Combined femoral and sciatic nerve block versus femoral and local infiltration anesthesia for pain control after total knee arthroplasty: a meta-analysis of randomized controlled trials
Li et al. Journal of Orthopaedic Surgery and Research
Combined femoral and sciatic nerve block versus femoral and local infiltration anesthesia for pain control after total knee arthroplasty: a meta-analysis of randomized controlled trials
Jian Li 1
Xinlian Deng 1
Tao Jiang 0
0 Department of Orthopaedics, The Second Affiliated Hospital of Zhejiang Chinese Medical University , 318 Chaowang Road, Hangzhou, Zhejiang 310005 , China
1 Department of Pain, Yidu Central Hospital of Weifang , Qingzhou, Shangdong Province , China
Background: The purpose of this systematic review and meta-analysis of randomized controlled trials (RCTs) was to evaluate the effect of combined femoral and sciatic nerve block (SNB) versus femoral and local infiltration anesthesia (LIA) after total knee arthroplasty (TKA). Methods: The electronic databases PubMed, Embase, Cochrane Library, and Web of Science were searched from their inception to 15 June 2016. Articles comparing combined femoral and SNB versus femoral and LIA for pain control were eligible for this meta-analysis. This systematic review and meta-analysis was performed according to the PRISMA statement criteria. The primary endpoint was the visual analogue scale (VAS) score with rest at 12, 24, and 48 h, which represents the pain control after TKA. Data regarding active knee flexion, length of hospital stay, anesthesia time, and morphine use at 24 and 48 h were also compiled. The complications of postoperative nausea and vomiting (PONV) and fall were also noted to assess the safety of morphine-sparing effects. After testing for publication bias and heterogeneity across studies, the data were aggregated for random-effects modeling when necessary. Results: Seven clinical trials with 615 patients were included in the meta-analysis. The pooled results indicated that SNB was associated with a lower VAS score at 12 h (MD = −6.96; 95% CI −8.36 to −5.56; P < 0.001) and 48 h (MD = −2. 41; 95% CI −3.90 to −0.91; P < 0.001) after TKA. There was no significant difference between the SNB group and the LIA group in terms of the VAS score at 24 h (MD = 0.67; 95% CI −0.31 to 1.66; P = 0.182). The anesthesia time in the LIA group was shorter than in the SNB group, and the difference was statistically significant (MD = 4.31, 95% CI 1.34 to 7.28, P = 0.004). There were no significant differences between the groups in terms of active knee flexion, length of hospital stay, morphine use, PONV, and the occurrence of falls. Conclusions: SNB may provide earlier anesthesia effects than LIA when combined femoral nerve block (FNB); however, there were no differences in morphine use, active knee flexion, and PONV between the groups. The LIA group spent less time under anesthesia, suggesting that LIA may offer a practical and potentially safer alternative to SNB.
Sciatic nerve block; Local infiltration anesthesia; Total knee replacement; Meta-analysis
Total knee arthroplasty (TKA) is a common procedure for
improving mobility and quality of life in patients with
osteoarthritis or rheumatoid arthritis. However, TKA itself
is always associated with moderate to severe pain after
surgery. It is reported that 60 and 30% of TKA patients
experience severe and moderate pain, respectively .
Currently, both femoral nerve block (FNB) and local
infiltration anesthesia (LIA) can provide effective analgesia,
facilitate early mobilization, and reduce the length of
hospital stay [2, 3]. Chan et al.  conducted a meta-analysis
to compare FNB with other analgesic techniques, and the
results indicated that there was insufficient data to draw a
definitive conclusion regarding FNB with LIA. Although
FNB is a well-accepted and commonly used technique for
regional anesthesia after TKA, previous studies indicate
that some patients experience significant postoperative
pain despite the administration of FNB [5, 6]. Compared
with peripheral nerve block, LIA is an alternative,
convenient anesthetic technique that is usually performed by
orthopedic surgeons. Meanwhile, the efficacy and safety of
LIA is comparable to that of epidural anesthesia, FNB, and
intrathecal morphine . Therefore, anesthesia via sciatic
nerve block (SNB) and LIA are two major options for
supplementing FNB to relieve pain after TKA [8–10].
