The Effects of Femoral Nerve Blockade in Conjunction with Epidural Analgesia After Total Knee Arthroplasty

YaDeau, Jacques T. MD, PhD*; Cahill, Janet B. PT†; Zawadsky, Mark W. MD‡; Sharrock, Nigel E. MBChB*; Bottner, Friedrich MD‡; Morelli, Christine M. BS*; Kahn, Richard L. MD*; Sculco, Thomas P. MD‡

Anesthesia & Analgesia:
doi: 10.1213/01.ANE.0000159150.79908.21
Regional Anesthesia: Research Report

Either epidural analgesia or femoral nerve blockade improves analgesia and rehabilitation after total knee arthroplasty. No study has evaluated the combination of femoral nerve blockade and epidural analgesia. In this prospective, randomized, blinded study we investigated combining femoral nerve blockade with epidural analgesia. Forty-one patients received a single-injection femoral nerve block with 0.375% bupivacaine and 5 μg/mL epinephrine; 39 patients served as controls. All patients received combined spinal-epidural anesthesia and patient-controlled epidural analgesia with 0.06% bupivacaine and 10 μg/mL hydromorphone. Average duration of epidural analgesia was 2 days. All patients received the same standardized physical therapy intervention. Median visual analog scale (VAS) scores with physical therapy were significantly lower for 2 days among patients who received a femoral nerve block versus controls: 3 versus 4 (day 1), 2.5 versus 4 (day 2); P < 0.05. Median VAS pain scores at rest were 0 in both groups on days 1 and 2. Flexion range of motion was improved on postoperative day 2 (70° versus 63°; P < 0.05). No peripheral neuropathies occurred. We conclude that the addition of femoral nerve blockade to epidural analgesia significantly improved analgesia for the first 2 days after total knee arthroplasty.

In Brief

IMPLICATIONS: The addition of femoral nerve blockade to epidural analgesia significantly improved analgesia for the first 2 days after total knee arthroplasty. Femoral nerve blockade improved flexion on postoperative day 2 but did not affect achievement of other physical therapy milestones.

Author Information

Departments of *Anesthesia, †Rehabilitation, and ‡Orthopaedic Surgery, Hospital for Special Surgery, Weill Medical College of Cornell University, New York, New York

Supported, in part, by the Research Fund of the Anesthesiology Department of the Hospital for Special Surgery.

Presented, in part, at the American Academy of Orthopedic Surgeons, March 2004, San Francisco, California.

Accepted for publication January 28, 2005.

Address correspondence to Jacques T. YaDeau, Hospital for Special Surgery, Weill Medical College of Cornell University, 535 E. 70th St., New York, NY 10021. Address e-mail to

Article Outline

Total knee arthroplasty (TKA) can cause severe postoperative pain (1). Improved control of postoperative pain facilitates more rapid achievement of functional outcomes (2,3). Multiple techniques of postoperative pain control have been used after TKA, including oral or IM opioids, patient-controlled IV opioids, patient-controlled epidural analgesia (PCEA), and single-dose or continuous femoral nerve block (CFNB). Reliance on opioids has been associated with inadequate pain control and frequent side effects such as nausea, sedation, and confusion (1). Postoperative epidural analgesia provided pain relief that was superior to pain relief from IV opioids (1,2,4). The addition of a FNB to postoperative IV analgesia after TKA resulted in significantly better pain control, less frequent morphine-related side effects, more rapid achievement of physical therapy (PT) milestones, and shorter hospital stay (5). However, other investigators found that FNB did not improve analgesia after TKA except in the recovery room (6).

