Peripheral nerve blocks for postoperative pain after major knee surgery.

Peripheral nerve blocks for postoperative pain after major knee surgery (Review) Xu J, Chen XM, Ma CK, Wang XR

This is a reprint of a Cochrane review, prepared and maintained by The Cochrane Collaboration and published in The Cochrane Library 2014, Issue 12 http://www.thecochranelibrary.com

Peripheral nerve blocks for postoperative pain after major knee surgery (Review) Copyright © 2014 The Cochrane Collaboration. Published by John Wiley & Sons, Ltd.

TABLE OF CONTENTS HEADER . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ABSTRACT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . PLAIN LANGUAGE SUMMARY . . . . . . . . . . . . . . . . . . . . . BACKGROUND . . . . . . . . . . . . . . . . . . . . . . . . . . . OBJECTIVES . . . . . . . . . . . . . . . . . . . . . . . . . . . . METHODS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Figure 1. . . . . . . . . . . . . . . . . . . . . . . . . . . . . RESULTS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Figure 2. . . . . . . . . . . . . . . . . . . . . . . . . . . . . Figure 3. . . . . . . . . . . . . . . . . . . . . . . . . . . . . DISCUSSION . . . . . . . . . . . . . . . . . . . . . . . . . . . . AUTHORS’ CONCLUSIONS . . . . . . . . . . . . . . . . . . . . . . ACKNOWLEDGEMENTS . . . . . . . . . . . . . . . . . . . . . . . REFERENCES . . . . . . . . . . . . . . . . . . . . . . . . . . . . CHARACTERISTICS OF STUDIES . . . . . . . . . . . . . . . . . . . . DATA AND ANALYSES . . . . . . . . . . . . . . . . . . . . . . . . . Analysis 1.1. Comparison 1 Peripheral nerve blocks adjunctive to systemic analgesia versus Outcome 1 Pain intensity at rest (0 - 23 hours postoperatively). . . . . . . . Analysis 1.2. Comparison 1 Peripheral nerve blocks adjunctive to systemic analgesia versus Outcome 2 Pain intensity at rest (24 - 47 hours postoperatively). . . . . . . . Analysis 1.3. Comparison 1 Peripheral nerve blocks adjunctive to systemic analgesia versus Outcome 3 Pain intensity at rest (48 - 72 hours postoperatively). . . . . . . . Analysis 1.4. Comparison 1 Peripheral nerve blocks adjunctive to systemic analgesia versus Outcome 4 Pain intensity on movement (0 - 23 hours postoperatively). . . . . . Analysis 1.5. Comparison 1 Peripheral nerve blocks adjunctive to systemic analgesia versus Outcome 5 Pain intensity on movement (24 - 47 hours postoperatively). . . . . Analysis 1.6. Comparison 1 Peripheral nerve blocks adjunctive to systemic analgesia versus Outcome 6 Pain intensity on movement (48 - 72 hours postoperatively). . . . . APPENDICES . . . . . . . . . . . . . . . . . . . . . . . . . . . . CONTRIBUTIONS OF AUTHORS . . . . . . . . . . . . . . . . . . . . DECLARATIONS OF INTEREST . . . . . . . . . . . . . . . . . . . . . DIFFERENCES BETWEEN PROTOCOL AND REVIEW . . . . . . . . . . . .


. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . systemic analgesia alone, . . . . . . . . . systemic analgesia alone, . . . . . . . . . systemic analgesia alone, . . . . . . . . . systemic analgesia alone, . . . . . . . . . systemic analgesia alone, . . . . . . . . . systemic analgesia alone, . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

1 1 2 3 4 4 6 9 10 11 15 16 17 17 20 51 52 53 54 55 56 57 57 59 59 59

i

[Intervention Review]

Peripheral nerve blocks for postoperative pain after major knee surgery Jin Xu1 , Xue-mei Chen1 , Chen-kai Ma2 , Xiang-rui Wang1 1 Department of Anesthesiology, Ren Ji Hospital, School of Medicine, Shanghai Jiao Tong University, Shanghai, China. 2 Department of Surgery, Royal Melbourne Hospital, The University of Melbourne, Melbourne, Australia

Contact address: Xiang-rui Wang, Department of Anesthesiology, Ren Ji Hospital, School of Medicine, Shanghai Jiao Tong University, 1630 Dongfang Road, Shanghai, 200127, China. [email protected]. Editorial group: Cochrane Pain, Palliative and Supportive Care Group. Publication status and date: New, published in Issue 12, 2014. Review content assessed as up-to-date: 10 February 2014. Citation: Xu J, Chen XM, Ma CK, Wang XR. Peripheral nerve blocks for postoperative pain after major knee surgery. Cochrane Database of Systematic Reviews 2014, Issue 12. Art. No.: CD010937. DOI: 10.1002/14651858.CD010937.pub2. Copyright © 2014 The Cochrane Collaboration. Published by John Wiley & Sons, Ltd.

ABSTRACT Background Major knee surgery is a common operative procedure to help people with end-stage knee disease or trauma to regain mobility and have improved quality of life. Poorly controlled pain immediately after surgery is still a key issue for this procedure. Peripheral nerve blocks are localized and site-specific analgesic options for major knee surgery. The increasing use of peripheral nerve blocks following major knee surgery requires the synthesis of evidence to evaluate its effectiveness and safety, when compared with systemic, local infiltration, epidural and spinal analgesia. Objectives To examine the efficacy and safety of peripheral nerve blocks for postoperative pain control following major knee surgery using methods that permit comparison with systemic, local infiltration, epidural and spinal analgesia. Search methods We searched the Cochrane Central Register of Controlled Trials (CENTRAL) (Issue 1, 2014), MEDLINE and EMBASE, from their inception to February 2014. We identified ongoing studies by searching trial registries, including the metaRegister of controlled trials (mRCT), clinicaltrials.gov and the WHO International Clinical Trials Registry Platform (ICTRP). Selection criteria We included participant-blind, randomized controlled trials of adult participants (15 years or older) undergoing major knee surgery, in which peripheral nerve blocks were compared to systemic, local infiltration, epidural and spinal analgesia for postoperative pain relief. Data collection and analysis Two review authors independently assessed study eligibility and extracted data. We recorded information on participants, methods, interventions, outcomes (pain intensity, additional analgesic consumption, adverse events, knee range of motion, length of hospital stay, hospital costs, and participant satisfaction). We used the 5-point Oxford quality and validity scale to assess methodological quality, as well as criteria outlined in the Cochrane Handbook for Systematic Reviews of Interventions. We conducted meta-analysis of two or more studies with sufficient data to investigate the same outcome. We used the I² statistic to explore the heterogeneity. If there was no significant heterogeneity (I² value 0% to 40%), we used a fixed-effect model for meta-analysis, but otherwise we used a randomPeripheral nerve blocks for postoperative pain after major knee surgery (Review) Copyright © 2014 The Cochrane Collaboration. Published by John Wiley & Sons, Ltd.

1

effects model. For dichotomous data, we present results as a summary risk ratio (RR) and a 95% confidence interval (95% CI). Where possible, we calculated the number needed to treat for an additional beneficial outcome (NNTB) or for an additional harmful outcome (NNTH), together with 95% CIs. For continuous data, we used the mean difference (MD) and 95% CI for similar outcome measures. We describe the findings of individual studies where pooling of data was not possible. Main results According to the eligibility criteria, we include 23 studies with 1571 participants, with high methodological quality overall. The studies compared peripheral nerve blocks adjunctive to systemic analgesia with systemic analgesia alone (19 studies), peripheral nerve blocks with local infiltration (three studies), and peripheral nerve blocks with epidural analgesia (one study). No study compared peripheral nerve blocks with spinal analgesia.Compared with systemic analgesia alone, peripheral nerve blocks adjunctive to systemic analgesia resulted in a significantly lower pain intensity score at rest, using a 100 mm visual analogue scale, at all time periods within 72 hours postoperatively, including the zero to 23 hours interval (MD -11.85, 95% CI -20.45 to -3.25, seven studies, 390 participants), the 24 to 47 hours interval (MD -12.92, 95% CI -19.82 to -6.02, six studies, 320 participants) and the 48 to 72 hours interval (MD -9.72, 95% CI -16.75 to -2.70, four studies, 210 participants). Subgroup analyses suggested that the high levels of statistical variation in our analyses could be explained by larger effects in people undergoing total knee arthroplasty compared with other types of surgery. Pain intensity was also significantly reduced on movement in the 48 to 72 hours interval postoperatively (MD -6.19, 95% CI -11.76 to 0.62, two studies, 112 participants). There was no significant difference on movement between these two groups in the time period of zero to 23 hours (MD -6.95, 95% CI -15.92 to 2.01, five studies, 304 participants) and 24 to 47 hours (MD -8.87, 95% CI -27.77 to 10.03, three studies, 182 participants). The included studies reported diverse types of adverse events, and we did not conduct a meta-analysis on specific types of adverse event. The numbers of studies and participants were also too few to draw conclusions on the other prespecified outcomes of: additional analgesic consumption; median time to remedication; knee range of motion; median time to ambulation; length of hospital stay; hospital costs; and participant satisfaction. There were insufficient data to compare peripheral nerve blocks and local infiltration or between peripheral nerve blocks and epidural analgesia. Authors’ conclusions All of the included studies reported the main outcome of pain intensity but did not cover all the secondary outcomes of interest. The current review provides evidence that the use of peripheral nerve blocks as adjunctive techniques to systemic analgesia reduced pain intensity when compared with systemic analgesia alone after major knee surgery. There were too few data to draw conclusions on other outcomes of interest. More trials are needed to demonstrate a significant difference when compared with local infiltration, epidural analgesia and spinal analgesia.

PLAIN LANGUAGE SUMMARY Peripheral nerve blocks for postoperative pain after major knee surgery There are different types of knee surgery which aim to treat joint stiffness or to reconstruct or repair the knee joint. Poorly controlled pain immediately after knee surgery is still a key issue. Pain-relieving medicines (analgesics) that are injected into the nerve are called peripheral nerve blocks. Nerve blocks are often used after major knee surgery, with or without other pain-relieving medicine. The current review included 23 high quality studies of 1571 adult participants (15 years or older) undergoing major knee surgery. The latest search date was February 2014. This review demonstrated that peripheral nerve blocks plus analgesics (taken orally or given by injection) reduced pain when compared with analgesia without the nerve block within the first 72 hours. However, the benefits may be limited to people undergoing total knee arthroplasty. Pain was also significantly reduced on movement in the 48 to 72 hours after surgery, but there was no difference on movement before then. The review did not find enough evidence to draw conclusions on morphine intake, knee range of motion, length of hospital stay, hospital costs, or patient satisfaction. The review did not find enough evidence to draw conclusions on adverse events. More trials are needed to demonstrate a significant difference when comparing nerve blocks with other ways of reducing pain. Peripheral nerve blocks for postoperative pain after major knee surgery (Review) Copyright © 2014 The Cochrane Collaboration. Published by John Wiley & Sons, Ltd.

2

BACKGROUND

Description of the condition Major knee surgery includes operations such as total knee arthroplasty, arthrolysis (Capdevila 1999) and anterior cruciate ligament reconstruction (Fowler 2008). People with end-stage knee disease or trauma can regain mobility and have improved quality of life after undergoing these procedures. Major knee surgery is becoming increasingly common; the annual number of total knee arthroplasties performed in the United States has doubled in the last decade (Weinstein 2013). Despite advances in surgical techniques, poorly controlled pain immediately after surgery is still a key issue associated with this procedure (Chan 2013). The incidence of moderate-to-severe pain after total knee arthroplasty is reported to be about 50%, greater than that reported for all surgeries in a general population or for total hip arthroplasty (Grosu 2014). Thus, an effective analgesic strategy is required. Patients are advised to undergo physical therapy after surgery to restore their ability to carry out daily activities. Physical therapy can include passive and active knee flexion and extension exercises. Postoperative muscle spasm is an obstacle to these exercises, as well as a major source of pain. Efficacious muscle relaxation should therefore be considered so as to optimize pain relief (Sakai 2013).

Description of the intervention There are several analgesic options for major knee surgery, such as systemic analgesia, local infiltration, and epidural and spinal analgesia. Systemic analgesia with opioids may cause various adverse events, including nausea, vomiting, pruritus, sedation and drowsiness (Chan 2013). Local infiltration analgesia has been proven to be most effective when the person is at rest, and may be less effective when they are walking or engaging in continuous passive movement (Yadeau 2013). There are many adverse events associated with epidural and spinal analgesia, such as perioperative hypotension, urinary retention, nausea, vomiting, and sensory and motor blockade of the non-operated leg. To some extent, these adverse events may cause harm to patients and interfere with early ambulation (Mugabure Bujedo 2012; Teng 2012). Moreover, the co-administration of anticoagulant drugs, used to prevent thromboembolic events, increases the potential risk of spinal epidural hematoma (Choi 2003). Peripheral nerve blocks are localized and site-specific. Their history dates back to 1930, when the technique was first reported to relieve the pain in obliterative vascular disease of the lower leg (Smithwick 1930). In 1980, Rosenblatt 1980 reported the first use of peripheral nerve blocks as the sole means of postoperative analgesia after a knee operation. For decades, the accurate performance of peripheral nerve blocks has been supported by peripheral nerve stimulation techniques. In

1989, ultrasonography was first utilized to confirm the location of the needle and observe the spread of local anesthetic while performing peripheral nerve blocks (Ting 1989). Subsequently, ultrasound-guided blocks have become increasingly popular with clinicians, owing to the precise action and faster onset of the block (Sala-Blanch 2012). Bupivacaine and ropivacaine are the most commonly used local anesthetics for peripheral nerve blocks. In addition, drugs such as opioids, epinephrine and clonidine are used as adjuvants to increase the duration of the analgesic effect, but there are no reports of their safety or efficacy (Aguirre 2012). There are several types of peripheral nerve block that aim to relieve lower extremity pain, including femoral nerve, sciatic nerve and lumbar plexus blocks. Among them, the femoral nerve block is considered one of the primary options following major knee surgery. The femoral nerve is the largest branch of the lumbar plexus and provides sensation to the anterior aspect of the knee, whereas the sciatic nerve innervates the posterior aspect. Therefore, one method of ensuring complete reduction of pain transmission in the knee is to perform a sciatic nerve block in combination with a femoral nerve block (Tantry 2012). Another combination of peripheral nerve blocks used to ensure optimum pain relief is the three-in-one technique, which blocks three branches of the lumbar plexus: the femoral nerve (L2 - L4), the obturator nerve (L2 - L4), and the lateral femoral cutaneous nerve (L2 - L3) (Hogan 2009). Peripheral nerve blocks can be administered as a single shot or continuously via a catheter and a pump. Continuous peripheral nerve blocks (CPNBs) were introduced by Ansbro in 1946, later than single-shot peripheral nerve blocks (SSPNBs), to increase the duration of the effect of the brachial plexus block (Ansbro 1946). Since then, continuous peripheral nerve blocks have evolved into reliable analgesic techniques that are widely used in the postoperative period. Continuous peripheral nerve blocks take approximately one-and-a-half to four times longer to perform than single shots, but the duration of the analgesic effect is increased (Chan 2013). Furthermore, the well-designed instrumentation used for a continuous peripheral nerve block provides patients with adequate and continuous analgesia after discharge (Dervin 2012).