However, there is no consensus regarding which anesthesia
method is preferable to relieve pain as an adjunct to FNB.
Thus, a meta-analysis of randomized controlled trials
(RCTs) was conducted to compare the efficacy and safety
of pain control with SNB versus LIA when combined with
FNB after TKA.
The electronic databases PubMed, Embase, Cochrane
Library, and Web of Science were searched from their
inception to 15 June 2016. The search terms included
local infiltration anesthesia, femoral nerve block, sciatic
nerve block, and total knee arthroplasty. The Boolean
operators “AND” and “OR” were used to couple these
terms. The details of the search strategy are displayed in
Additional file 1. There were no restrictions regarding
language and publication date. We also manually
retrieved reference lists from the identified studies and
relevant review studies to identify additional relevant
studies. Two investigators independently assessed the
titles and abstracts of the studies identified by the
retrieval. The full text of the remaining studies was then
reviewed to ensure that they met the eligibility criteria.
Disagreements were settled by consulting a third
Fig. 1 The flow diagram of the included studies. RCT randomized clinical trial
General anesthesia Spinal anesthesia
Table 1 The general characteristic of the included studies
No of patients Gender Age
(SNB:LIA) male/female (mean, SNB:LIA)
Anesthesia Surgery method FNB The drug and
doses of SNB
The drug and
doses of LIA
NS not stated, SNB sciatic nerve block, LIA local infiltration anesthesia, FNB femoral nerve block, PCA patient controlled anesthesia
20 ml of 0.75% ropivacaine,
physiological saline 20 mL,
adrenaline 0.3 mg, morphine
hydrochloride and dexamethasone
Ropivacaine Epinephrine Ketorolac Morphine sulfate Concomitant pain management
PCA for 48 h
0.2% ropivacaine infusion
60 mg oral loxoprofen
sodium every 8 h
Continuous femoral block
of 0.2% ropivacaine
20 mL of 0.375% 20 mL of 0.375% 20 mL of 0.375% levobupivacaine
20 mL of 0.5%
20 mL of 0.5%
50 mL of 0.2% ropivacaine
PCA for 48 h
Standard mid-vastus 20 ml of 0.375%
20 ml of 0.375%
200 mg of ropivacaine and
0.5 ml of adrenaline
60 ml 0.5% ropivacaine with
0.3 mg epinephrine
Midvastus approach 20 mL of 0.375% 20 mL of 0.375% 100 mL of 0.2% ropivacaine by
ropivacaine ropivacaine adding 0.5 mL of adrenaline
20 mL of 0.75%
10 mL of 0.75%
Table 2 The quality assessment of the included studies
Author and year
Random sequence generation
Mahadevan, D 2012 
Safa, B 2014 
Tanikawa, H 2014 
Gi, E 2014 
Nagafuchi, M 2015 
Randomization web site
Uesugi, K 2014 
Spangehl, M. J 2015  Computerized random
Opaque sealed envelopes Yes
Inclusion criteria and study selection
All RCTs comparing combined femoral and SNB
versus femoral and LIA for pain control were eligible for
this meta-analysis. If there were more than one
eligible trials from one team, the study with most recent
publication data were enrolled for analysis. Studies
that included bilateral TKA, revision of TKA, or
other anesthetic methods were excluded. All
nonrandomized trials were also excluded.
