be determined if these duct the study. higher risk patients can be managed with selective tibial nerve Attestation: John D’Alessio approved the final manuscript. block and monitored and treated appropriately, but the Conflicts of Interest: The author has no conflicts of interest to results provide an ethical basis to now study this question. declare. The anesthesiologist performing the block could not be Name: David G. Freitas, MD. blinded to study group, and unconscious bias may have been Contribution: This author helped conduct the study. introduced in the performance of the blocks, but the consis- Attestation: David Freitas approved the final manuscript. tency of the results for tibial sensory and motor blocks in the Conflicts of Interest: The author has no conflicts of interest to two groups suggest the block technique was comparable. A declare. different volume of local anesthetic was used in the 2 groups, Name: John T. Barnett, MD. because it is common to use only enough local anesthetic to Contribution: This author helped conduct the study. surround a target nerve when limited spread or minimal Attestation: John Barnett approved the final manuscript. dosage of local anesthetic is the goal.10 If an equal volume had Conflicts of Interest: The author has no conflicts of interest to been used in both groups, and, hence, a larger volume than declare. necessary to surround the tibial nerve selectively in the TN Name: Robert S. Weller, MD. group, more spread to the peroneal component might have Contribution: This author helped design the study, analyze occurred. In addition, a higher concentration of ropivacaine was the data, and write the manuscript. used than many anesthesiologists use for postoperative analge- Attestation: Robert S. Weller reviewed the analysis of the data sia, but this choice increased the potential for complete motor and approved the final manuscript. block of the peroneal nerve and improved the ability to discrimi- Conflicts of Interest: Robert S. Weller reported receiving nate an effect due specifically to the site of injection. In addition, research support from I-Flow Corporation and honoraria from the study may have been underpowered to prove equivalence of Sonosite Corporation. analgesia as a secondary endpoint. The aggressive pre- and This manuscript was handled by: Terese T. Horlocker, MD. postoperative multimodal analgesic regimen routinely used in these patients is likely to have contributed to the low pain scores. REFERENCES In conclusion, this study demonstrated that selective tibial 1. Cook P, Stevens J, Gaudron C. Comparing the effects of nerve block performed below the bifurcation of the sciatic femoral nerve block versus femoral and sciatic nerve block on nerve in the popliteal fossa, instead of sciatic nerve block pain and opiate consumption after total knee arthroplasty. above the bifurcation, avoided complete motor block of the J Arthroplasty 2003;18:583–6 2. Weber A, Fournier R, Van Gessel E, Gamulin Z. Sciatic nerve peroneal nerve and provided comparable pain relief after block and the improvement of femoral nerve block analgesia TKA in patients who were not at increased risk of peroneal after total knee replacement. Eur J Anaesthesiol 2002;19:834–6 injury. It remains to be determined if selective tibial nerve 3. Allen H, Liu S, Ware P, Nairn C, Owens B. Peripheral nerve block will similarly allow for monitoring and intervention in blocks improve analgesia after total knee replacement surgery. patients at high risk for peroneal nerve injury. Anesth Analg 1998;87:93–7 4. Nercessian O, Ugwonali O, Park S. Peroneal nerve palsy after DISCLOSURES total knee arthroplasty. J Arthroplasty, 2005;20:1068–73 5. Schinsky M, Macaulay W, Parks M, Kiernan H, Nercessian O. Name: Sanjay K. Sinha, MB, BS. Nerve injury after primary total knee arthroplasty. J Arthro- Contribution: This author helped design the study, conduct plasty 2001;16:1048–54 the study, analyze the data, and write the manuscript. 6. Dellon AL. Postarthroplasty “palsy” and systemic neuropathy Attestation: Sanjay K. Sinha has seen the original study data, a peripheral-nerve management algorithm. Ann Plast Surg reviewed the analysis of the data, approved the final manuscript, 2005;55:638–42 and is the author responsible for archiving the study files. 7. Amis A, Bull A, Gupte C Knee. In: Standring S, ed. Gray’s Conflicts of Interest: The author has no conflicts of interest to Anatomy: The Anatomical Basis of Clinical Practice, 40th ed. declare. Philadelphia: Elsevier Churchill Livingstone, 2008: 1393–1410 Name: Jonathan H. Abrams, MD. 8. Danelli G, Fanelli A, Ghisi D, Moschini E, Rossi M, Ortu A, Contribution: This author helped design the study, conduct Baciarello M, Fanelli G. Ultrasound vs nerve stimulation multiple injection technique for posterior popliteal sciatic the study, analyze the data, and write the manuscript. nerve block. Anaesthesia 2009;64:638–42 Attestation: Jonathan H. Abrams has seen the original study 9. Prasad A, Perlas A, Ramlogan R, Brull R, Chan V. Ultrasound- data, reviewed the analysis of the data, and approved the final guided popliteal block distal to sciatic nerve bifurcation short- manuscript. ens onset time. Reg Anesth Pain Med. 2010;35:267–71 Conflicts of Interest: The author has no conflicts of interest to 10. Marhofer P, Greher M, Kapral S. Ultrasound guidance in declare. regional anaesthesia. Br J Anaesth 2005;94:7–17

206 www.anesthesia-analgesia.org ANESTHESIA & ANALGESIA