How the intervention might work Peripheral nerve blocks following major knee surgery reduce local pain transmission by blocking one or more major nerves supplying the lower limb. Clinical trials have also shown that they may reduce the postoperative inflammatory response (Bagry 2008; Martin 2008). Peripheral nerve blocks offer a number of advantages for postoperative analgesia following major knee surgery. It has been reported that they result in better analgesic control, fewer opioid-related side effects, earlier improvements in knee flexion, and less pain during rehabilitation (Chan 2013; Sakai 2013). Additionally, they


3

avoid motor blockade to the non-operated leg, thereby encouraging early ambulation and relieving psychological stress to some degree. It is worth noting that performing a peripheral nerve block is limited by certain factors such as the operative position of the patient. In addition, safety factors, such as the potential infection risk and the risk of exceeding the safe dose of local anesthetic, may even potentially become life-threatening (Tantry 2012).

insertion is not obligatory but studies should ensure that participants have the same expectations of treatment throughout the entire treatment process. For example, all procedures could be performed behind a drape to block the participant’s view, dressings could be placed on injection sites and blocks could be performed by anesthetists who are not involved with data collection (Allen 1998). Types of participants

Why it is important to do this review Peripheral nerve blocks may be effective for postoperative pain after major knee surgery. Most randomized controlled trials (RCTs) conclude that they relieve postoperative pain and decrease consumption of opioids, although studies have included a relatively small number of participants. A number of meta-analyses have been conducted on related topics in recent years. However, most of them focus on specific types of peripheral nerve block, such as the femoral nerve (Chan 2014; Paul 2010), sciatic nerve (Abdallah 2011) and continuous peripheral nerve block (Richman 2006), and the surgical procedures studied have mainly been limited to total knee arthroplasty. It is important to evaluate the effect of a range of peripheral nerve blocks after major knee surgery.

We include studies of adult participants (15 years or older) who underwent major knee surgery. The procedure could be either total knee arthroplasty, arthrolysis, anterior cruciate ligament reconstruction or any other major surgeries performed on the knee. Types of interventions We include studies that compare the analgesic effect of peripheral nerve block versus systemic, local infiltration, epidural or spinal analgesia following major knee surgery. We include all subtypes of peripheral nerve block. Co-interventions are permitted in the review, but are required to be the same in both the intervention and comparator groups. Types of outcome measures We consider pain intensity measured as greater than 30 mm on a 100 mm visual analogue scale (VAS) to equate to ’at least moderate pain’. To clarify, we consider an intensity of less than or equal to 30 mm to be ’no worse than mild pain’ (Collins 1997; Moore 2013).

OBJECTIVES To examine the efficacy and safety of peripheral nerve blocks for postoperative pain control following major knee surgery using methods that permit comparison with systemic, local infiltration, epidural and spinal analgesia.

METHODS

Primary outcomes

• Pain intensity assessed on a 100 mm VAS on the day of surgery and within the 72 hours following surgery (divided into three time intervals: zero to 23 hours, 24 to 47 hours, 48 to 72 hours) at rest or on movement. We standardized pain intensity data described by other means than a 100 mm VAS to such a scale. Secondary outcomes

Criteria for considering studies for this review

Types of studies We include participant-blind, prospective, randomized controlled clinical trials. We exclude concurrent cohort studies and observational studies, as well as case series and case reports without a control group. In order to maintain participant blinding, placebo catheters normally need to be inserted, which may cause unnecessary discomfort and raise the risk of infection (Chan 2013). Therefore, this

• Proportion of participants with ’no worse than mild pain’ (pain intensity of less than or equal to 30 mm of 100 mm VAS). • Additional analgesic consumption within 72 hours after surgery (divided into three time intervals: zero to 23 hours, 24 to 47 hours, 48 to 72 hours) and median time to remedication. • Adverse events within 72 hours after surgery (divided into three time intervals: zero to 23 hours, 24 to 47 hours, 48 to 72 hours). • Knee range of motion within 72 hours after surgery (divided into three time intervals: zero to 23 hours, 24 to 47 hours, 48 to 72 hours) and median time to ambulation. • Length of hospital stay and hospital costs.


4

• Participant satisfaction.

Search methods for identification of studies

We carried out handsearching of the reference lists of retrieved articles for relevant trials not identified by the electronic searches. We identified any ongoing trials by searching trial registries on 10th February 2014, including the metaRegister of controlled trials (mRCT), clinicaltrials.gov and the WHO International Clinical Trials Registry Platform (ICTRP).

Electronic searches We searched the following electronic databases from the date of inception: • The Cochrane Central Register of Controlled Trials (CENTRAL), on The Cochrane Library Issue 1 of 12, 2014 • MEDLINE & MEDLINE in Process (OVID) 1946 to 10th February 2014 • EMBASE (OVID) 1974 to 10th February 2014 See Appendix 1 for the search strategies. We applied no language restrictions. Searching other resources

Data collection and analysis Selection of studies We initially determined eligibility by reading the titles retrieved from each search. We then screened all remaining articles by reading each abstract. Two review authors (JX, XC) evaluated the studies independently, resolving disagreements by discussion or, if necessary, by recourse to a third review author (XW). We documented the selection process in sufficient detail to complete a PRISMA flowchart (Liberati 2009; Figure 1). We created a table of Characteristics of included studies.


5

Figure 1. Study flow diagram.


6

Data extraction and management We extracted data from the included trials (JX, XC), resolving disagreements through recourse to a third review author (XW). We used a standard data extraction form to tabulate the extracted data. This form included: • participant characteristics (total numbers, age and gender of each person); • type of knee surgery; • type of analgesia and related details; • pain intensity, both at rest and on movement, from the day of surgery to 72 hours after surgery; • additional analgesic requirements (name and amount of analgesic consumed, median time to remedication); • safety (incidence and degree of adverse events); • rehabilitation indices (median time to ambulation and knee range of motion); • length of hospital stay and hospital costs; • participant satisfaction. Assessment of risk of bias in included studies We used the Oxford quality and validity scales to assess the risk of bias in included studies (Jadad 1996). Two review authors (JX, XC) assessed each study independently and resolved any disagreement through discussion, with recourse to a third review author (XW) if necessary. The scales are as follows. • Is the study randomized? If yes add one point. • Is the randomization procedure reported and is it appropriate? If yes add one point, if no deduct one point. • Is the study double-blind? If yes add one point. • Is the double-blind method reported and is it appropriate? If yes add one point, if no deduct one point. • Are the reasons for participant withdrawals and drop-outs described? If yes add one point. We also used the criteria outlined in the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2011) to further assess the risk of bias. The criteria are as follows:

in the process, e.g. odd or even date of birth; date of admission; hospital or clinical record number); unclear risk of bias.

Allocation concealment (checking for possible selection bias)

We assessed the method used in each included study to conceal the allocation sequence. The judgment was as follows: low risk of bias (assignment could not be foreseen, e.g. central allocation; sequentially-numbered drug containers of identical appearance; sequentially-numbered, opaque, sealed envelopes); high risk of bias (assignment could possibly be foreseen, e.g. using an open random allocation schedule; assignment envelopes were used without appropriate safeguards; alternation or rotation); unclear risk of bias.

Blinding of participants and personnel (checking for possible performance bias)

We assessed the method used in each included study to blind study participants and personnel from knowledge of which intervention a participant received. The judgment was as follows: low risk of bias (no blinding or incomplete blinding, but the outcome is not likely to be influenced by lack of blinding; blinding of participants and key study personnel ensured, and unlikely that the blinding could have been broken); high risk of bias (no blinding or incomplete blinding, or blinding of key study participants and personnel attempted, but likely that the blinding could have been broken, and the outcome is likely to be influenced by lack of blinding); unclear risk of bias.

Blinding of outcome assessment (checking for possible detection bias)

We assessed the method used in each included study to blind outcome assessors from knowledge of which intervention a participant received. The judgment was as follows: low risk of bias (no blinding of outcome assessment, but the outcome measurement is not likely to be influenced by lack of blinding; blinding of outcome assessment ensured, and unlikely that the blinding could have been broken); high risk of bias (no blinding of outcome assessment, or blinding of outcome assessment, but likely that the blinding could have been broken, and the outcome measurement is likely to be influenced by lack of blinding); unclear risk of bias.

Sequence generation (checking for possible selection bias)

We assessed the method used in each included study to generate the allocation sequence. The judgment was as follows: low risk of bias (a random component was described in the process, e.g. random number table; computer random number generator; coin tossing); high risk of bias (a non-random component is described

Incomplete outcome data (checking for possible attrition bias)

We assessed the completeness of outcome data in each included study. The judgment was as follows: low risk of bias (e.g. no missing outcome data; reasons for missing outcome data unlikely to be


7

related to true outcome; missing outcome data balanced in numbers across intervention groups, with similar reasons for missing data across groups); high risk of bias (e.g. reasons for missing outcome data likely to be related to true outcome, with either imbalance in numbers or reasons for missing data across intervention groups); unclear risk of bias.

calculated the standard deviation from the formula SD =

√

N

× upper limit - lower limit)/3.92 (Higgins 2011). We report individual study results if pooling of data was not possible.

Unit of analysis issues Selective reporting (checking for possible reporting bias)

We stated how the possibility of selective outcome reporting is examined and what is found. The judgment was as follows: low risk of bias (the study protocol is available and all of the prespecified outcomes that are of interest in the review are reported in the prespecified way; the study protocol is not available but the published report includes all expected outcomes); high risk of bias (e.g. not all of the prespecified primary outcomes are reported; one or more primary outcomes are reported using measurements, analysis methods or subsets of the data that were not prespecified; one or more reported primary outcomes are not prespecified); unclear risk of bias.

Size of study (checking for possible biases related to small size)

We assessed studies as follows: low risk of bias (more than or equal to 200 participants per treatment arm); high risk of bias (fewer than 50 participants per treatment arm); unclear risk of bias (50 to 199 participants per treatment arm).

We had intended to describe special issues in the analysis of studies with non-standard designs, but we found no such studies.

Dealing with missing data If some data were not reported or not clearly reported for some outcomes or groups, we attempted to contact the study authors for further information. We had intended to conduct an intention-to-treat (ITT) analysis to estimate whether the intervention effect was biased if some participants were excluded from analysis in the randomized trials, or if they were lost to follow-up. We would have analyzed available data on all participants in each arm, regardless of what happened subsequently (Newell 1992). We had intended to use the baseline observation carried forward (BOCF) approach to avoid the analgesic effect calculation of the original intervention being affected by remedication. After remedication, pain intensity would revert to its initial value and pain relief would become zero for all subsequent time points (Moore 2005). However, there was insufficient information to implement this.

Other sources of bias

We stated any important concerns about bias that are not addressed by the other domains. The judgment was as follows: low risk of bias (the study appears to be free of other sources of bias); high risk of bias (at least one important risk of bias due to problems not covered elsewhere in the criteria above); unclear risk of bias.

Measures of treatment effect We carried out statistical analysis of treatment effects using the Review Manager 5 software (RevMan 2014) where two or more studies investigated the same outcome. For dichotomous data, we present results as a summary risk ratio (RR) and 95% confidence interval (95% CI). Where possible, we calculated number needed to treat for an additional beneficial outcome (NNTB) or for an additional harmful outcome (NNTH), together with 95% CIs. For continuous data, we used the mean difference (MD) and the 95% CI for similar outcome measures. When the continuous data were presented as the median and interquartile range, we estimated the mean as equivalent to the median and the standard deviation as a quarter of the interquartile range. When the continuous data were presented as a 95% CI, we

Assessment of heterogeneity We used the I² statistic (Higgins 2003) to identify whether there was statistical heterogeneity. If there was no significant heterogeneity (I² value 0% to 40%), we used a fixed-effect model for metaanalysis. We used a random-effects model if the I² value was above 40%.

Assessment of reporting biases The methods of assessing the risk of reporting bias are described above in ’Selective reporting (checking for possible reporting bias)’ (Assessment of risk of bias in included studies). When we suspected reporting bias, we performed a sensitivity analysis to explore whether the related studies contributed severe bias. We intended to use funnel plots to explore publication bias. However, all analyses contained fewer than 10 studies, so the power of the tests was too low to distinguish chance from real asymmetry (Higgins 2011).


8

Data synthesis We used a fixed-effect model to conduct meta-analysis if data were homogeneous, but otherwise a random-effects model. Where the data were unsuitable for meta-analysis, we described the findings of multiple studies separately.

Subgroup analysis and investigation of heterogeneity We took into consideration several potential sources of heterogeneity. Firstly, the types and doses of analgesics applied in each intervention were not standardized between studies. Secondly, peripheral nerve blocks may be administered as single-shot or continuous doses, and continuous peripheral nerve blocks increased the duration of the analgesic effect beyond that of single-shot peripheral nerve blocks. Thirdly, applying the intervention to different locations, such as femoral nerve block only, or combining a femoral with a sciatic or obturator nerve block, may have different analgesic effects. Fourthly, major knee surgery included various types of operation. Peripheral nerve blocks may exert different effects on the participants undergoing different surgeries. However, we only managed to conduct subgroup analysis according to the types of surgery. There was insufficient information (a minimum of two studies and 200 participants) (Moore 1998) to carry out subgroup analysis to check whether other differences affect the results.

Sensitivity analysis We intended to perform sensitivity analyses to explore the effect of quality score (two versus three or more) and trial size (39 or fewer versus 40 or more per treatment arm) for important outcomes in the review.