Two reviewers extracted the data independently using a
predefined data extraction form. Disagreements were
resolved through discussion or consensus with a third
reviewer. The data extracted included the first author,
publication year, study characteristics (number of patients and
percent of female patients), participant characteristics (i.e.,
mean age, type of anesthesia, operative approach, and type
of prosthesis), and the length of follow-up. For studies
Fig. 2 The forest plot comparing SNB and LIA for VAS score after TKA. An inverse variance fixed-effects model was used. Mean differences with
95% CIs are reported. WMD weighted mean difference
Fig. 3 Funnel plot with pseudo 95% confidence limits
with insufficient information, the reviewers tried to
contact the first author via e-mail or telephone to obtain the
original data. After duplicates were excluded, two
reviewers independently read the titles and abstracts of the
selected literature. Most of the articles were excluded
based on the topic of the article provided in the title or
abstract, and disagreements regarding whether an article
should be included were resolved via discussion or
consultation with a senior reviewer. Postoperative pain
intensity was measured using a 100-point visual analogue scale
(VAS). The 10-point VAS score was converted to a
100point VAS score. Data in other forms (i.e., median,
interquartile range, and mean ± 95% confidence interval
(CI)) were converted to mean ± SD as described in the
Cochrane handbook . If the data were not reported
numerically, we extracted them from the published figures
using the “GetData Graph Digitizer” software .
Two independent reviewers assessed the methodological
quality of the included trials according to the Cochrane
Collaboration recommendations . The following
information was evaluated: random sequence generation,
allocation concealment, blinding of outcome assessments,
Fig. 4 The publication bias between the studies, indicated by the funnel plot. WMD weighted mean difference
SFNB single femoral nerve block, CFNB continuous femoral nerve block, VAS visual analogue scale, MD mean difference
were expressed as the relative risk (RR) with 95% CIs.
Statistical significance was set at P < 0.05 to summarize
the findings across the trials. Stata 12.0 software (Stata
Corp., College Station, TX) was used for the
metaanalysis. Statistical heterogeneity was tested using the I2
statistic. A value of I2 > 50% was considered to indicate
statistical heterogeneity, and a random effects model
was applied. Then, sensitivity analysis was conducted to
identify potential sources of heterogeneity. When there
was no statistical evidence of heterogeneity, a
fixedeffects model was adopted. A subgroup analysis was
Heterogeneity P value (I2)
Fig. 5 The result of Begg’s test for the VAS score after TKA
incomplete outcome data, selective reporting, and other
biases. An independent arbiter was consulted to reconcile
Continuous outcomes, such as the VAS at 12, 24, and
48 h, morphine consumption at 24 and 48 h, active knee
flexion, the length of hospital stay, and anesthesia time,
were expressed as the mean difference (MD) with the
respective 95% CIs. Discontinuous outcomes (the rate of
postoperative nausea and vomiting (PONV) and fall)
Table 3 Subgroup analysis for VAS with rest at 12, 24, and 48 h
conducted to identify whether the type of FNB
(continuous FNB versus single-shot FNB) and anesthesia (general
anesthesia versus spinal anesthesia) affected the VAS at
12, 24, and 48 h.
Trial sequential analysis
Because cumulative meta-analyses carry a risk of
producing random errors, mainly because of sparse data
and repetitive testing of cumulative data [13–15], trial
sequential analysis was performed in case the data
were too sparse to draw firm conclusions. Trial
sequential analysis is comparable to interim analysis in
a single trial, and the trial sequential monitoring
boundary can be applied to meta-analysis to
determine whether the P value is small enough to show
the anticipated effect and whether the trial should be
terminated early . If the trial sequential analysis
boundary or the futility zone is crossed, more trials
Search results and quality assessment
In the initial search, 125 potentially relevant trials were
identified, and 32 duplications were removed using the
Endnote X7 software (Thomson Research Soft Company,
America). The titles and abstracts of 93 articles were
reviewed to determine whether they met the inclusion
criteria; the search details are shown in Fig. 1. During the
search process, a total of three disagreements occurred
and were resolved by reading the full-length of article and
conferring with the senior author. Finally, only seven
RCTs involving 615 patients were included for
metaanalysis [9, 17–22]. The number of patients in the SNB
group and the LIA group were 307 and 308, respectively.