In comparison with epidural analgesia after TKA, CFNB plus sciatic nerve block caused less nausea and had more patients with no pain (7). Analgesia after CFNB without a sciatic nerve block was equivalent (8) or inferior (9) to epidural analgesia. CFNB alone did not reduce nausea compared to epidural analgesia (8,9). CFNB plus sciatic nerve block allowed earlier ambulation and earlier discharge (versus epidural) (7), but CFNB alone did not improve rehabilitation versus epidural (8). Epidural 1% lidocaine (with clonidine and morphine) caused more dysesthesia, urinary retention, and hypotension than occurred when the same mixture was used for CFNB (9). CFNB with bupivacaine improved range of motion (ROM) on day 1 but did not reduce pain or side effects when compared with placebo (10). CFNB may (8,9), or may not (7) cause less urinary retention than epidural analgesia. One study found a 40% rate of epidural catheter-related problems (lateralized, kinked, or difficult insertion) versus 0% with CFNB (8). However, insertion of CFNB may have an 18% complication rate (11).

In our review of the literature we did not find studies of a FNB combined with PCEA after TKA. We investigated whether the addition of a single injection FNB to continuous epidural analgesia improved postoperative pain control, accelerated achievement of rehabilitation milestones, or lessened side effects.

Back to Top | Article Outline


This prospective, randomized, blinded study was approved by the Hospital for Special Surgery IRB. The primary outcome variable was flexion ROM (a major rehabilitation functional milestone). A Student’s t-test sample size calculation (Sigma Stat, Jandel Scientific) using published data (9) indicated that 80 patients gave a power of 0.838 to detect a 10° difference in mean flexion ROM on day 2, with a 15° standard deviation, α of 0.05. The 10° difference was chosen because it exceeds the accepted 5° variability in goniometric measures and because, clinically, a 10° increase in knee flexion allows the patient to more closely approximate the ROM necessary for functional tasks (12). Secondary outcome variables included rehabilitation functional milestones, pain reported by patients (visual analog scale [VAS] scores at rest and with PT), and side effects such as nausea, pruritus, and confusion. PT outcomes measured included knee flexion ROM, continuous passive motion range, ambulation distance, postoperative day (POD) achieving unassisted transfer in and out of bed, walker ambulation, cane ambulation, non-reciprocal stair negotiation, and length of stay.

After written informed consent was obtained, 80 patients of a single surgeon were enrolled. Inclusion criteria consisted of primary TKA for a diagnosis of osteoarthritis and age <85 yr. Exclusion criteria consisted of previous knee trauma, previous surgery to the operative knee including arthroscopic or open meniscectomy, peripheral neuropathy, chronic preoperative opioid usage, a nonpalpable femoral artery, or previous lower extremity vascular bypass surgery.

Patients were randomized into two groups using computer-generated random numbers and a closed envelope design. The patients, the physical therapists, the pain management team, and the chart analysts were not informed of the patients’ group assignment. Patients were told during the informed consent discussion that a FNB, if performed, would result in motor blockade and decreased sensation of the anterior thigh. Forty-one patients received a single-dose FNB before combined spinal-epidural anesthesia. Extent and type of sedation before the FNB was at the discretion of the anesthesiologist. A 22-gauge insulated Stimuplex (B Braun) needle was inserted lateral to the femoral pulse at the level of the inguinal crease. Bupivacaine 0.375% with 5 μg/mL epinephrine, 30 mL, was injected after eliciting a quadriceps contraction at <0.5 mAmp. The FNB was performed in the operating room (OR). Success of the block could not be evaluated in the OR, as the combined spinal-epidural anesthetic was initiated immediately after conclusion of the FNB. Thirty-nine control patients did not receive a FNB. The block was not evaluated after surgery, as we wished to preserve the blinding of the patients as much as possible.

Patients in both groups received combined spinal-epidural anesthesia with subarachnoid administration of 12.5 mg of 0.5% bupivacaine. The combined spinal-epidural anesthetic was sufficient for surgery in all patients. The epidural catheter was not used during surgery. Postoperative PCEA was provided with 0.06% bupivacaine and 10 μg/mL hydromorphone. Initial settings were a continuous rate of 3–6 mL/h with 4 additional 5 mL patient-controlled boluses allowed per hour. PCEA settings were subsequently adjusted by the pain service as needed. Patients were offered oral analgesics (hydrocodone/acetaminophen or oxycodone/acetaminophen) on the second POD. Epidural analgesia was discontinued after successful transition to oral analgesics, typically on either the second or the third POD. Urinary catheters were placed in all patients in the recovery room and were typically continued until the epidural catheter was removed. As all patients had urinary catheters, effects of the analgesic technique on urinary retention were not evaluated.