RESULTS Description of studies See: Characteristics of included studies; Characteristics of excluded studies; Characteristics of studies awaiting classification; Characteristics of ongoing studies

Results of the search We screened 407 papers from CENTRAL, MEDLINE, and EMBASE. After reading the titles and abstracts, we identified 27 papers as potentially eligible for inclusion. We excluded four papers after reading the full text. Two studies were only available in the form of abstract and we failed to get further information from the authors. See: Characteristics of studies awaiting classification. Finally, we identified 23 studies that met the inclusion criteria. The

study selection process is summarized in Figure 1. We also found three ongoing studies, which will be considered for inclusion in the update of this review. See: Characteristics of ongoing studies. Included studies All of the included studies recruited participants older than 15 years and the sex ratio was balanced. Participants were in good general health, and were excluded if they had an American Society of Anesthesiologists (ASA) score worse than . Two studies (Good 2007; Widmer 2012) did not report the ASA score of their participants. With regard to the types of surgery, 14 studies (Allen 1998; Carli 2010; Chan 2012; Ganapathy 1999; Good 2007; Hirst 1996; Ilfeld 2010; Jenstrup 2012; Moghtadaei 2014; Ng 2001; Szczukowski 2004; Wang 2002; Widmer 2012; Xie 2012) evaluated people undergoing total knee arthroplasty. Six studies (Espelund 2013; Lundblad 2011; Matava 2009; Mulroy 2001; Williams 2006; Wulf 2010) evaluated people undergoing anterior cruciate ligament reconstruction. Two studies (Schultz 1991; Tierney 1987) evaluated those undergoing open knee surgery, and one study (Akkaya 2008) evaluated people undergoing partial menisectomy arthroscopically. The included studies varied in the types of peripheral nerve blocks. There was respectively one included study used saphenous nerve block (Akkaya 2008), infrapatellar nerve block (Lundblad 2011), and femoral nerve block plus sciatic nerve block (Allen 1998). Two studies (Espelund 2013; Jenstrup 2012) used adductor canal block. Five studies (Ganapathy 1999; Hirst 1996; Ng 2001; Schultz 1991; Xie 2012) used three-in-one block. Thirteen studies used femoral nerve block, among which, three studies (Carli 2010; Ilfeld 2010; Williams 2006) administered block continuously via a catheter and a pump, while the other 10 studies (Chan 2012; Good 2007; Matava 2009; Moghtadaei 2014; Mulroy 2001; Szczukowski 2004; Tierney 1987; Wang 2002; Widmer 2012; Wulf 2010) used a single shot. Anesthetic used in the blocks was also variable. Fourteen studies (Allen 1998; Chan 2012; Ganapathy 1999; Good 2007; Hirst 1996; Matava 2009; Mulroy 2001; Ng 2001; Schultz 1991; Szczukowski 2004; Tierney 1987; Wang 2002; Wulf 2010; Xie 2012) used bupivacaine; three studies (Akkaya 2008; Lundblad 2011; Williams 2006) used levo bupivacaine; eight studies (Carli 2010; Espelund 2013; Ilfeld 2010; Jenstrup 2012; Moghtadaei 2014; Ng 2001; Widmer 2012; Wulf 2010) used ropivacaine. To locate the nerve, 16 studies (Akkaya 2008; Allen 1998; Carli 2010; Ganapathy 1999; Good 2007; Hirst 1996; Ilfeld 2010; Matava 2009; Moghtadaei 2014; Mulroy 2001; Ng 2001; Schultz 1991; Szczukowski 2004; Wang 2002; Widmer 2012; Wulf 2010) used nerve stimulators; three studies (Espelund 2013; Jenstrup 2012; Lundblad 2011) used ultrasound guidance. There was one study (Chan 2012) combining ultrasound imaging and nerve stimulators. One study (Tierney 1987) located the nerve without the aid of any instrument. The remaining two studies (Williams 2006; Xie 2012) did not describe


9

the exact methods. In terms of block timing, eight studies (Akkaya 2008; Allen 1998; Good 2007; Hirst 1996; Ilfeld 2010; Lundblad 2011; Widmer 2012; Xie 2012) performed this before the induction of anesthesia; two studies (Matava 2009; Wulf 2010) performed this after anesthetic induction but before surgical incision; and 12 studies (Carli 2010; Espelund 2013; Ganapathy 1999; Jenstrup 2012; Moghtadaei 2014; Mulroy 2001; Ng 2001; Schultz 1991; Szczukowski 2004; Tierney 1987; Wang 2002; Williams 2006) performed this at the end of surgery. Chan 2012 had two intervention groups, with one performing block after anesthesia but before the surgical procedure, and the other group performing this postoperatively. Nineteen studies (Akkaya 2008; Allen 1998; Chan 2012; Espelund 2013; Ganapathy 1999; Good 2007; Hirst 1996; Ilfeld 2010; Jenstrup 2012; Lundblad 2011; Matava 2009; Mulroy 2001; Ng 2001; Szczukowski 2004; Tierney 1987; Wang 2002; Williams 2006; Wulf 2010; Xie 2012) compared peripheral nerve blocks as adjunctive techniques to systemic analgesia with systemic analgesia alone. Three studies (Carli 2010; Moghtadaei 2014; Widmer 2012) compared peripheral nerve blocks with local infiltration and one study (Schultz 1991) compared peripheral nerve blocks with epidural analgesia; all four used systemic analgesia as remedication. Ten included studies (Allen 1998; Ganapathy 1999; Hirst 1996; Ilfeld 2010; Jenstrup 2012; Mulroy 2001; Schultz 1991; Szczukowski 2004; Tierney 1987; Wulf 2010) presented the outcomes in the form of graphs without exact data. We attempted to acquire the data through email and acquired the data for one study (Jenstrup 2012). We describe the details of the included studies in the ’ Characteristics of included studies’ table.

Excluded studies All four excluded studies (Bogoch 2002; De Lima 2008; Frassanito 2010; Hunt 2009) were randomized trials comparing peripheral nerve blocks with other analgesia methods. Among them, Bogoch 2002 recruited participants undergoing either total hip arthroplasty or total knee arthroplasty, but with the outcomes for participants undergoing total knee arthroplasty not presented separately. The remaining three studies claimed to be participantblind. However, sham block was not performed in the comparator treatment group, and the measures to maintain participant blinding were not described. We describe the details of the excluded studies in the ’ Characteristics of excluded studies’ table.

Risk of bias in included studies All 23 included studies were randomized and either double-blind or participant-blind. Two review authors scored each study independently for quality, using the Oxford quality and validity scale, with a maximum possible score of five. Nine included studies (Carli 2010; Espelund 2013; Ilfeld 2010; Jenstrup 2012; Lundblad 2011; Matava 2009; Moghtadaei 2014; Szczukowski 2004; Williams 2006) scored five, nine studies (Chan 2012; Good 2007; Mulroy 2001; Ng 2001; Schultz 1991; Tierney 1987; Wang 2002; Widmer 2012; Xie 2012) scored four, and the remaining five studies (Akkaya 2008; Allen 1998; Ganapathy 1999; Hirst 1996; Wulf 2010) scored three. The quality scores for individual studies are recorded in the ’Characteristics of included studies’ table. Risk of bias is also presented in the ’Risk of bias’ graph and ’Risk of bias’ summary (Figure 2; Figure 3).

Figure 2. Risk of bias graph: review authors’ judgements about each risk of bias item presented as percentages across all included studies.


10

Figure 3. Risk of bias summary: review authors’ judgements about each risk of bias item for each included study.


11

Allocation All 23 included studies reported that they were randomized, while only 15 of them (Akkaya 2008; Carli 2010; Chan 2012; Espelund 2013; Ilfeld 2010; Jenstrup 2012; Lundblad 2011; Matava 2009; Moghtadaei 2014; Mulroy 2001; Ng 2001; Szczukowski 2004; Tierney 1987; Widmer 2012; Williams 2006) described a detailed method to generate a random sequence and to conceal allocation.

Blinding Fifteen of the included studies (Carli 2010; Espelund 2013; Good 2007; Ilfeld 2010; Jenstrup 2012; Lundblad 2011; Matava 2009; Moghtadaei 2014; Schultz 1991; Szczukowski 2004; Tierney 1987; Wang 2002; Widmer 2012; Williams 2006; Xie 2012) were double-blind and adequately described how this was achieved. Although three studies (Chan 2012; Hirst 1996; Ng 2001) were stated to be double-blind, they only adequately described how they blinded the participants, but did not describe the exact method to blind the investigators. Five studies (Akkaya 2008; Allen 1998; Ganapathy 1999; Mulroy 2001; Wulf 2010) achieved participantblinding through use of a sham block with saline or by performing nerve block behind a drape to block the participant’s view. But these five studies did not report whether or not the investigators were blinded to the treatment assignment. For the blinding of outcome assessment, 19 of the included studies (Allen 1998; Carli 2010; Espelund 2013; Ganapathy 1999; Good 2007; Ilfeld 2010; Jenstrup 2012; Lundblad 2011; Matava 2009; Moghtadaei 2014; Mulroy 2001; Ng 2001; Schultz 1991; Szczukowski 2004; Tierney 1987; Wang 2002; Widmer 2012; Williams 2006; Xie 2012) stated that outcomes were assessed by blinded assessors, while the remaining four studies (Akkaya 2008; Chan 2012; Hirst 1996; Wulf 2010) did not describe this.

Selective reporting All the included studies reported our main outcome of pain intensity but did not cover all the secondary outcomes of interest. We rated all of them as unclear for risks of selective reporting. Other potential sources of bias According to the ’Checking for possible biases related to small size’ section of Assessment of risk of bias in included studies, only two included studies had 50 to 199 participants per treatment arm, and were therefore assessed as being at unclear risk of bias. The remaining 21 studies had fewer than 50 participants per treatment arm, and were assessed as being at high risk of bias.

Effects of interventions

1. Peripheral nerve blocks adjunctive to systemic analgesia versus systemic analgesia alone We include 19 studies comparing peripheral nerve blocks as adjunctive techniques to systemic analgesia with systemic analgesia alone (Akkaya 2008; Allen 1998; Chan 2012; Espelund 2013; Ganapathy 1999; Good 2007; Hirst 1996; Ilfeld 2010; Jenstrup 2012; Lundblad 2011; Matava 2009; Mulroy 2001; Ng 2001; Szczukowski 2004; Tierney 1987; Wang 2002; Williams 2006; Wulf 2010; Xie 2012). Some trials did not report every outcome prespecified in our protocol, and some trials did not report the outcome at every time interval of interest. Some trials presented data in graph form that we could not accurately extract. We were therefore only able to include eight of the 19 studies in any of our meta-analyses. All meta-analyses were for the primary outcome of pain intensity. Primary outcomes:

Incomplete outcome data Eleven of the included studies (Akkaya 2008; Allen 1998; Carli 2010; Ganapathy 1999; Hirst 1996; Matava 2009; Ng 2001; Schultz 1991; Szczukowski 2004; Tierney 1987; Wang 2002) reported no missing outcome data. Another 11 studies (Chan 2012; Espelund 2013; Good 2007; Ilfeld 2010; Jenstrup 2012; Lundblad 2011; Moghtadaei 2014; Mulroy 2001; Williams 2006; Wulf 2010; Xie 2012) reported missing outcome data. Reasons for the missing data were acceptable, and did not seem to have potential impact on outcomes. One study (Widmer 2012) had 55 participants who met the inclusion criteria and were allocated, but analyzed only 54 participants. The reason for lost follow-up was not described.

Pain intensity (Analyses 1.1 to 1.6) Pain intensity was assessed on a visual analogue scale (VAS) by participants themselves or by blinded assessors. Different authors reported this outcome on different scales. We standardized all of them to a 100 mm VAS so that we could analyze them in the review. Where the pain score was given but was not defined as measuring pain at rest or on movement, we considered it to be at rest. Data were generally reported as the average pain intensity of a time period. However, in studies in which the only data available were measured at a single time point within the interval of interest, we


12

used this measurement to represent the average pain intensity of that time period. Where measurements were reported at several time points within the interval of interest, we selected the one that measured at the time point closest to the mid-point of that interval. To clarify, an outcome measured at the time point closest to the 12th hour represented the ’zero to 23 hours’ interval, the time point closest to the 36th hour represented the ’24 to 47 hours’ interval, and the time point closest to the 60th hour represented the ’48 to 72 hours’ interval. Seven studies examined pain scores at rest in the interval from zero to 23 hours, involving 390 participants with 196 in the peripheral nerve blocks adjunctive to systemic analgesia group and 194 in the systemic analgesia alone group, respectively. Meta-analysis showed that the former group had a mean VAS score 11.85 points lower than the latter group (95% confidence interval (CI) -20.45 to 3.25, I² value 87%, random-effects model) (Analysis 1.1). Pain intensity at rest in the 24 to 47 hours interval was reported in six studies (320 participants, 162 with peripheral nerve blocks adjunctive to systemic analgesia and 158 with systemic analgesia alone, respectively), and showed significantly lower mean VAS in the adjunctive group than that in the control group (mean difference (MD) -12.92, 95% CI -19.82 to -6.02, I² value 72%, randomeffects model) (Analysis 1.2). Four studies (210 patients, 109 with peripheral nerve blocks adjunctive to systemic analgesia and 101 with systemic analgesia alone, respectively) analyzed pain intensity at rest in the 48 to 72 hours interval. Mean VAS of the adjunctive group was reported to be 9.72 points lower than that of the control group (95% CI -16.75 to -2.70, I² value 54%, randomeffects model) (Analysis 1.3). Five studies reported pain scores on movement in the zero to 23 hours interval (304 patients, 150 with peripheral nerve blocks adjunctive to systemic analgesia and 154 with systemic analgesia alone, respectively) and there appeared to be no significant difference between these two groups (Analysis 1.4). Only three studies (182 participants) reported results of VAS on movement in the 24 to 47 hours, and two studies (112 participants) reported results of VAS in the 48 to 72 hours interval. For the former comparison (24 to 47 hours), there was no significant difference between the groups (Analysis 1.5), while the latter comparison (48 to 72 hours) showed that the adjunctive group had a 6.19 points lower mean VAS than the control group (95% CI -11.76 to -0.62, I² value 0%, fixed-effect model) (Analysis 1.6). There was substantial heterogeneity between studies in most of the analyses, except for Analysis 1.6. The heterogeneity was likely due to different study protocols, different types of surgery, different times at which blocks were performed, and different time points of postoperative pain assessment. The results should therefore be interpreted cautiously. We stratified pain intensity by the type of surgery, and were able to create two subgroups: total knee arthroplasty with eight studies, and anterior cruciate ligament reconstruction with four studies. The number of studies for the remaining types of surgery was insufficient to create other subgroups, because only one study inves-

tigated partial menisectomy arthroscopically. In the sub-category of total knee arthroplasty, peripheral nerve blocks adjunctive to systemic analgesia showed significantly superior pain relief to systemic analgesia alone at rest in all three time intervals within 72 hours (MD -19.50, 95% CI -35.01 to -3.99 ( three studies, 183 participants, Analysis 1.1); MD -15.05, 95% CI -24.11 to -5.98 (four studies, 224 participants, Analysis 1.2); MD -13.73, 95% CI -18.21 to -9.26 (three studies, 154 participants, Analysis 1.3) respectively). In the zero to 23 hours postoperative interval, there was no significant difference in the VAS for participants undergoing total knee arthroplasty on movement (MD -12.93, 95% CI 26.62 to 0.75, two studies, 153 participants, Analysis 1.4) or for participants undergoing anterior cruciate ligament reconstruction at rest (MD -2.82, 95% CI -8.03 to 2.38, three studies, 167 participants, Analysis 1.1) or on movement (MD 0.10, 95% CI -7.49 to 7.70, two studies, 111 participants, Analysis 1.4). There were insufficient data to conduct a subanalysis in the 24 to 47 hours and 48 to 72 hours intervals postoperatively for anterior cruciate ligament reconstruction. Sensitivity analysis All of the included studies were ranked as equal to or better than 3/5 on the Oxford quality and validity scale. Only one study ( Chan 2012) in the meta-analysis had 40 or more participants per treatment arm, and removing the outcome of this study did not affect the results. Secondary outcomes:

Proportion of participants with ’no worse than mild pain’ Two studies (Lundblad 2011; Williams 2006) reported the proportion of participants with ’no worse than mild pain’. Lundblad 2011 observed that in the zero to 23 hours interval postoperatively, the percentage of participants with ’no worse than mild pain’ was significantly higher in the adjunctive group as compared to the control group at rest but not on movement. Data were only available in graph form. Williams 2006 recorded that in the 24 to 47 hours interval postoperatively, the proportion of ’no worse than mild pain’ in the systemic analgesia alone group, single peripheral nerve block adjunctive to systemic analgesia group, and continuous peripheral nerve block adjunctive to systemic analgesia group was 36.8%, 52.6%, and 73.7% respectively. In the 48 to 72 hours interval postoperatively, the proportions were 39.5%, 46.8%, and 68% respectively. Additional analgesic consumption within 72 hours after surgery and median time to remedication Eighteen included studies recorded additional analgesic consumption within 72 hours postoperatively. Only three of them (Akkaya 2008; Chan 2012; Jenstrup 2012) presented data in the form of a mean and standard deviation; Akkaya 2008 reported less tramadol


13

consumption in the adjunctive group in the zero to 23 hours interval (mean consumption: 108.5 mg versus 217.1 mg). Chan 2012 and Jenstrup 2012 reported less morphine consumption in the adjunctive group in the zero to 23 hours interval (mean consumption:15.71 mg versus 29.82 mg, and 40 mg versus 56 mg, respectively). Among the remaining 15 studies, four studies (Espelund 2013; Hirst 1996; Ilfeld 2010; Matava 2009) reported additional analgesic use did not differ between the two groups, while 11 studies (Allen 1998; Ganapathy 1999; Good 2007; Mulroy 2001; Ng 2001; Szczukowski 2004; Tierney 1987; Wang 2002; Williams 2006; Wulf 2010; Xie 2012) found less additional analgesic consumption in the adjunctive group than in the control group. In addition, three studies (Chan 2012; Tierney 1987; Wulf 2010) reported that participants in the adjunctive group had a significantly longer time to first request for additional analgesic than the control group.

Adverse events Seventeen of the included studies reported various types of adverse events, including nausea, vomiting, sedation, pruritus, urinary retention, fever, and motor block. We were unable to conduct a meta-analysis on specific type of adverse event. Among them, 11 studies (Akkaya 2008; Allen 1998; Espelund 2013; Ganapathy 1999; Good 2007; Jenstrup 2012; Matava 2009; Mulroy 2001; Ng 2001; Williams 2006; Xie 2012) observed no significant difference in the incidence of adverse events between the adjunctive group and the control group. Four studies (Chan 2012; Hirst 1996; Szczukowski 2004; Wang 2002) found that those in the adjunctive group had significant reductions in adverse events compared with the control group. Two studies (Ilfeld 2010; Wulf 2010) reported that motor block was significantly increased in the adjunctive group compared with the control group, and Ilfeld 2010 observed that three of the 39 participants in the adjunctive group fell within 72 hours postoperatively. None of the included studies reported any withdrawal of cases due to the adverse events.

Szczukowski 2004 reported that there was no significant difference in range of knee motion over 24 to 47 hours or over 48 to 72 hours postoperatively. Good 2007 also reported no significant difference in range of knee motion over 48 to 72 hours postoperatively. However, Ilfeld 2010 and Szczukowski 2004 presented the outcomes in a form that could not be standardized to the mean and the standard deviation; pooling of data was therefore not possible. Length of hospital stay and hospital costs Four included studies (Matava 2009; Ng 2001; Szczukowski 2004; Wang 2002) reported the length of hospital stay, while two of them presented the outcomes in a form that could not be standardized to the mean and the standard deviation. Wang 2002 showed that length of hospital stay was significantly shorter in the adjunctive group than in the control group (mean length: 3 days versus 4 days). However, Matava 2009, Ng 2001 and Szczukowski 2004 reported no significant difference between the two groups. Only one study (Matava 2009) reported hospital costs, and found no significant difference between the two groups. Participant satisfaction Four studies (Allen 1998; Hirst 1996; Matava 2009; Xie 2012) recorded overall satisfaction with the pain management on a 5point scale or a 10-point scale ranked by participants, while three of them presented the outcomes in a form that could not be standardized to the mean and the standard deviation. Allen 1998, Hirst 1996 and Matava 2009 reported no difference between the adjunctive group and the control group. Xie 2012 reported that participants in the low-dose block adjunctive to systemic analgesia group were more likely to have higher satisfaction than those in the systemic analgesia alone group (odds ratio (OR) 5.64, 95% CI 1.60 to 19.8) and than participants in high-dose block adjunctive to systemic analgesia group (OR 5.34, 95% CI 1.53 to 18.7). 2. Peripheral nerve blocks versus local infiltration

Knee range of motion within 72 hours after surgery Knee range of motion included range of flexion and range of extension, which was reported in six studies (Chan 2012; Ganapathy 1999; Good 2007; Ilfeld 2010; Szczukowski 2004; Wang 2002) within 72 hours postoperatively. Most of them did not mention whether the knee range of motion was active or passive, except for Ilfeld 2010, who recorded that the knee range of motion was achieved in a passive way. Among all six studies, only Wang 2002 reported knee range of flexion, and found a significantly higher degree in the adjunctive group than in the control group in the 48 to 72 hours interval postoperatively (mean degree: 70° versus 60°). Ganapathy 1999 observed increased range of motion achieved by the participants in the adjunctive group over 24 to 47 hours postoperatively, but without exact data. Chan 2012, Ilfeld 2010 and

We include three studies comparing peripheral nerve blocks with local infiltration (Carli 2010; Moghtadaei 2014; Widmer 2012). There were too few data to conduct a meta-analysis. We therefore describe the findings of the individual studies separately.

Primary outcomes:

Pain intensity Two studies (Carli 2010; Moghtadaei 2014) reported pain intensity within 72 hours postoperatively. Carli 2010 reported a trend of less pain at rest in the block group (mean VAS: 20 mm versus 50 mm in the 24 to 47 hours interval, 10 mm versus 30 mm in the


14

48 to 72 hours interval), but on walking, there was no significant difference in pain intensity between the two groups. Pain intensity at greatest knee flexion was also similar between these two groups in the 24 to 47 hours interval and the 48 to 72 hours interval postoperatively. Moghtadaei 2014 reported that participants in the local infiltration group suffered less pain in the first six hours after surgery (mean VAS: 30 mm versus 40 mm), while the block group reported better pain relief within 12 hours postoperatively (mean VAS: 50 mm versus 60 mm). Eventually, there appeared to be no significant difference in pain score between the two groups 24 hours after surgery.

Secondary outcomes:

Participant satisfaction Only one study (Moghtadaei 2014) reported a satisfaction scale ranked by participants, which did not reveal a significant difference between the two groups within 48 hours after surgery.

3. Peripheral nerve blocks versus epidural analgesia We include only one study comparing peripheral nerve blocks with epidural analgesia (Schultz 1991). This study observed the pain intensity and supplemental morphine consumption only during the first 18 hours postoperatively, and did not find a significant difference between the three-in-one nerve block group and the epidural analgesia group. However, the block group had a statistically significantly lower incidence of nausea, vomiting, pruritus and urinary retention than the epidural analgesia group.

Additional analgesic consumption within 72 hours after surgery Two studies (Moghtadaei 2014; Widmer 2012) reported additional analgesic consumption. Moghtadaei 2014 reported consumption of morphine or tramadol, and found significantly less consumption in the local infiltration group than in the block group in the first 24 hours after surgery (mean consumption:10 mg versus 12.5 mg). However, no difference was demonstrated between the two groups within 48 hours postoperatively. Widmer 2012 reported consumption of fentanyl, and observed that the block group consumed significantly less fentanyl than the local infiltration group in the first 24 hours postoperatively (mean consumption: 973 mg versus 1502 mg).

DISCUSSION We included 23 studies (1571 participants) examining the effectiveness and safety of peripheral nerve blocks on pain relief after major knee surgery. Nineteen of the 23 studies compared peripheral nerve blocks as adjunctive techniques to systemic analgesia with systemic analgesia alone, three studies compared peripheral nerve blocks with local infiltration, and one study compared peripheral nerve blocks with epidural analgesia. The studies were of high methodological quality overall. Given the heterogeneity and the small numbers of studies and participants, evidence for or against the efficacy and safety of peripheral nerve blocks is currently inconclusive, and should be interpreted cautiously.

Knee range of motion within 72 hours after surgery and median time to ambulation Two studies (Carli 2010; Moghtadaei 2014) reported median time to ambulation. Median time for each participant to have the first walk was about 24 hours in Carli 2010 and 12 hours in Moghtadaei 2014, with no significant difference between the block group and the local infiltration group. Carli 2010 reported knee range of motion within 72 hours postoperatively and observed no significant difference between the two groups over 24 to 47 hours and 48 to 72 hours postoperatively. Whether the knee range of motion was achieved actively or passively was not mentioned.

Length of hospital stay Two studies (Carli 2010; Moghtadaei 2014) reported length of hospital stay. Both studies showed no significant difference between the two groups.

Summary of main results Compared with systemic analgesia alone, peripheral nerve blocks as adjunctive techniques to systemic analgesia resulted in a lower pain intensity score at rest in the zero to 23 hours interval, the 24 to 47 hours interval and the 48 to 72 hours interval postoperatively. Pain intensity was also reduced on movement in the 48 to 72 hours interval. There appeared to be no significant differences for pain on movement over the time period of zero to 23 hours and 24 to 47 hours. In the subcategory of total knee arthroplasty, meta-analysis favored peripheral nerve blocks adjunctive to systemic analgesia at rest in all three intervals within 72 hours postoperatively. In the zero to 23 hours postoperative interval, there was no significant difference in the mean visual analogue scale (VAS) for participants undergoing total knee arthroplasty on movement or for participants undergoing anterior cruciate ligament reconstruction at rest and on movement.


15

Overall completeness and applicability of evidence Studies were powered to identify the pain relief efficacy of peripheral nerve blocks as adjunctive techniques to systemic analgesia when compared with systemic analgesia alone following major knee surgery, but the numbers of studies and participants were too few to draw conclusions on other outcomes prespecified in our review. Three studies met the inclusion criteria comparing peripheral nerve blocks with local infiltration, and only one study met the inclusion criteria comparing peripheral nerve blocks with epidural analgesia. Data from these four studies were inadequate to be pooled for analysis. We have therefore described the findings of individual studies separately. None of the studies comparing peripheral nerve blocks with spinal analgesia was included in our review.

studies covered by these reviews, because of methodological issues. Despite some differences in approach, the findings of these two systematic reviews were similar to our review in terms of pain relief efficacy of peripheral nerve blocks adjunctive to systemic analgesia compared with systemic analgesia alone. Another previous review (Fowler 2008), comparing peripheral nerve blocks with epidural analgesia after major knee surgery, showed that peripheral nerve blocks provided comparable postoperative analgesia with epidural techniques, and caused neuraxial complication less frequently than epidural analgesia. However, none of the eight included studies was participant-blind, and therefore did not meet the inclusion criteria of this review.

AUTHORS’ CONCLUSIONS Quality of the evidence

Implications for practice

The evidence for the analgesic efficacy of peripheral nerve blocks following major knee surgery was based on studies of high methodological quality, but sample sizes were somewhat limited (21 of the 23 included studies had fewer than 50 participants per treatment arm). Moreover, there was considerable heterogeneity among studies, likely due to different study protocols, different types of surgery, different times at which blocks were performed, and different time points and scales of postoperative pain assessment. The results should therefore be interpreted with caution.

Our review provides evidence in support of the use of peripheral nerve blocks as adjunctive analgesic methods to systemic analgesia following major knee surgery within the first 72 hours. However, the benefits may be limited to those undergoing total knee arthroplasty. The current evidence is insufficient to draw conclusions on other outcomes (morphine consumption, adverse events, knee range of motion, length of hospital stay, hospital costs, and patient satisfaction). The comparison with local infiltration, epidural and spinal analgesia is limited by a lack of evidence.

Implications for research Potential biases in the review process We carried out extensive searches to identify relevant studies, and minimised the potential biases by having two review authors complete each assessment of eligibility and data extraction. However, there always remains the possibility of unidentified studies. In addition, all of the included studies used remedication once the pain relief effect of the original intervention seemed insufficient; any benefit seen for peripheral nerve blocks may therefore be a combination of the benefit of peripheral nerve blocks and remedication used. We had intended to use the ’baseline observation carried forward’ approach to avoid interference by the effect of remedication, but there was insufficient information to do so. We found no other potential bias in the review process.

Agreements and disagreements with other studies or reviews Two previous systematic reviews compared femoral nerve block with an intravenous patient-controlled analgesia opioid, epidural analgesia, local infiltration analgesia, and oral analgesia after total knee arthroplasty (Chan 2014; Paul 2010). We excluded several

1. Although the number of studies currently available is adequate, the number of participants per treatment arm in each study is limited. Small studies are thought to be at increased risk of bias. Multicenter trials with collaborative research groups may solve this problem by increasing the number of participants in each study. 2. Presentation of continuous outcomes in the form of graphs instead of exact data hinders the pooling of results. Measurements of continuous outcomes therefore need to be standardized to be more accessible, such as the mean and standard deviation. 3. Sham block or placebo block is needed to encourage the participant to have the same expectations of treatment throughout the entire treatment process. However, as the insertion of the placebo catheter is an invasive procedure, the potential ethical problems can not be ignored. The advantages and disadvantages of sham block or placebo block should be weighed carefully. 4. The increasingly prevalent ’enhanced recovery’ protocol values both good pain control and length of hospital stay (Malviya 2011). The paucity of data for ’length of hospital stay’


16

makes it difficult to determine the role of peripheral nerve blocks in this protocol. Further research should include these data.