The baseline data of the included studies were
comparable, and all the patients in the included studies were
prepared for primary unilateral TKA (Additional file 2). Four
studies [17, 19, 21, 23] performed continuous FNB, and
the remaining three studies [18, 20, 22] performed
singleshot FNB. Two studies administered intraoperative
Fig. 6 The trial sequence analysis of the visual analogue scale (VAS) scores with rest at 12 h, showing that the accumulative Z-curve crossed the
trial sequential monitoring boundary for harm and surpassed the required information size
Fig. 7 The forest plot comparing sciatic nerve block (SNB) and local infiltration anesthesia (LIA) for active knee flexion after TKA. An inverse
variance fixed-effects model was used. Mean differences with 95% CIs are reported
Fig. 8 The forest plot comparing sciatic nerve block (SNB) and placebo in terms of morphine consumption at 24 and 48 h after TKA. An inverse
variance fixed-effects model was used. Mean differences with 95% CIs are reported. WMD weighted mean difference
anesthesia in spinal anesthesia and the rest studies in
general anesthesia. Six studies perform the FNB with nerve
stimulator [9, 17, 18, 20–22] and four studies with
ultrasound devices [18–20, 22]. All studies administered with
oral NSAIDs for postoperative anesthesia. Two studies
administered acetaminophen [9, 22], two studies
administered with diclofenac [18, 20], two studies with loxoprofen
sodium [19, 21], and the rest one study  administered
with celecoxib. The general characteristic can be seen in
The quality assessment was as follows: three studies’
random sequence generation methods were not clear,
and allocation concealment was performed in all the
included studies. The other biases are shown in
VAS score with rest at 12, 24, and 48 h
Five trials with 409 patients reported the VAS score with
rest at 12 h for the SNB group and the LIA group. The
pooled results indicated that SNB was associated with a
lower VAS score at 12 h after TKA (MD = −6.96; 95% CI
−8.36 to −5.56; P < 0.001, Fig. 2) and high heterogeneity
(P = 0.000, I2 = 82.5%). Seven studies with 615 patients
reported the VAS scores with rest at 24 h in the SNB
group and the LIA group; the meta-analysis results
indicated that there was no significant difference between
the SNB group and the LIA group in terms of the VAS
score at 24 h (MD = 0.67; 95% CI −0.31 to 1.66; P = 0.182,
Fig. 2). Four studies with 481 patients were included in a
meta-analysis that indicated that SNB can decrease
VAS with rest at 48 h by a mean of 2.41 mm; this
result was statistically significant (MD = −2.41; 95% CI
−3.90 to −0.91; P < 0.001, Fig. 2) and had low
heterogeneity (P = 0.210, I2 = 33.7%).
A funnel plot was then used to identify whether the
studies showed publication bias, and the results
indicated that there was no publication bias in the included
studies (Fig. 3). Begg’s test further confirmed the results
(P > 0.05, Fig. 4). Because there was a large degree of
heterogeneity among the studies, a sensitivity analysis
was performed to find the source of the heterogeneity.
The results indicated that the study by Safa B may affect
the final result (Fig. 5). We then excluded the data from
the study by Safa B, and results are shown in Additional
file 3. The subgroup analysis results are shown in Table 3.
The SNB group had a lower VAS score with rest at 12 h
when general anesthesia was administered compared
with the LIA group. Furthermore, continuous FNB was
associated with a lower VAS score with rest at 12 h than
Fig. 9 Forest plot comparing the length of hospital stay between the SNB group and the LIA group. WMD weighted mean difference
Trial sequential analysis (TSA)
The trial sequential analysis results indicated that there
was no need to perform additional studies to further
identify the effects of SNB versus LIA for pain control
after TKA. The cumulative Z-score crossed the Z-cure
but did not reach the TSA value (Fig. 6).