All patients in both groups received the same standardized postoperative PT. A PT evaluation in the recovery room of lower extremity strength and sensation was recorded. Patients’ lower extremity strength and sensation were reassessed daily. Although formal quadriceps strength was not assessed because of pain and the nature of the surgical procedure, quadriceps buckling in standing (inability to stand as a result of inadequate quadriceps strength) was viewed as an indication of profound weakness. The following standardized rehabilitation protocol was applied to both groups. The continuous passive motion (CPM) machine was initiated (0°–60°) in the recovery room, and was increased as tolerated on subsequent days. There was no specific target ROM. ROM was progressed based on patient comfort and ability in a blinded fashion. The CPM was used for 4–6 h daily. Transfer training and weight bearing as tolerated gait training were initiated on the first POD. Patients were advanced from a rollator walker to a standard cane according to their functional ability. Active knee flexion, passive extension, and muscle strengthening exercises were also initiated on the first POD. Knee ROM was measured using a universal goniometer in the seated position; hip and knee were aligned with the foot placed on the floor. Patients received PT from a physical therapist or a rehabilitation technician twice daily during the week and daily on weekends. Daily PT sessions (with a physical therapist) included gait and transfer training, ROM exercises, and strengthening exercises. A second daily mobilization practice session under the supervision of a rehabilitation technician repeated the transfers and ambulation done in the previous session by the physical therapist.

Patient demographics and surgical details were recorded. All patients were followed by standard pain service protocols, which included repeated documented assessment by nurses for pain, nausea, pruritus, sedation, and confusion. Patients were evaluated by the attending anesthesiologist on the pain service for satisfactory recovery from anesthesia, including complete resolution of the nerve block. VAS pain scores were determined by nurses every 2 h after surgery. Pain scores that did not coincide with PT sessions were analyzed as pain at rest. VAS scores with activity were recorded during PT sessions. Analgesic requirements, volume of epidural medication, VAS scores, and side effects attributed to analgesics were recorded. Achievement of rehabilitation functional milestones was recorded by physical therapists. Physical therapists were blinded as to study assignment.

The data distributions of all the variables were examined. Means and sd or medians and quartiles were obtained. Data for Table 1 are presented as mean ± sd if data passed a normality test. Student’s t-test was used to evaluate normally distributed data. Data for Table 1 that did not pass a normality test are presented as median (25%–75% percentiles). A Mann-Whitney rank sum test was used to evaluate data that were not normally distributed. Data for Table 2 are presented as median (25%–75% percentiles). Correction was made for multiple comparisons as indicated. The ROM data were analyzed using repeated-measures analysis of variance. The pain data are ordinal and were not normally distributed. A Mann-Whitney U-test was performed between those with a block and those without for each day. The VAS scale is an integer scale; therefore nonparametric statistics were used to evaluate VAS scores.

Back to Top | Article Outline


Forty-one patients received a FNB and epidural analgesia; 39 patients received epidural analgesia alone. Combined spinal-epidural anesthesia was achieved in all patients. Patients in each group were similar in age, ASA physical status, weight, surgical and anesthesia times, and component use (Table 1).

Patients who received a FNB had significantly lower VAS scores with PT on POD 1 (median, 3 versus 4) and 2 (median, 2.5 versus 4) (Fig. 1). On POD 1, VAS scores ≥6 during PT were reported in 11 of 39 (28%) control patients versus 4 of 41 (10%) FNB patients (P = 0.045, Fisher’s exact test). On POD 2, VAS scores ≥6 during PT were reported in 12 of 39 (31%) control patients versus two of 41 (5%) FNB patients (P = 0.0062, Fisher’s exact test). Median pain scores at rest were low in both groups (Fig. 1). The volume of epidural medication delivered to both groups was similar (Fig. 2).