ACKNOWLEDGEMENTS The review authors wish to thank Dr Annette Swinkels and Ms Yi-jing Lu for their helpful suggestions on the protocol for this review. The review authors wish to thank Ms Joanne Abbott for

assistance in the search process. The review authors wish to thank Dr Tudor Phillips for his helpful suggestions on the review. The review authors wish to thank Ms Anna Hobson for her editorial support. CRG Funding Acknowledgement: The National Institute for Health Research (NIHR) is the largest single funder of the Cochrane PaPaS Group. Disclaimer: The views and opinions expressed therein are those of the authors and do not necessarily reflect those of the NIHR, NHS or the Department of Health.

REFERENCES

References to studies included in this review Akkaya 2008 {published data only} Akkaya T, Ersan O, Ozkan D, Sahiner Y, Akin M, Gumus H, et al.Saphenous nerve block is an effective regional technique for post-menisectomy pain. Knee Surgery, Sports Traumatology, Arthroscopy 2008;16(9):855–8.

Hirst 1996 {published data only} Hirst GC, Lang SA, Dust WN, Cassidy JD, Yip RW. Femoral nerve block. Single injection versus continuous infusion for total knee arthroplasty. Regional Anesthesia 1996;21(4):292–7.

Allen 1998 {published data only} Allen HW, Liu SS, Ware PD, Nairn CS, Owens BD. Peripheral nerve blocks improve analgesia after total knee replacement surgery. Anesthesia and Analgesia 1998;87(1): 93–7.

Ilfeld 2010 {published data only} Ilfeld BM, Mariano ER, Girard PJ, Loland VJ, Meyer RS, Donovan JF, et al.A multicenter, randomized, triplemasked, placebo-controlled trial of the effect of ambulatory continuous femoral nerve blocks on discharge-readiness following total knee arthroplasty in patients on general orthopaedic wards. Pain 2010;150(3):477–84.

Carli 2010 {published data only} Carli F, Clemente A, Asenjo JF, Kim DJ, Mistraletti G, Gomarasca M, et al.Analgesia and functional outcome after total knee arthroplasty: periarticular infiltration vs continuous femoral nerve block. British Journal of Anaesthesia 2010; Vol. 105, issue 2:185–95.

Jenstrup 2012 {published and unpublished data} Jenstrup MT, Jæger P, Lund J, Fomsgaard JS, Bache S, Mathiesen O, et al.Effects of adductor-canal-blockade on pain and ambulation after total knee arthroplasty: a randomized study. Acta Anaesthesiologica Scandinavica 2012;56(3):357–64.

Chan 2012 {published data only} Chan MH, Chen WH, Tung YW, Liu K, Tan PH, Chia YY. Single-injection femoral nerve block lacks preemptive effect on postoperative pain and morphine consumption in total knee arthroplasty. Acta Anaesthesiologica Taiwanica 2012;50 (2):54–8.

Lundblad 2011 {published data only} Lundblad M, Forssblad M, Eksborg S, Lönnqvist PA. Ultrasound-guided infrapatellar nerve block for anterior cruciate ligament repair: a prospective, randomised,doubleblind, placebo-controlled clinical trial. European Journal Anaesthesiology 2011;28(7):511–8.

Espelund 2013 {published data only} Espelund M, Fomsgaard JS, Haraszuk J, Mathiesen O, Dahl JB. Analgesic efficacy of ultrasound-guided adductor canal blockade after arthroscopic anterior cruciate ligament reconstruction: a randomised controlled trial. European Journal of Anaesthesiology 2013;30(7):422–8.

Matava 2009 {published data only} Matava MJ, Prickett WD, Khodamoradi S, Abe S, Garbutt J. Femoral nerve blockade as a preemptive anesthetic in patients undergoing anterior cruciate ligament reconstruction: a prospective, randomized, double-blinded, placebo-controlled study. The American Journal of Sports Medicine 2009;37(1):78–86.

Ganapathy 1999 {published data only} Ganapathy S, Wasserman RA, Watson JT, Bennett J, Armstrong KP, Stockall CA. Modified continuous femoral three-in-one block for postoperative pain after total knee arthroplasty. Anesthesia and Analgesia 1999;89(5): 1197–202. Good 2007 {published data only} Good RP, Snedden MH, Schieber FC, Polachek A. Effects of a preoperative femoral nerve block on pain management and rehabilitation after total knee arthroplasty. American Journal of Orthopedics 2007;36(10):554–7.

Moghtadaei 2014 {published data only} Moghtadaei M, Farahini H, Faiz SHR, Mokarami F, Safari S. Pain management for total knee arthroplasty: Singleinjection femoral nerve block versus local infiltration analgesia. Iranian Red Crescent Medical Journal 2014;16(1): e13247. Mulroy 2001 {published data only} Mulroy MF, Larkin KL, Batra MS, Hodgson PS, Owens BD. Femoral nerve block with 0.25% or 0.5% bupivacaine improves postoperative analgesia following outpatient


17

arthroscopic anterior cruciate ligament repair. Regional Anesthesia and Pain Medicine 2001;26(1):24–9. Ng 2001 {published data only} Ng HP, Cheong KF, Lim A, Lim J, Puhaindran ME. Intraoperative single-shot “3-in-1” femoral nerve block with ropivacaine 0.25%, ropivacaine 0.5% or bupivacaine0.25% provides comparable 48-hr analgesia after unilateral total knee replacement. Canadian Journal of Anaesthesia 2001;48 (11):1102–8. Schultz 1991 {published data only} Schultz P, Anker-Møller E, Dahl JB, Christensen EF, Spangsberg N, Faunø P. Postoperative pain treatment after open knee surgery: continuous lumbar plexus block with bupivacaine versus epidural morphine. Regional Anesthesia 1991;16(1):34–7. Szczukowski 2004 {published data only} Szczukowski MJ Jr, Hines JA, Snell JA, Sisca TS. Femoral nerve block for total knee arthroplasty patients: a method to control postoperative pain. The Journal of Arthroplasty 2004;19(6):720–5. Tierney 1987 {published data only} Tierney E, Lewis G, Hurtig JB, Johnson D. Femoral nerve block with bupivacaine 0.25 per cent for postoperative analgesia after open knee surgery. Canadian Journal of Anaesthesia 1987;34(5):455–8. Wang 2002 {published data only} 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. Regional Anesthesia and Pain Medicine 2002;27(2):139–44. Widmer 2012 {published data only} Widmer BJ, Scholes CJ, Pattullo GG, Oussedik SI, Parker DA, Coolican MR. Is femoral nerve block necessary during total knee arthroplasty?: a randomized controlled trial. The Journal of Arthroplasty 2012;27(10):1800–5. Williams 2006 {published data only} Williams BA, Kentor ML, Vogt MT, Irrgang JJ, Bottegal MT, West RV, et al.Reduction of verbal pain scores after anterior cruciate ligament reconstruction with 2-day continuous femoral nerve block: a randomized clinical trial. Anesthesiology 2006;104(2):315–27. Wulf 2010 {published data only} Wulf H, Löwe J, Gnutzmann KH, Steinfeldt T. Femoral nerve block with ropivacaine or bupivacaine in day case anterior crucial ligament reconstruction. Acta Anaesthesiologica Scandinavica 2010;54(4):414–20. Xie 2012 {published data only} Xie Z, Hussain W, Cutter TW, Apfelbaum JL, Drum ML, Manning DW. Three-in-one nerve block with different concentrations of bupivacaine in total knee arthroplasty: randomized, placebo-controlled, double-blind trial. The Journal of Arthroplasty 2012;27(5):673–8.e1.

References to studies excluded from this review

Bogoch 2002 {published data only} Bogoch ER, Henke M, Mackenzie T, Olschewski E, Mahomed NN. Lumbar paravertebral nerve block in the management of pain after total hip and knee arthroplasty: a randomized controlled clinical trial. The Journal of Arthroplasty 2002;17(4):398–401. De Lima 2008 {published data only} De Lima E Souza R, Correa CH, Henriques MD, De Oliveira CB, Nunes TA, Gomez RS. Single-injection femoral nerve block with 0.25% ropivacaine or 0.25% bupivacaine for postoperative analgesia after total knee replacement or anterior cruciate ligament reconstruction. The Journal of Clinical Anesthesia 2008;20(7):521–7. Frassanito 2010 {published data only} Frassanito L, Vergari A, Zanghi F, Messina A, Bitondo M, Antonelli M. Post-operative analgesia following total knee arthroplasty: comparison of low-dose intrathecal morphine and single-shot ultrasound-guided femoral nerve block: a randomized, single blinded, controlled study. European Review for Medical and Pharmacological Sciences 2010;14(7): 589–96. Hunt 2009 {published data only} Hunt KJ, Bourne MH, Mariani EM. Single-injection femoral and sciatic nerve blocks for pain control after total knee arthroplasty. The Journal of Arthroplasty 2009;24(4): 533–8.

References to studies awaiting assessment Tugay 2006 {published data only} Tugay N, Saricaoglu F, Satilmis T, Alpar U, Akarcali I, Citaker S, et al.Effects on the independence level in functional activities in the early postoperative period in patients with total knee arthroplasty. Neurosciences 2006;11 (3):175–9. Wongswadiwat 2012 {published data only} Wongswadiwat M, Pathanon P, Sriraj W, Yimyaem PR, Bunthaothuk S. Single injection fascia iliaca block for pain control after arthroscopic anterior cruciate ligament reconstruction: a randomized, controlled trial. Journal of the Medical Association of Thailand 2012;95(11):1418–24.

References to ongoing studies NCT01616836 {published data only} NCT01616836. Optimizing pain and rehabilitation after knee arthroplasty (OPRA). clinicaltrials.gov/show/ NCT01616836 (accessed 28th November 2014). [: NCT01616836] NCT01683071 {published data only} NCT01683071. Femoral nerve block with liposome bupivacaine for postsurgical analgesia following total knee arthroplasty. clinicaltrials.gov/show/NCT01683071 (accessed 29th November 2014). [: NCT01683071] NCT02008617 {unpublished data only} NCT02008617. Analgesic benefits of genicular nerve blocks of the posterior knee for patients undergoing anterior


18

cruciate ligament (ACL) reconstruction. clinicaltrials.gov/ show/NCT02008617 (accessed 29th November 2014).

Additional references Abdallah 2011 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. Regional Anesthesia and Pain Medicine 2011;36(5): 493–8. Aguirre 2012 Aguirre J, Del Moral A, Cobo I, Borgeat A, Blumenthal S. The role of continuous peripheral nerve blocks. Anesthesiology Research and Practice 2012;2012:560879. Ansbro 1946 Ansbro FP. A method of continuous brachial plexus block. American Journal of Surgery 1946;71:716–22. Bagry 2008 Bagry H, De la Cuadra Fontaine JC, Asenjo JF, Bracco D, Carli F. Effect of a continuous peripheral nerve block on the inflammatory response in knee arthroplasty. Regional Anesthesia and Pain Medicine 2008;33(1):17–23. Capdevila 1999 Capdevila X, Barthelet Y, Biboulet P, Ryckwaert Y, Rubenovitch J, D’Athis F. Effects of perioperative analgesic technique on the surgical outcome and duration of rehabilitation after major knee surgery. Anesthesiology 1999; 91(1):8–15. Chan 2013 Chan EY, Fransen M, Sathappan S, Chua NH, Chan YH, Chua N. Comparing the analgesia effects of singleinjection and continuous femoral nerve blocks with patient controlled analgesia after total knee arthroplasty. Journal of Arthroplasty 2013;28(4):608–13. Chan 2014 Chan EY, Fransen M, Parker DA, Assam PN, Chua N. Femoral nerve blocks for acute postoperative pain after knee replacement surgery. Cochrane Database of Systematic Reviews 2014, Issue 5. [DOI: 10.1002/ 14651858.CD009941.pub2] Choi 2003 Choi PT, Bhandari M, Scott J, Douketis J. Epidural analgesia for pain relief following hip or knee replacement. Cochrane Database of Systematic Reviews 2003, Issue 3. [DOI: 10.1002/14651858.CD003071] Collins 1997 Collins SL, Moore RA, McQuay HJ. The visual analogue pain intensity scale: what is moderate pain in millimetres?. Pain 1997;72(1-2):95–7. Dervin 2012 Dervin GF, Madden SM, Crawford-Newton BA, Lane AT, Evans HC. Outpatient unicompartment knee arthroplasty with indwelling femoral nerve catheter. Journal of Arthroplasty 2012;27(6):1159–65.

Fowler 2008 Fowler SJ, Symons J, Sabato S, Myles PS. Epidural analgesia compared with peripheral nerve blockade after major knee surgery: a systematic review and meta-analysis of randomized trials. British Journal of Anaesthesia 2008;100 (2):154–64. Grosu 2014 Grosu I, Lavand’homme P, Thienpont E. Pain after knee arthroplasty: an unresolved issue. Knee Surgery, Sports Traumatology, Arthroscopy 2014;22(8):1744–58. [DOI: 10.1007/s00167-013-2750-2] Higgins 2003 Higgins JPT, Thompson SG, Deeks JJ, Altman DG. Measuring inconsistency in meta-analyses. BMJ 2003;327 (7414):557–60. Higgins 2011 Higgins JPT, Green S (editors). Cochrane Handbook for Systematic Reviews of Interventions Version 5.1.0 [updated March 2011]. The Cochrane Collaboration, 2011. Available from www.cochrane-handbook.org. Hogan 2009 Hogan MV, Grant RE, Lee L Jr. Analgesia for total hip and knee arthroplasty: a review of lumbar plexus, femoral, and sciatic nerve blocks. American Journal of Orthopedics 2009; 38(8):E129–33. Jadad 1996 Jadad AR, Moore RA, Carroll D, Jenkinson C, Reynolds DJ, Gavaghan DJ, et al.Assessing the quality of reports of randomized clinical trials: is blinding necessary?. Controlled Clinical Trials 1996;17(1):1–12. Liberati 2009 Liberati A, Altman DG, Tetzlaff J, Mulrow C, Gøtzsche PC, Ioannidis JP, et al.The PRISMA statement for reporting systematic reviews and meta-analyses of studies that evaluate health care interventions: explanation and elaboration. PLoS Medicine 2009;6(7):e1000100. [DOI: 10.1371/ journal.pmed.1000100] Malviya 2011 Malviya A, Martin K, Harper I, Muller SD, Emmerson KP, Partington PF. Enhanced recovery program for hip and knee replacement reduces death rate. Acta Orthopaedica 2011;82 (5):577–81. Martin 2008 Martin F, Martinez V, Mazoit JX, Bouhassira D, Cherif K, Gentili ME, et al.Antiinflammatory effect of peripheral nerve blocks after knee surgery: clinical and biologic evaluation. Anesthesiology 2008;109(3):484–90. Moore 1998 Moore RA, Gavaghan D, Tramèr MR, Collins SL, McQuay HJ. Size is everything--large amounts of information are needed to overcome random effects in estimating direction and magnitude of treatment effects. Pain 1998;78(3):209–16.