Active knee flexion
A total of five studies with 245 patients and three studies
with 321 patients reported active knee flexion in the
SNB group and the LIA group on day 3 and month 3,
respectively. The pooled results indicate that there was
no significant difference between SNB and LIA in terms
of active knee flexion on day 3 (MD = 0.00; 95% CI
−4.15 to 4.16; P = 0.999, Fig. 7) or month 3 (MD = −1.40;
95% CI −3.56 to 1.49; P = 0.421, Fig. 7).
Morphine consumption at 24 and 48 h
Four studies with 323 patients and two studies with 212
patients reported the morphine consumption at 24 and
48 h, respectively. The pooled results indicated that
there was no significant difference between morphine
consumption at 24 h (MD = −0.36; 95% CI −2.23 to 1.52;
P = 0.708, Fig. 8) and 48 h (MD = 0.94; 95% CI −6.93 to
8.80; P = 0.816, Fig. 8) between the SNB group and the
Length of hospital stay
Six studies with 405 patients reported the length of
hospital stay for the two groups. The result indicated that
there was no significant difference between the SNB and
LIA groups in terms of the length of hospital stay (MD
= 0.21; 95% CI −0.05 to 0.47; P = 0.115, Fig. 9), and
heterogeneity was moderate (P = 0.006, I2 = 69.3%).
Six trials with 498 patients reported the occurrence of
PONV for the two groups. The meta-analysis indicated
that there was no statistically significant difference
between the SNB and LIA groups (RR = 0.89, 95% CI 0.61
to 1.32, P = 0.575, Fig. 10).
Four trials with 304 patients reported the occurrence
of falls for the two groups. The meta-analysis
indicated that there was no statistically significant
difference between the SNB and LIA groups (RR = 1.99,
95% CI 0.51 to 7.71, P = 0.320, Fig. 11) in terms of
the occurrence of falls.
Fig. 10 Forest plot comparing PONV between the SNB group and the LIA group. RR risk ratio
Fig. 11 Forest plot comparing the occurrence of fall between the two groups. Mean differences with 95% CIs are reported. RR risk ratio
Three studies with 110 patients reported the anesthesia
time for the LIA group and the SNB group. The pooled
results indicated that the anesthesia time of the LIA
group was shorter than that of the SNB group, and the
difference was statistically significant (MD = 4.31, 95%
CI 1.34 to 7.28, P = 0.004, Fig. 12).
This is the first systematic review and meta-analysis to
compare combined femoral and SNB versus combined
femoral with LIA for pain control after TKA. On the
basis of the pooled estimates, combined femoral and
SNB, compared with combined femoral with LIA, was
associated with a reduction in pain scores, equivalent on
an 110-point VAS to 6.96 point (95% CI, −8.36 to −5.56
point) at 12 h and 2.41 point (95% CI, −3.90 to −0.91) at
48 h. However, this reduction is not of clinical
importance. There was no significant difference between active
knee flexion, morphine consumption at 24 and 48 h,
length of hospital stay, and the occurrence of PONV.
And combined femoral and LIA, compared with
combined femoral with SNB, are associated with less
The only positive result is that SNB can decrease the
VAS score at 12 and 48 h compared with LIA, and the
difference is statistically significant. These outcomes
concur with the morphine consumption at 24 and 48 h.
Peripheral nerve block was induced before surgery and can
decrease the morphine consumption during the early
period. Morphine consumption was also used as a marker
to test the efficacy of adjunctive analgesia [25–27].
Abdallah et al.  conducted a systematic review, and the
results indicated that there was insufficient evidence to
support the effect of adding SNB to FNB for anesthesia
following TKA. Meanwhile, there was no statistically
significant difference between the morphine-related
complications of PONV.
Regarding active knee flexion, there was no significant
difference between the two groups on day 3 and month
3 after TKA. The present results did not find any
significant difference in the progress of rehabilitation, knee
mobilization, and length of hospital stay. These results
also indicated that improved early anesthesia cannot
facilitate early rehabilitation. The results are consistent
with past reports that concluded that SNB or LIA had
no benefit in terms of knee function or length of
hospital stay [23, 28, 29].