On POD 1, some members of both groups of patients had decreased strength or sensation in the operative leg. Among patients who received FNB, 29% had buckling because of decreased quadriceps strength and 34% had decreased sensation. Among control patients, 3% had buckling because of decreased quadriceps strength and 13% had decreased sensation. Strength and sensation were determined by the physical therapist. If the patient demonstrated quadriceps buckling on standing, ambulation training was deferred until the patient had adequate quadriceps control in standing. The study design did not control for time from block until assessment. All patients had adequate quadriceps strength in standing without buckling that allowed for safe progressive gait training by POD 2. It is not part of the current practice to use Jordan splints or knee immobilizers for gait training.

Patients who had received a FNB had significant improvement in flexion ROM on POD 2 (Table 2). No differences were noted for other PT outcomes including ambulation distance, independent transfer, CPM flexion, CPM progression, use of a walker, and progression to cane use, stair use, or length of stay (Table 2).

The incidence of side effects was not significantly different between groups. During the first three PODs, nausea was reported (one or more times) in 27% of the FNB patients and 28% of the controls. Corresponding numbers for pruritus were 49% (FNB patients) and 39% (controls). One (3%) of the control patients and none of the FNB patients became confused during the first 3 PODs.

Catheters were removed on POD 1 in 4 patients (2 in the FNB group; one for back pain, one for bilateral numbness; 2 in the control group; one for confusion, one for epidural site leakage). The rate of premature discontinuation was 5%. Sixty-one catheters were removed on POD 2 (34 in the FNB group, 27 in the control group). Fifteen catheters were removed on POD 3 (5 in the FNB group, 10 in the control). No patient withdrew from the study.

No patients had residual numbness, dysesthesia, or weakness associated with FNB or combined spinal-epidural blockade.

Back to Top | Article Outline


Single-injection FNB combined with epidural analgesia using a low concentration of local anesthetic (0.06% bupivacaine + 10 μg/mL hydromorphone) improved pain relief with PT and improved one measure of rehabilitation compared with epidural analgesia alone. Epidural analgesia alone provided excellent pain relief at rest. Previous studies have demonstrated that either epidural analgesia (1,2,4) or FNB (5) improved analgesia and rehabilitation after TKA compared with systemic opioids.

The duration of analgesia observed in this study from the FNB (through POD 2) was longer than had been anticipated. A single-injection FNB typically provided analgesia for no longer than 24 h after anterior cruciate ligament reconstruction or TKA (5,13–15). Only 34% of patients in this study had signs of sensory blockade when evaluated on POD 1 (which may have been <24 h postinjection), and no patients had residual FNB on POD 2. In another study, all patients reported weakness of thigh muscles after TKA even if given a nerve block with saline (10).

FNB benefited one measure of PT (active knee flexion ROM on POD 2) and did not slow any aspect of PT. Increased knee flexion allows patients to perform tasks with less compensatory motion and allows patients to emphasize other goals, such as obtaining complete extension and independent mobilization. In addition, the criterion for discontinuing the use of CPM at this institution is active flexion to 90° for 2 consecutive days. Thus early increases in flexion result in earlier discontinuation of this passive modality and shift the emphasis of the rehabilitation program to more active exercises.

This study did not demonstrate reduction of analgesic-related side effects by adding FNB to PCEA. Patients given FNB had less pain, used similar doses of epidural narcotic, and had the same rate of side effects as did control patients. With PCEA, patients self-administer medication boluses until their pain is entirely controlled or they encounter side effects that induce them to stop activating the bolus mechanism. It is possible that patients administered epidural hydromorphone/bupivacaine to themselves until side effects became prominent.

Routine FNB does not require much additional time, as evidenced by the similar anesthesia times in the 2 groups (109 min for control versus 110 min for FNB group). Anesthesia time included time to perform the FNB. The small difference between the two groups indicated that in the context of TKA, little additional time is needed in the OR to perform a FNB.