19

Moore 2005 Moore RA, Edwards JE, McQuay HJ. Acute pain: individual patient meta-analysis shows the impact of different ways of analysing and presenting results. Pain 2005;116(3):322–31. Moore 2013 Moore RA, Straube S, Aldington D. Pain measures and cut-offs - ’no worse than mild pain’ as a simple, universal outcome. Anaesthesia 2013;68(4):400–12. Mugabure Bujedo 2012 Mugabure Bujedo B. A clinical approach to neuraxial morphine for the treatment of postoperative pain. Pain Research and Treatment 2012;2012:612145. Newell 1992 Newell DJ. Intention-to-treat analysis: implications for quantitative and qualitative research. International Journal of Epidemiology 1992;21(5):837–41. Paul 2010 Paul JE, Arya A, Hurlburt L, Cheng J, Thabane L, Tidy A, et al.Femoral nerve block improves analgesia outcomes after total knee arthroplasty: a meta-analysis of randomized controlled trials. Anesthesiology 2010;113(5):1144–62. RevMan 2014 The Nordic Cochrane Centre, The Cochrane Collaboration. Review Manager (RevMan). 5.3. Copenhagen: The Nordic Cochrane Centre, The Cochrane Collaboration, 2014. Richman 2006 Richman JM, Liu SS, Courpas G, Wong R, Rowlingson AJ, McGready J, et al.Does continuous peripheral nerve block provide superior pain control to opioids? A meta-analysis. Anesthesia and Analgesia 2006;102(1):248–57. Rosenblatt 1980 Rosenblatt RM. Continuous femoral anesthesia for lower extremity surgery. Anesthesia and Analgesia 1980;59(8): 631–2. Sakai 2013 Sakai N, Inoue T, Kunugiza Y, Tomita T, Mashimo T. Continuous femoral versus epidural block for attainment of 120° knee flexion after total knee arthroplasty: a randomized controlled trial. Journal of Arthroplasty 2013;28 (5):807–14. Sala-Blanch 2012 Sala-Blanch X, De Riva N, Carrera A, López AM, Prats A, Hadzic A. Ultrasound-guided popliteal sciatic block

with a single injection at the sciatic division results in faster block onset than the classical nerve stimulator technique. Anesthesia and Analgesia 2012;114(5):1121–7. Smithwick 1930 Smithwick RH, White JC. Elimination of pain in obliterative vascular disease of lower extremity: technique for alcohol injection of sensory nerves of lower leg. Surgery, Gynecology, and Obstetrics 1930;51:394–403. Tantry 2012 Tantry TP, B G M, Hukkery R. Use of a single injection femoral nerve block in the patients of total knee replacement with concomitant epidural analgesia. Journal of Clinical Diagnostic Research 2012;6(10):1744–8. Teng 2012 Teng WN, Su YP, Kuo IT, Lin SM, Tsou MY, Chan KH, et al.Patient controlled epidural analgesia for bilateral versus unilateral total knee arthroplasty: a retrospective study of pain control. Journal of the Chinese Medical Association 2012;75(3):114–20. Ting 1989 Ting PL, Sivagnanaratnam V. Ultrasonographic study of the spread of local anaesthetic during axillary brachial plexus block. British Journal of Anaesthesia 1989;63(3):326–9. Weinstein 2013 Weinstein AM, Rome BN, Reichmann WM, Collins JE, Burbine SA, Thornhill TS, et al.Estimating the burden of total knee replacement in the United States. Journal of Bone and Joint Surgery. American Volume 2013;95(5):385–92. Yadeau 2013 Yadeau JT, Goytizolo EA, Padgett DE, Liu SS, Mayman DJ, Ranawat AS, et al.Analgesia after total knee replacement: local infiltration versus epidural combined with a femoral nerve blockade: a prospective, randomised pragmatic trial. The Bone & Joint Journal 2013;95-B(5):629–35.

References to other published versions of this review Xu 2014 Xu J, Chen XM, Ma CK, Wang XR. Peripheral nerve blocks for postoperative pain after major knee surgery. Cochrane Database of Systematic Reviews 2014, Issue 1. [DOI: 10.1002/14651858.CD010937] ∗ Indicates the major publication for the study


20

CHARACTERISTICS OF STUDIES

Characteristics of included studies [ordered by study ID] Akkaya 2008 Methods

Randomized, participant-blind, 2 parallel groups

Participants

Partial menisectomy arthroscopically N = 40 M = 20, F = 20 Mean age 43.6 years (experimental group) and 47.9 years (control group) ASA - (experimental group) and - (control group) Country: Turkey

Interventions

Sartorius nerve block (experimental group): 10 ml of 0.5% levo bupivacaine injected before surgery, N = 20 Saline (control group): 1 ml saline injected to the same region, N = 20

Outcomes

Pain intensity at rest and on movement at 0 hour, 2 hours, 4 hours, 6 hours, and 12 hours postoperatively Additional analgesic consumption Adverse events

Notes

Oxford quality and validity scale = 3

Risk of bias Bias

Authors’ judgement

Support for judgement

Random sequence generation (selection Low risk bias)

Quote: “An envelope was drawn and the patient was allocated into a group.”

Allocation concealment (selection bias)

Quote: “An envelope was drawn and the patient was allocated into a group.”

Low risk

Blinding of participants and personnel Unclear risk (performance bias) All outcomes

Comment: participants were blinded. Blinding of personnel was not described

Blinding of outcome assessment (detection Unclear risk bias) All outcomes

Comment: not described.

Incomplete outcome data (attrition bias) All outcomes

Low risk

Comment: no missing outcome data.

Selective reporting (reporting bias)

Unclear risk

Comment: proportion of participants with ’no worse than mild pain’, knee range of motion, length of hospital stay, hospital


21

Akkaya 2008

(Continued)

costs, and participant satisfaction were not reported Other bias

High risk

Comment: high risk of bias for size of study.

Allen 1998 Methods

Randomized, double-blind, 3 parallel groups

Participants

Total knee arthroplasty N = 36 M = 22, F = 14 Mean age 66 years (Femoral nerve block (Group FNB)), 69 years (Group Sciatic nerve block (SNB) + FNB), 68 years (control group) ASA Country: the United States

Interventions

Group FNB: 30 ml of 0.25% bupivacaine with 1:400,000 epinephrine, and a sham sciatic nerve block performed, N = 12 Group SNB+FNB: 30 ml of 0.25% bupivacaine with 1:400,000 epinephrine, N = 12 Control group: sham injections performed at both the inguinal and gluteal regions, N = 12

Outcomes

Pain intensity at rest at 4 hours and 8 hours, and on days 1 and 2 postoperatively Pain intensity on movement on day 1 and day 2 postoperatively Additional analgesic consumption Adverse events Participant satisfaction

Notes


Risk of bias Bias



Random sequence generation (selection Unclear risk bias)




Unclear risk



Quote: “All procedures were performed behind a drape to block patient’s view of procedure. Dressings were placed on all injection or sham injection sites.” Comment: blinding of personnel was not described.

22

Allen 1998

(Continued)

Blinding of outcome assessment (detection Low risk bias) All outcomes

Quote: “Assessed by blinded investigator”


Low risk



Unclear risk

Comment: proportion of participants with ’no worse than mild pain’, knee range of motion, length of hospital stay, and hospital costs were not reported

Other bias

High risk


Carli 2010 Methods


Participants

Total knee arthroplasty N = 40 M = 11, F = 29 Mean age 70.8 years (Group I) and 71.1 years (Group F) ASA Country: Canada Dates: from February 2007 to November 2008

Interventions

Femoral nerve block (Group F): first a loading dose of 8 ml of 0.2% ropivacaine, then a continuous infusion of 0.2% ropivacaine at a rate of 8 ml per hour for 48 hours, N = 20 Periarticular and intra-articular infiltration (Group I): 100 ml of 0.2% ropivacaine, 1 ml of 30 mg ml−1 ketorolac, and 0.5 ml of 1 mg ml−1 epinephrine, N = 20

Outcomes

Pain intensity at rest, on walking, and at greatest knee flexion at days 1 and 2 postoperatively Additional analgesic consumption Knee range of motion on day 1 and day 2 postoperatively Length of hospital stay

Notes

Oxford quality and validity scale = 5 Conflicts of interest: A.C. was a recipient of a fellowship from the Department of Anesthesia of the Sacred Heart Catholic University of Rome

Risk of bias Bias


Random sequence generation (selection Low risk bias) Peripheral nerve blocks for postoperative pain after major knee surgery (Review) Copyright © 2014 The Cochrane Collaboration. Published by John Wiley & Sons, Ltd.

Support for judgement Quote: “Computer-generated tables”

23

Carli 2010

(Continued)


Low risk

Quote: “Sealed brown envelopes”

Blinding of participants and personnel Low risk (performance bias) All outcomes

Quote: “The staff and the patients were not aware of the treatment group assignment.” Quote: “All patients received a femoral nerve catheter and a periarticular knee catheter.”


Comment: Assessors were blinded to the treatment group assignment


Low risk



Unclear risk

Comment: proportion of participants with ’no worse than mild pain’, adverse events, hospital costs, and participant satisfaction were not reported

Other bias

High risk


Chan 2012 Methods


Participants

Total knee arthroplasty N = 82 M = 22, F = 60 Mean age 68.1 years (pre-treatment group), 67.3 years (post-treatment group), 70.9 years (pre-control group), and 71.8 years (post-control group) ASA Country: China

Interventions

Pre-treatment group: femoral nerve block with 0.4 ml kg−1 of 0.375% bupivacaine plus 1:200,000 epinephrine after anesthesia but before the operation, N = 20 Post-treatment group: femoral nerve block with 0.4 ml kg−1 of 0.375% bupivacaine plus 1:200,000 epinephrine immediately after the operation, N = 21 Pre-control group: femoral nerve block with normal saline in the same volume as the tested drugs before the operation, N = 20 Post-control group:femoral nerve block with normal saline in the same volume as the tested drugs after the operation, N = 21

Outcomes

Pain intensity at rest and on movement at 2 hours, 4 hours, 6 hours, 24 hours, 48 hours, and 72 hours postoperatively Additional analgesic consumption and the time of 1st request for morphine Adverse events Knee range of motion on days 1 and 2 postoperatively


24

Chan 2012

(Continued)

Notes

Oxford quality and validity scale = 4 Funding: GHKS98-065, VGHKS97-084 from Kaohsiung Veterans General Hospital

Risk of bias Bias




Quote: “Patients were randomly assigned to four groups using computer-generated numbers enclosed in envelopes”


Quote: “Patients were randomly assigned to four groups using computer-generated numbers enclosed in envelopes”

Low risk


Quote: “Double-blinded” Quote: “All patients in the four groups received the femoral nerve block (true or sham)” Comment: The exact method of blinding the investigators was not described




Low risk

Comment: Reasons for the missing data were acceptable. The missing data did not seem to have potential impact on outcomes


Unclear risk

Comment: proportion of participants with ’no worse than mild pain’, length of hospital stay, hospital costs, and participant satisfaction were not reported

Other bias

High risk


Espelund 2013 Methods


Participants

Anterior cruciate ligament reconstruction N = 49 M = 38, F = 11 Mean age 33 years (Group control) and 28 years (Group adductor canal blockade (ACB) ) ASA -


25

Espelund 2013

(Continued)

Country: Denmark Dates: from June 2010 to March 2012 Interventions

Group ACB: 30 ml ropivacaine 7.5 mg ml−1 ,N = 25 Group control: 30 ml 0.9% saline, N = 24

Outcomes

Pain intensity at rest 0 - 24 hours postoperatively Pain intensity on movement 1 - 24 hours postoperatively Additional analgesic consumption Adverse events

Notes

Oxford quality and validity scale = 5 Funding: this trial was supported by the Department of Anaesthesiology, University of Copenhagen, Glostrup Hospital, Capital Region of Denmark

Risk of bias Bias




Quote: “Computer-generated block randomization list”


Low risk

Quote: “Computer-generated block randomisation list”


Quote: “No investigator, person treating or nursing the patients was aware of group assignment”


Quote: “No investigator, person treating or nursing the patients was aware of group assignment”


Low risk



Unclear risk

Comment: proportion of participants with ’no worse than mild pain’, knee range of motion, length of hospital stay, hospital costs, and participant satisfaction were not reported

Other bias

High risk



26

Ganapathy 1999 Methods


Participants

Total knee arthroplasty N = 62 M = 31, F = 31 Mean age 70 years (control group), 66 years (0.1% bupivacaine group) and 66 years (0. 2% bupivacaine group) ASA Country: Canada

Interventions

0.1% bupivacaine group: femoral 3-in-1 block with first a loading dose of 30 ml of 0. 1% bupivacaine, then a continuous infusion of 0.1% bupivacaine at a rate of 10 ml per hour for 48 hours, N = 20 0.2% bupivacaine group: femoral 3-in-1 block with first a loading dose of 30 ml of 0. 2% bupivacaine, then a continuous infusion of 0.2% bupivacaine at a rate of 10 ml per hour for 48 hours, N = 22 Control group: femoral 3-in-1 block with first a loading dose of 30 ml of saline, then a continuous infusion of saline at a rate of 10 ml per hour for 48 hours, N = 20

Outcomes

Pain intensity at rest and with activity at day of surgery and at days 1 and 2 postoperatively Additional analgesic consumption Adverse events Knee range of flexion on the day of surgery and day 1 postoperatively

Notes


Risk of bias Bias







Unclear risk




Comment: Assessors were blinded to the treatment group assignment


Low risk



Unclear risk

Comment: proportion of participants with ’no worse than mild pain’, length of hospital stay, hospital costs, and participant sat-


27

Ganapathy 1999

(Continued)

isfaction were not reported Other bias

High risk


Good 2007 Methods

Randomized, double-blind, 2 parallel groups.

Participants

Total knee arthroplasty N = 42 M = 26, F = 16 Mean age 61 years (experimental group) and 68 years (control group) ASA not described Country: the United States Dates: from December 1997 to March 1999

Interventions

Femoral nerve block (experimental group): 40 ml of 0.5% bupivacaine hydrochloride with epinephrine 1:200,000 before surgery, N = 22 Saline (control group): 40 ml solution of 0.9% normal saline before surgery, N = 20

Outcomes

Pain intensity on day 1, day 2, and day 3 postoperatively (assessed at rest or on movement is not described) Additional analgesic consumption Adverse events Knee range of motion on day 2 postoperatively

Notes


Risk of bias Bias




Quote: “randomized” Comment: not specifically described.