It has been reported that the rate of peripheral nerve
injury is 2.9/10,000 for FNB and 2.4/10,000 for SNB, and
the incidence of permanent nerve damage is 1.5/10,000
. Sciatic nerve injury is also a generally known
Fig. 12 Forest plot comparing the anesthesia times of the two groups. Mean difference with 95% CIs are reported. WMD weighted
complication after TKA, with an incidence of 1.3 to
2.2% [31, 32]. Thus, LIA is a relatively safe anesthetic
technique for pain control after TKA. The occurrence of
falls did not differ significantly between the two groups
after TKA. Spanghel et al.  found that four patients
in the SNB group and no patients in the LIA group fell
after TKA, and one patient suffered from a lumbar
vertebral fracture after a fall. Tanikawa et al.  reported
that one patient in the LIA group and one in the SNB
group fell after TKA. The patients in each group were
not permitted to ambulate without assistance from
nurses or physiotherapists to prevent falls. Furthermore,
the duration of the motor block of toe motion in the
LIA group was less than that in the SNB group, and the
difference was statistically significant. Nagafuchi et al.
 found that one patient in the LIA group fell after
surgery. Although the results of this meta-analysis
indicated that there was no significant difference between
SNB and LIA in terms of fall, a larger sample may have
shown a trend toward more falls among patients who
underwent FNB combined with SNB.
There were several limitations to this meta-analysis:
(1) only seven RCTs were included, and the sample
sizes in each trial were not large, which could affect
the final results; (2) the duration of follow-up in some
studies was unclear, and long-term follow-up was
necessary for this analysis; (3) the publication bias
that existed in the meta-analysis also influenced the
results; (4) variations in the methods for obtaining,
preparing, and applying the perioperative anesthetic
protocol currently constitute a limitation to
performing any comparisons between studies; (5) there is
gross variability in SNB and FNB techniques
(singleshot or continuous), location and needle approach
and in the concentration, and volume and frequency
of local anesthetic administered; (6) no studies applied
the postoperative recovery scale, and high-quality
studies that include a multi-domain quality of
recovery assessment tool are preferred.
In conclusion, there were no differences in pain
scores with rest at 24 h, morphine consumption,
active knee flexion, the occurrence of fall, and PONV
between the groups. Although some analgesic efficacy
at 12 and 48 h were seen with the use of combined
femoral and SNB, these were unlikely to be of clinical
importance. LIA may offer a practical and potentially
safer alternative to SNB. Further high-quality RCTs
are needed to identify the optimal dose of LIA for
reducing pain after TKA.
Additional file 1: The search strategy of the included studies. (DOCX 14 kb)
Additional file 3: The results of VAS with rest at 12 h, 24 h and 48 h
after excluded the study of Safa B. (TIF 3014 kb)
FNB: Femoral nerve block; LIA: Local infiltration anesthesia; MD: Mean
difference; PONV: Postoperative nausea and vomiting; RCTs: Randomized
controlled trials; RR: Relative risk; SNB: Sciatic nerve block; TKA: Total knee
arthroplasty; TSA: Trial sequential analysis; VAS: Visual analogue scale.
No funding was required for this retrospective study.
Availability of data and materials
We state that the data will not be shared since all the raw data are present
in the figures included in the article.
JL conceived the study design. XLD and JL performed the study, collected
the data, and contributed to the study design. TJ prepared the manuscript.
JL and TJ edited the manuscript. All authors read and approved the final
The authors declare that they have no competing interests.
Consent for publication
1. Singelyn FJ , Deyaert M , Joris D , Pendeville E , Gouverneur JM . 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 .
2. Nakagawa S , Arai Y , Inoue H , et al. Comparative effects of periarticular multimodal drug injection and single-shot femoral nerve block on pain following total knee arthroplasty and factors influencing their effectiveness . Knee Surg Relat Res . 2016 ; 28 : 233 - 8 .