There are several methodological issues to consider. This study evaluated the addition of a single-injection FNB to postoperative epidural analgesia. We did not study CFNB catheters. All patients received postoperative epidural analgesia. All patients received urinary catheters. We could not assess the role of the epidural analgesia in causing nausea, vomiting, or urinary retention. The epidural catheter was not tested by administration of concentrated local anesthetic, as this was considered likely to delay the patients’ recovery from the anesthetic. In an attempt to maintain blinding, we also did not assess the success of the FNB. We did not standardize the duration of epidural analgesia; instead the duration was determined by clinical assessment of the patients’ need for epidural analgesia.

In conclusion, addition of FNB to epidural analgesia significantly improved analgesia for the first 2 PODs after TKA and improved knee flexion on POD 2. FNB did not affect achievement of other PT functional milestones.

Back to Top | Article Outline


1. Raj PP, Knarr DC, Vigdorth E, et al. Comparison of continuous epidural infusion of a local anesthetic and administration of systemic narcotics in the management of pain after total knee replacement surgery. Anesth Analg 1987;66:401–6.
2. Mahoney OM, Noble PC, Davidson J, Tullos HS. The effect of continuous epidural analgesia on postoperative pain, rehabilitation, and duration of hospitalization in total knee arthroplasty. Clin Orthop 1990;260:30–7.
3. Williams-Russo P, Sharrock NE, Haas SB, et al. Randomized trial of epidural versus general anesthesia: outcomes after primary total knee replacement. Clin Orthop 1996;331:199–208.
4. Pettine KA, Wedel DJ, Cabanela ME, Weeks JL. The use of epidural bupivacaine following total knee arthroplasty. Orthop Rev 1989;18:894–901.
5. Wang H, Boctor B, Verner J. The effect of single-injection femoral nerve block on rehabilitation and length of hospital stay after total knee replacement. Reg Anesth Pain Med 2002;27:139–44.
6. Hirst GC, Lang SA, Dust WN, et al. Femoral nerve block: single injection versus continuous infusion for total knee arthroplasty. Reg Anesth 1996;21:292–7.
7. Chelly JE, Greger J, Gebhard R, et al. Continuous femoral blocks improve recovery and outcome of patients undergoing total knee arthroplasty. J Arthroplasty 2001;16:436–45.
8. Singelyn FJ, Deyaert M, Joris D, et al. Effects of intravenous patient-controlled analgesia with morphine, continuous epidural analgesia, and continuous three-in-one block on postoperative pain and knee rehabilitation after unilateral total knee arthroplasty. Anesth Analg 1998;87:88–92.
9. Capdevila X, Barthelet Y, Biboulet P, et al. Effects of perioperative analgesic technique on the surgical outcome and duration of rehabilitation after major knee surgery. Anesthesiology 1999;91:8–15.
10. Ganapathy S, Wasserman RA, Watson JT, et al. Modified continuous femoral three-in-one block for postoperative pain after total knee arthroplasty. Anesth Analg 1999;89:1197–202.
11. Cuvillon P, Ripart J, Lalourcey L, et al. The continuous femoral nerve block catheter for postoperative analgesia: bacterial colonization, infectious rate and adverse effects. Anesth Analg 2001;93:1045–9.
12. Kettelkamp DB, Johnson RJ, Smidt GL, et al. An electrogoniometric study of knee motion in normal gait. J Bone Joint Surg Am 1970;52:775–90.
13. Mulroy MF, Larkin KL, Batra MS, et al. Femoral nerve block with 0.25% or 0.5% bupivacaine improves postoperative analgesia following outpatient arthroscopic anterior cruciate ligament repair. Reg Anesth Pain Med 2001;26:24–9.
14. Allen HW, Liu SS, Ware PD, et al. Peripheral nerve blocks improve analgesia after total knee replacement surgery. Anesth Analg 1998;87:93–7.
15. Allen JG, Denny NM, Oakman N. Postoperative analgesia following total knee arthroplasty: a study comparing spinal anesthesia and combined sciatic femoral 3-in-1 block. Reg Anesth Pain Med 1998;23:142–6.
© 2005 International Anesthesia Research Society