Unclear risk


Quote: “The orthopedic surgeon, anesthesiologist, physical therapists, clinical nursing staff, and patients were blinded to the treatment.”


Quote: “The orthopedic surgeon, anesthesiologist, physical therapists, clinical nursing staff, and patients were blinded to the treatment.”


28

Good 2007

(Continued)


Low risk

Comment: The study had 3 participants in the nerve block group and one in placebo group discharged early because of increased recovery rate, which happened after 72 hours postoperatively, thus may not have potential impact on outcomes


Unclear risk


Other bias

High risk


Hirst 1996 Methods


Participants

Total knee arthroplasty N = 33 M = 10, F = 23 Mean age 66.9 years (Group 1), 71.2 years (Group 2) and 70.1 years (Group 3) ASA Country: Canada

Interventions

Group 1: single-injection femoral 3-in-1 nerve block with 20 ml of 0.5% bupivacaine with 1:200,000 epinephrine, N = 11 Group 2: continuous femoral 3-in-1 nerve block, first a loading dose of 20 ml of 0. 5% bupivacaine with 1:200,000 epinephrine, then a continuous infusion of 0.125% bupivacaine at a rate of 6 ml per hour for 48 hours, N = 11 Group 3: a ’mock’ femoral 3-in-1 nerve block, N = 11

Outcomes

Pain intensity at rest and on motion at 12 hours, 24 hours, 48 hours, 72 hours postoperatively Additional analgesic consumption Adverse events Participant satisfaction

Notes


Risk of bias Bias




Support for judgement Quote: “Randomized” Comment: not specifically described.

29

Hirst 1996

(Continued)


Unclear risk



Quote: “Double-blinded” Comment: Control group had a ’mock’ femoral 3-in-1 nerve block. The exact method to blind the investigators was not described




Low risk



Unclear risk


Other bias

High risk


Ilfeld 2010 Methods

Randomized, double-blind, two parallel groups.

Participants

Total knee arthroplasty N = 77 M = 26, F = 51 Mean age 61 years (ropivacaine group) and 66 years (placebo group) ASA Country: the United States

Interventions

Ropivacaine group: continuous femoral nerve block with 0.2% ropivacaine, N = 39 Placebo group: continuous femoral nerve block with normal saline, N = 38

Outcomes

Pain intensity at 24 hours, 48 hours and 72 hours postoperatively Additional analgesic consumption Adverse events Knee range of motion on days 1 and 2 postoperatively

Notes

Oxford quality and validity scale = 5 Funding: the National Institutes of Health Grant GM077026 from the National Institute of General Medical Sciences (Bethesda, Maryland); National Institutes of Health Grants RR00082, RR000827, and RR025208 from the National Center for Research Resources (Bethesda, Maryland); the Department of Anesthesiology, University of California San Diego (San Diego, California); Stryker Instruments (Kalamazoo, Michigan); and Teleflex


30

Ilfeld 2010

(Continued)

Medical (Research Triangle Park, North Carolina) Conflicts of interest: Teleflex Medical and Stryker Instruments provided funding and donated portable infusion pumps for the investigation, these two companies had absolutely no input into any aspect of study conceptualization, design, and implementation; data collection, analysis, and interpretation; or manuscript preparation. The funding from these companies was used by each institution to help defray research coordinator expenses. Drs. Mariano and Loland conduct continuous peripheral nerve block workshops for Stryker Instruments Risk of bias Bias




Quote: “Subjects were randomized to groups stratified by institution/hospital using computer-generated tables and provided to investigational pharmacists via the PAINfRE.com Web site”


Low risk

Quote: “Subjects were randomized to groups stratified by institution/hospital using computer-generated tables and provided to investigational pharmacists via the PAINfRE.com Web site”


Quote: “Investigators, subjects, and all clinical staff were masked to treatment group assignment.”


Quote: “Investigators, subjects, and all clinical staff were masked to treatment group assignment.”


Low risk



Unclear risk


Other bias

High risk



31

Jenstrup 2012 Methods


Participants

Total knee arthroplasty N = 71 M = 37, F = 34 Mean age 67 years ASA Country: Denmark Dates: from August 2010 to March 2011

Interventions

Adductor canal block: 0.75% ropivacaine, N = 34 Placebo: saline, N = 37

Outcomes

Pain intensity at rest at 2 hours, 4 hours, 8 hours, and 24 hours postoperatively Pain intensity on movement at 2 hours, 4 hours, 8 hours, and 24 hours postoperatively Additional analgesic consumption Adverse events

Notes

Oxford quality and validity scale = 5 Funding: provided solely from institutional and departmental sources

Risk of bias Bias




Quote: “a computer generated block randomization list”


Low risk

Quote: “a computer generated block randomization list”


Quote: “All investigators, staff, and patients were blinded to the treatment groups. ”


Quote: “All investigators, staff, and patients were blinded to the treatment groups. ”


Low risk

Comment: Reasons for missing data were acceptable. The missing data did not seem to have potential impact on outcomes


Unclear risk



32

Jenstrup 2012

(Continued)

Other bias

High risk


Lundblad 2011 Methods


Participants

Anterior cruciate ligament reconstruction N = 62 M = 31, F = 31 Mean age 30.8 years (Group control) and 25.9 years (Group infrapatellar nerve block (IPNB)) ASA Country: Sweden

Interventions

Group IPNB: 10 ml of 0.5% levo bupivacaine, N = 30 Group control: 10 ml saline, N = 32

Outcomes

Pain intensity at rest and on movement 1 - 24 hours postoperatively Proportion of participants with ’no worse than mild pain’

Notes

Oxford quality and validity scale = 5 Funding: the study was solely supported by departmental (Karolinska University Hospital) and local governmental healthcare (Stockholms läns landsting) funds

Risk of bias Bias




Quote: “Computer-generated numbers”


Low risk

Comment: the participants and investigators could not foresee assignment


Comment: the participants and investigators were blinded to the treatment


Comment: the assessors were blinded to the treatment.



Low risk


random

33

Lundblad 2011

(Continued)


Unclear risk

Comment: additional analgesic consumption, knee range of motion, adverse events, length of hospital stay, hospital costs, and participant satisfaction were not reported

Other bias

High risk


Matava 2009 Methods


Participants

Anterior cruciate ligament reconstruction N = 56 M = 35, F = 21 Mean age 27 years (Femoral nerve block (Group FNB)) and 26.5 years (Group control) ASA Country: the United States

Interventions

Group FNB: 0.5% bupivacaine with 1:200,000 epinephrine, N = 31 Group control: saline, N = 25

Outcomes

Pain intensity at baseline, 2 hours, 4 hours, 8 hours, 12 hours, 24 hours, 48 hours, and 72 hours postoperatively Additional analgesic consumption Adverse events Length of hospital stay Participant satisfaction scores Total charges

Notes

Oxford quality and validity scale = 5 Funding: the study was funded by a BJC Innovations in Health Care Program grant

Risk of bias Bias




Quote: “Computer-generated numbers”


Quote: “Sealed study envelope”

Low risk

random


Quote: “The surgeons and patients were blinded to group assignment.”

Blinding of outcome assessment (detection Low risk bias)

Comment: assessors were blinded to the treatment.


34

Matava 2009

(Continued)

All outcomes Incomplete outcome data (attrition bias) All outcomes

Low risk



Unclear risk

Comment: proportion of patients with ’no worse than mild pain’, knee range of motion were not reported

Other bias

High risk


Moghtadaei 2014 Methods


Participants

Total knee arthroplasty N = 36 M = 25, F = 11 Mean age 67.4 years (Single femoral nerve block (Group F) and 64 years (Intra-periarticular infiltration (Group I)) ASA Country: IR Iran Dates: from January to June 2013

Interventions

Group F: 20 ml of 10mg ml−1 ropivacaine, N = 18 Group I: a combination of 300 mg ropivacaine, 30 mg ketorolac and 0.5 mg epinephrine diluted to a volume of 150 ml, N = 18

Outcomes

Pain intensity at 6 hours, 12 hours, and 24 hours postoperatively Additional analgesic consumption Knee range of motion Time to walk Length of hospital stay Participant satisfaction scores

Notes

Oxford quality and validity scale = 5 Funding: Iran University of Medical Sciences Thesis grants

Risk of bias Bias




Quote: “Computerized blocked random number list”


Quote: “Computerized blocked random number list”

Low risk


35

Moghtadaei 2014

(Continued)


Quote: “Patients, researches, physiotherapists and all nursing staffs were blinded to the grouping.”


Quote: “Patients, researches, physiotherapists and all nursing staffs were blinded to the grouping.”


Low risk



Unclear risk

Comment: proportion of participants with ’no worse than mild pain’, adverse events, and hospital costs were not reported

Other bias

High risk


Mulroy 2001 Methods


Participants

Anterior cruciate ligament reconstruction N = 53 M = 38, F = 15 Mean age 38 years (sham group), 36 years (0.25% bupivacaine group) and 38 years (0. 5% bupivacaine group) ASA Country: the United States

Interventions

0.25% bupivacaine group: femoral nerve block with 25 ml 0.25% bupivacaine and 1: 200,000 epinephrine, N = 21 0.5% bupivacaine group: femoral nerve block with 25 ml 0.5% bupivacaine and 1:200, 000 epinephrine, N = 20 Sham group: a sham block, N = 12

Outcomes

Pain intensity on the day of surgery, days 1 and 2 postoperatively Additional analgesic consumption Adverse events

Notes


Risk of bias Bias




36

Mulroy 2001

(Continued)


Quote: “ randomized by sealed envelopes”


Quote: “ randomized by sealed envelopes”

Low risk




Quote: “evaluated by a blinded observer”


Low risk

Comment: 2 participants in the block group were excluded for failure of the block. The missing data did not seem to have potential impact on outcomes


Unclear risk


Other bias

High risk


Ng 2001 Methods


Participants

Total knee arthroplasty N = 48 M = 7, F = 41 Mean age 65.3 years (Group C), 64.3 years (Group R1), 64.8 years (Group R2), and 62. 8 years (Group B) ASA Country: Singapore

Interventions

Group C: 3-in-1 femoral nerve block with 30 ml of saline, N = 12 Group R1: 3-in-1 femoral nerve block with 30 ml of ropivacaine 0.25% (75 mg), N = 12 Group R2: 3-in-1 femoral nerve block with 30 ml of ropivacaine 0.5% (150 mg), N = 12 Group B: 3-in-1 femoral nerve block with 30 ml of bupivacaine 0.25% (75 mg), N = 12

Outcomes

Pain intensity at 1 hour, 4 hours, 8 hours, 24 hours, and 48 hours postoperatively Additional analgesic consumption Adverse events


37

Ng 2001

(Continued)

Length of hospital stay Notes


Risk of bias Bias




Quote: “Sealed envelopes”.


Quote: “Sealed envelopes”.

Low risk


Quote: “Double-blinded”. Comment: Control group had a sham block. The exact method of blinding the investigators was not described


Quote: “Investigator was blinded to the patient’s grouping.”


Low risk



Unclear risk

Comment: proportion of participants with ’no worse than mild pain’, knee range of motion, hospital costs, and participant satisfaction were not reported

Other bias

High risk


Schultz 1991 Methods


Participants

Open knee surgery N = 22 M = 13, F = 9 Median age 25 years (lumbar plexus block (Group LPB)) and 30 years (epidural analgesia (Group EPI)) ASA Country: Denmark

Interventions

3-in-1 Group LPB: first a bolus dose of 2 mg kg-1 (5 mg ml−1 ) bupivacaine, then 0.14 ml kg−1 per hour of 2.5 mg ml−1 bupivacaine, N = 11 Group EPI: first a bolus dose of 10 ml (0.4 mg ml−1 ) morphine, then 8 mg morphine over 24 hours, N = 11


38

Schultz 1991

(Continued)

Outcomes

Pain intensity within 18 hours postoperatively Additional analgesic consumption Adverse events

Notes


Risk of bias Bias




Quote: “Randomized” Comment: not specifically described.



Unclear risk


Quote: “Double-blinded” “Pain treatment was given as either bupivacaine in the femoral catheter or morphine in the epidural catheter, with saline in the other catheter.” “The patients and the personnel were unaware of the pain treatment given”


Quote:“ Double-blinded” “Pain treatment was given as either bupivacaine in the femoral catheter or morphine in the epidural catheter, with saline in the other catheter.” “The patients and the personnel were unaware of the pain treatment given”


Low risk



Unclear risk


Other bias

High risk


Szczukowski 2004 Methods


Participants

Total knee arthroplasty N = 40 M = 15, F = 25 Mean age 68 years (femoral nerve block (Group FNB)) and 66 years (Group Control)


39

Szczukowski 2004

(Continued)

ASA Country: the United States Interventions

Group FNB: a single-injection femoral nerve block with 30 ml of 0.5% bupivacaine with epinephrine 1:200,000, N = 19 Group Control: a single-injection femoral nerve block with saline, N = 21

Outcomes

Pain intensity on the day of surgery, days 1 and 2 postoperatively Additional analgesic consumption Adverse events Knee range of motion on days 1 and 2 postoperatively Length of hospital stay

Notes


Risk of bias Bias




Quote: “Patients were randomized using the research randomizer application”


Low risk



Quote: “the surgeons, anesthesiologists, nurses, and all operating room staff were blinded to the contents of the vials.”


Quote: “the surgeons, anesthesiologists, nurses, and all operating room staff were blinded to the contents of the vials.”


Low risk



Unclear risk

Comment: proportion of participants with ’no worse than mild pain’, hospital costs, and participant satisfaction were not reported

Other bias

High risk



40

Tierney 1987 Methods


Participants

Open knee surgery N = 20 Sex ratio not described Mean age not described ASA Country: Canada

Interventions

Group FNB: femoral nerve block with 0.25% bupivacaine, N = 10 Group Control: femoral nerve block with saline, N = 10

Outcomes

Additional analgesic consumption and median time to remedication

Notes


Risk of bias Bias




Quote: “The hospital pharmacy prepared twenty 20ml ampoules, ten containing bupivacaine 0.25 percent plain and ten normal saline.”


Low risk



Quote: “The anesthetist and other staff, including the postoperative nurses, were blinded as to their contents.”


Quote: “The anesthetist and other staff, including the postoperative nurses, were blinded as to their contents.”