3. Albrecht E , Guyen O , Jacot-Guillarmod A , Kirkham KR . The analgesic efficacy of local infiltration analgesia vs femoral nerve block after total knee arthroplasty: a systematic review and meta-analysis . Br J Anaesth . 2016 ; 116 : 597 - 609 .
4. Chan EY , Fransen M , Parker DA , Assam PN , Chua N. Femoral nerve blocks for acute postoperative pain after knee replacement surgery . Cochrane Database Syst Rev . 2014 ; 13 :Cd009941.
5. Davies AF , Segar EP , Murdoch J , Wright DE , Wilson IH . Epidural infusion or combined femoral and sciatic nerve blocks as perioperative analgesia for knee arthroplasty . Br J Anaesth . 2004 ; 93 : 368 - 74 .
6. Tierney E , Lewis G , Hurtig JB , Johnson D. Femoral nerve block with bupivacaine 0.25 per cent for postoperative analgesia after open knee surgery . Can J Anaesth . 1987 ; 34 : 455 - 8 .
7. Andersen LO , Kehlet H. Analgesic efficacy of local infiltration analgesia in hip and knee arthroplasty: a systematic review . Br J Anaesth . 2014 ; 113 : 360 - 74 .
8. Andersen LO , Husted H , Otte KS , Kristensen BB , Kehlet H. A compression bandage improves local infiltration analgesia in total knee arthroplasty . Acta Orthop . 2008 ; 79 : 806 - 11 .
9. Mahadevan D , Walter RP , Minto G , Gale TC , McAllen CJ , Oldman M. Combined femoral and sciatic nerve block vs combined femoral and periarticular infiltration in total knee arthroplasty. a randomized controlled trial . J Arthroplasty . 2012 ; 27 : 1806 - 11 .
10. McNamee DA , Parks L , Milligan KR . Post-operative analgesia following total knee replacement: an evaluation of the addition of an obturator nerve block to combined femoral and sciatic nerve block . Acta Anaesthesiol Scand . 2002 ; 46 : 95 - 9 .
11. Higgins JPT GS . Cochrane handbook for systematic reviews of interventions version 5 .1.0. 2011 ; [http://handbook.cochrane.org/].
12. Abdallah FW , Brull R. Is sciatic nerve block advantageous when combined with femoral nerve block for postoperative analgesia following total knee arthroplasty? A systematic review . Reg Anesth Pain Med . 2011 ; 36 : 493 - 8 .
13. Wetterslev J , Thorlund K , Brok J , Gluud C. Trial sequential analysis may establish when firm evidence is reached in cumulative meta-analysis . J Clin Epidemiol . 2008 ; 61 : 64 - 75 .
14. Brok J , Thorlund K , Wetterslev J , Gluud C. Apparently conclusive meta-analyses may be inconclusive-trial sequential analysis adjustment of random error risk due to repetitive testing of accumulating data in apparently conclusive neonatal meta-analyses . Int J Epidemiol . 2009 ; 38 : 287 - 98 .
15. Higgins J , Whitehead A , Simmonds M. Sequential methods for randomeffects meta‐analysis . Stat Med . 2011 ; 30 : 903 - 21 .
16. Thorlund K , Imberger G , Walsh M , et al. The number of patients and events required to limit the risk of overestimation of intervention effects in metaanalysis-a simulation study . PLoS One . 2011 ; 6 : e25491 .
17. Safa B , Gollish J , Haslam L , McCartney CJL . Comparing the effects of single shot sciatic nerve block versus posterior capsule local anesthetic infiltration on analgesia and functional outcome after total knee arthroplasty a prospective, randomized, double-blinded, controlled trial . J Arthroplasty . 2014 ; 29 : 1149 - 53 .
18. Tanikawa H , Sato T , Nagafuchi M , Takeda K , Oshida J , Okuma K. Comparison of local infiltration of analgesia and sciatic nerve block in addition to femoral nerve block for total knee arthroplasty . J Arthroplasty . 2014 ; 29 : 2462 - 7 .
19. Gi E , Yamauchi M , Yamakage M , et al. Effects of local infiltration analgesia for posterior knee pain after total knee arthroplasty: comparison with sciatic nerve block . J Anesth . 2014 ; 28 : 696 - 701 .
20. Nagafuchi M , Sato T , Sakuma T , et al. Femoral nerve block-sciatic nerve block vs. femoral nerve block-local infiltration analgesia for total knee arthroplasty: a randomized controlled trial . BMC Anesthesiol . 2015 ; 15 : 182 .
21. Uesugi K , Kitano N , Kikuchi T , Sekiguchi M , Konno S. Comparison of peripheral nerve block with periarticular injection analgesia after total knee arthroplasty: a randomized, controlled study . Knee . 2014 ; 21 : 848 - 52 .
22. Spangehl MJ , Clarke HD , Hentz JG , Misra L , Blocher JL , Seamans DP . The Chitranjan Ranawat Award: periarticular injections and femoral & sciatic blocks provide similar pain relief after TKA: a randomized clinical trial . Clin Orthop Relat Res . 2015 ; 473 : 45 - 53 .
23. Mahadevan D , Walter R , Gale T , et al. Sciatic nerve blockade versus posterior capsular local anaesthetic infiltration in total knee arthroplasty: a randomised controlled trial . Reg Anesth Pain Med . 2010 ; 35 :E51.
24. Safa B , Haslam L , Gollish J , McCartney C. A prospective, randomized trial, comparing analgesic efficacy and postoperative functional recovery of either single shot sciatic nerve block or posterior capsule infiltration combined with femoral block for total knee arthroplasty . Regional Anesthesia and Pain Medicine . 2011 ; 36 : 508 - 20 .
25. Edwards ND , Wright EM . Continuous low-dose 3-in-1 nerve blockade for postoperative pain relief after total knee replacement . Anesth Analg . 1992 ; 75 : 265 - 7 .
26. Dahl JB , Daugaard JJ , Rasmussen B , Egebo K , Carlsson P , Kehlet H. Immediate and prolonged effects of pre- versus postoperative epidural analgesia with bupivacaine and morphine on pain at rest and during mobilisation after total knee arthroplasty . Acta Anaesthesiol Scand . 1994 ; 38 : 557 - 61 .
27. Serpell MG , Millar FA , Thomson MF . Comparison of lumbar plexus block versus conventional opioid analgesia after total knee replacement . Anaesthesia . 1991 ; 46 : 275 - 7 .
28. Carli F , Clemente A , Asenjo JF , et al. Analgesia and functional outcome after total knee arthroplasty: periarticular infiltration vs continuous femoral nerve block . Br J Anaesth . 2010 ; 105 : 185 - 95 .
29. Toftdahl K , Nikolajsen L , Haraldsted V , Madsen F , Tonnesen EK , Soballe K. Comparison of peri- and intraarticular analgesia with femoral nerve block after total knee arthroplasty: a randomized clinical trial . Acta Orthop . 2007 ; 78 : 172 - 9 .
30. Auroy Y , Benhamou D , Bargues L , et al. Major complications of regional anesthesia in France: The SOS Regional Anesthesia Hotline Service . Anesthesiology. 2002 ; 97 : 1274 - 80 .
31. Schinsky MF , Macaulay W , Parks ML , Kiernan H , Nercessian OA. Nerve injury after primary total knee arthroplasty . J Arthroplasty . 2001 ; 16 : 1048 - 54 .
32. Horlocker TT , Cabanela ME , Wedel DJ . Does postoperative epidural analgesia increase the risk of peroneal nerve palsy after total knee arthroplasty? Anesth Analg . 1994 ; 79 : 495 - 500 .