Low risk



Unclear risk

Comment: pain intensity,proportion of participants with ’no worse than mild pain’, knee range of motion, adverse events, hospital costs, and participant satisfaction were not reported

Other bias

High risk



41

Wang 2002 Methods


Participants

Total knee arthroplasty N = 30 M = 11, F = 19 Mean age 66 years (Group control) and 67 years (femoral nerve block (Group FNB)) ASA Country: the United States

Interventions

Group FNB: 40 ml of 0.25% bupivacaine, N = 15 Group control: 40 ml saline, N = 15

Outcomes

Pain intensity at rest and on movement at day 1, day 2, and day 3 postoperatively Additional analgesic consumption Knee range of flexion on day 2 postoperatively Adverse events Length of hospital stay

Notes


Risk of bias Bias




Quote: “randomly assigned”. Comment: not specifically described.



Unclear risk


Comment: all study personnel and participants were blinded to the treatment group assignments


Comment: assessors were blinded to the treatment group assignments


Low risk



Unclear risk

Comment: proportion of participants with ’no worse than mild pain’, hospital costs, and participant satisfaction were not reported

Other bias

High risk



42

Widmer 2012 Methods


Participants

Total knee arthroplasty N = 54 M = 30, F = 24 mean age 72.1 years (Group F) and 69.4 years (Group L) ASA not described Country: Australia

Interventions

Group F: femoral nerve block with 30 ml of 0.375% ropivacaine and periarticular anesthetic infiltration with 100 ml of 0.2% ropivacaine, N = 27 Group L: periarticular anesthetic infiltration with 100 ml of 0.2% ropivacaine, N = 27

Outcomes

Pain intensity at rest in the first 24 hours postoperatively Postoperative fentanyl requirement in the first 24 hours postoperatively Adverse events

Notes


Risk of bias Bias




Quote: “Randomized by coded envelope”


Low risk

Quote: “Randomized by coded envelope”


Quote: “Both the investigators and patients were blinded to treatment modality.”




Unclear risk

Comment: 55 participants met the inclusion criteria, and the study claimed no dropout in either group, but only 54 participants were analyzed


Unclear risk


Other bias

High risk



43

Williams 2006 Methods


Participants

Anterior cruciate ligament reconstruction N = 233 M = 137, F = 96 Mean age 28 years ASA Country: the United States

Interventions

Femoral nerve block, levo bupivacaine (0.25%) bolus with saline infusion, N = 79 Femoral nerve block, levo bupivacaine (0.25%) bolus and infusion, N = 76 Placebo group: femoral nerve block, 30 ml saline bolus plus saline infusion (270 ml at 5 ml h−1 ), N = 78

Outcomes

Pain intensity on movement at day 1, day 2, day 3 and day 4 postoperatively Proportion of participants with ’no worse than mild pain’ on postoperative days 1, 2, 3, 4 Additional analgesic consumption Adverse events

Notes


Risk of bias Bias




Quote: “Random sequence were generated by a computer program.”


Quete: “Sealed envelopes”.

Low risk


Comment: personnel and participants were blinded to the treatment group assignment




Low risk



Unclear risk

Comment: knee range of motion, length of hospital stay, hospital costs, and participant satisfaction were not reported

Other bias

Unclear risk

Comment: unclear risk of bias for size of study.


44

Wulf 2010 Methods


Participants

Anterior cruciate ligament reconstruction N = 280 Sex ratio not described Mean age 31 years (Group bupivacaine (BU) 2.5), 33 years (Group ropivacaine (RO) 2. 0), 33 years (Group RO 7.5), and 30 years (Group saline (NaCl)) ASA Country: Germany

Interventions

Group BU 2.5: single-shot femoral nerve block with 30 ml of 0.25% bupivacaine, N = 80 Group RO 2.0: single-shot femoral nerve block with 30 ml of 0.2% ropivacaine, N = 80 Group RO 7.5: single-shot femoral nerve block with 30 ml of 0.75% ropivacaine, N = 80 Group NaCl: single-shot femoral nerve block with 30 ml of 0.9% NaCl, N = 80

Outcomes

Pain intensity at 1 hour, 2 hours, 4 hours, 8 hours, 10 hours, 12 hours and on day 1 postoperatively Additional analgesic consumption and time to 1st request for escape medication Adverse events

Notes

Oxford quality and validity scale = 3 Funding: this study was an investigator initiated trial funded by institutional funding sources Conflicts of interest: H. W. has received an honorarium for lectures and advisory board activity and received funding for clinical studies from AstraZeneca, Teleflex (former arrow), B. Braun and others

Risk of bias Bias




Quote: “computer-generated randomization list”


Comment: not described

Unclear risk




Comment: not described



Low risk


45

Wulf 2010

(Continued)


Unclear risk


Other bias

Unclear risk

Comment: unclear risk of bias for size of study.

Xie 2012 Methods


Participants

Total knee arthroplasty N = 105 M = 29, F = 76 Mean age 72.6 years (Control group), 68.1 (Low-dose group), and 68.8 years (Highdose group) ASA Country: the United States

Interventions

Low-dose group: 3-in-1 block with 30 ml 0.25% bupivacaine and 1:200,000 epinephrine, N = 35 High-dose group: 3-in-1 block with 30 ml 0.50% bupivacaine and 1:200,000 epinephrine, N = 35 Control group: 3-in-1 block with 30 ml saline, N = 35

Outcomes

Pain intensity at rest in the PACU Additional analgesic consumption Adverse events

Notes


Risk of bias Bias




Quote: “randomized” Comment: not specifically described.



Unclear risk



The anesthesiologists, surgeons and participants were blinded to the treatment assignment

46

Xie 2012

(Continued)


Comment: blinding of outcome assessment ensured.


Low risk



Unclear risk


Other bias

High risk


ASA: American Society of Anesthesiologists score F: female M: male N: number PACU: post-anesthesia care unit

Characteristics of excluded studies [ordered by study ID]

Study

Reason for exclusion

Bogoch 2002

Surgeries in this study included total hip arthroplasty and total knee arthroplasty. The outcomes of participants undergoing total knee arthroplasty were not presented separately

De Lima 2008

The study claimed to be double-blind. However, sham block was not performed in the control group, and the measures to maintain participant blinding seemed inadequate

Frassanito 2010

The study claimed to be single-blind. However, sham block was not performed in the spinal analgesia group, and the measures to maintain participant blinding seemed inadequate

Hunt 2009

Participants in the intervention group received both femoral nerve block and sciatic nerve block, while participants in the control group only received sham femoral nerve block. It may affect participant blinding


47

Characteristics of studies awaiting assessment [ordered by study ID] Tugay 2006 Methods


Participants

Total knee arthroplasty N = 23

Interventions

Group I received pre-emptive single injection FNB Group II received postoperative single injection FNB Group III served as a control group

Outcomes

Pain intensity at the 15th minute, 30th minute, 1st, 4th, 6th, 12th, 24th, and 48th hours postoperatively

Notes

Presently only available in abstract form

Wongswadiwat 2012 Methods


Participants

Arthroscopic anterior cruciate ligament reconstruction N = 47

Interventions

The participants received either 0.2 % bupivacaine with adrenaline or 0.9% NSS 40 ml

Outcomes

Morphine consumption, time to 1st rescue analgesia, pain intensity, and complication within 24 hours

Notes

Presently only available in abstract form

FNB: femoral nerve block N: number NSS: normal saline solution

Characteristics of ongoing studies [ordered by study ID] NCT01616836 Trial name or title

Optimizing Pain and Rehabilitation After knee arthroplasty (OPRA)

Methods


Participants

Total knee arthroplasty 18 - 70 years old Both genders ASA -


48

NCT01616836

(Continued)

Interventions

Active comparator: bolus femoral nerve block (ropivacaine 0.5% 20 mL), continuous infusion 48 hours (ropivacaine 0.2%, 5 mL h−1 ), placebo local infiltration Active comparator: femoral nerve block (ropivacaine 0.5% 20 mL), placebo femoral nerve block infusion, placebo local infiltration Active comparator: placebo fascia iliac block, placebo fascia iliaca infusion, local infiltration analgesia

Outcomes

Pain intensity at 9:00 on postoperative day 2 Opioid consumption within 4 days postoperatively Knee range of motion within 4 days postoperatively

Starting date

June 7, 2012

Contact information

Colin J McCartney, MBChB 416-480-4864 [email protected] Stephen Choi, MD 416-480-4864 [email protected]

Notes

Identified by searching clinicaltrials.gov; identifier: NCT01616836

NCT01683071 Trial name or title

Femoral nerve block with liposome bupivacaine for postsurgical analgesia following total knee arthroplasty

Methods


Participants

Total knee arthroplasty 18 - 70 years old Both genders ASA -

Interventions

Active comparator: Group 1 liposome bupivacaine injectable suspension 67 mg Active comparator: Group 2 liposome bupivacaine injectable suspension 133 mg Active comparator: Group 3 liposome bupivacaine injectable suspension 266 mg Placebo comparator: Group 4 Placebo (preservative-free normal saline)

Outcomes

Pain intensity within 72 hours postoperatively

Starting date

September 7, 2012

Contact information

Erol Onel, MD, Pacira Pharmaceuticals Inc

Notes



49

NCT02008617 Trial name or title

Analgesic benefits of genicular nerve blocks of the posterior knee for patients undergoing anterior cruciate ligament (ACL) reconstruction

Methods


Participants

Anterior cruciate ligament reconstruction 18 - 70 years old Both genders ASA -

Interventions

Active comparator: ultrasound guided posterior genicular nerve infiltration with 30 mL of bupivacaine 0. 20% with epinephrine 1:300,000 Sham comparator: ultrasound-guided posterior genicular nerve infiltration posterior knee with 30 mL of preservative-free normal saline

Outcomes

Opioid consumption within 24 hours postoperatively Pain intensity within 24 hours postoperatively Participant satisfaction within 24 hours postoperatively

Starting date

December 6, 2013

Contact information

Rohit Rahangdale, MD, 312-695-2528 [email protected]

Notes


ASA: American Society of Anesthesiologists score N: number,


50

DATA AND ANALYSES

Comparison 1. Peripheral nerve blocks adjunctive to systemic analgesia versus systemic analgesia alone

No. of studies

No. of participants

1 Pain intensity at rest (0 - 23 hours postoperatively) 1.1 Total knee arthroplasty

7

390

Mean Difference (IV, Random, 95% CI)

3

183


1.2 Anterior cruciate ligament reconstruction 1.3 Partial menisectomy arthroscopically 2 Pain intensity at rest (24 - 47 hours postoperatively) 2.1 Total knee arthroplasty

3

167


1

40


6

320


4

224


2.2 Anterior cruciate ligament reconstruction 2.3 Partial menisectomy arthroscopically 3 Pain intensity at rest (48 - 72 hours postoperatively) 3.1 Total knee arthroplasty

1

56


1

40


4

210


3

154


3.2 Anterior cruciate ligament reconstruction 4 Pain intensity on movement (0 23 hours postoperatively) 4.1 Total knee arthroplasty

1

56


-13.73 [-18.21, -9. 26] -0.5 [-10.60, 9.60]

5

304


-6.95 [-15.92, 2.01]

2

153


4.2 Anterior cruciate ligament reconstruction 4.3 Partial menisectomy arthroscopically 5 Pain intensity on movement (24 - 47 hours postoperatively) 6 Pain intensity on movement (48 - 72 hours postoperatively)

2

111


-12.93 [-26.62, 0. 75] 0.10 [-7.49, 7.70]

1

40


-7.5 [-19.91, 4.91]

3

182


2

112

Mean Difference (IV, Fixed, 95% CI)

-8.87 [-27.77, 10. 03] -6.19 [-11.76, -0.62]

Outcome or subgroup title

Statistical method


Effect size -11.85 [-20.45, -3. 25] -19.50 [-35.01, -3. 99] -2.82 [-8.03, 2.38] -15.60 [-22.84, -8. 36] -12.92 [-19.82, -6. 02] -15.05 [-24.11, -5. 98] -0.10 [-11.59, 11. 39] -15.50 [-22.27, -8. 73] -9.72 [-16.75, -2.70]

51

Analysis 1.1. Comparison 1 Peripheral nerve blocks adjunctive to systemic analgesia versus systemic analgesia alone, Outcome 1 Pain intensity at rest (0 - 23 hours postoperatively). Review:

Peripheral nerve blocks for postoperative pain after major knee surgery

Comparison: 1 Peripheral nerve blocks adjunctive to systemic analgesia versus systemic analgesia alone Outcome: 1 Pain intensity at rest (0 - 23 hours postoperatively)

Study or subgroup

Peripheral nerve blocks

Mean Difference

Systemic analgesia

Weight

IV,Random,95% CI

Mean Difference

N

Mean(SD)

N

Mean(SD)

IV,Random,95% CI

Chan 2012

41

18 (11)

41

40 (14)

16.3 %

-22.00 [ -27.45, -16.55 ]

Jenstrup 2012

34

38.5 (25)

37

41.3 (25.2)

13.2 %

-2.80 [ -14.49, 8.89 ]

Wang 2002

15

42 (29)

15

78 (16)

10.5 %

-36.00 [ -52.76, -19.24 ]

1 Total knee arthroplasty

Subtotal (95% CI)

90

93

40.0 % -19.50 [ -35.01, -3.99 ]

Heterogeneity: Tau2 = 152.78; Chi2 = 12.34, df = 2 (P = 0.002); I2 =84% Test for overall effect: Z = 2.46 (P = 0.014) 2 Anterior cruciate ligament reconstruction Espelund 2013

25

11.4 (10.4)

24

13.1 (7.9)

16.4 %

-1.70 [ -6.86, 3.46 ]

Lundblad 2011

30

10.4 (18.7)

32

10 (17.5)

14.6 %

0.40 [ -8.63, 9.43 ]

Matava 2009

31

25.7 (20.4)

25

36.7 (22.4)

13.4 %

-11.00 [ -22.34, 0.34 ]

44.5 %

-2.82 [ -8.03, 2.38 ]

15.5 %

-15.60 [ -22.84, -8.36 ]

Subtotal (95% CI)

86

81

Heterogeneity: Tau2 = 5.71; Chi2 = 2.65, df = 2 (P = 0.27); I2 =24% Test for overall effect: Z = 1.06 (P = 0.29) 3 Partial menisectomy arthroscopically Akkaya 2008

Subtotal (95% CI)

20

3.4 (1.9)

20

20

19 (16.4)

20

15.5 % -15.60 [ -22.84, -8.36 ]

194

100.0 % -11.85 [ -20.45, -3.25 ]

Heterogeneity: not applicable Test for overall effect: Z = 4.23 (P = 0.000024)

Total (95% CI)

196

Heterogeneity: Tau2 = 110.31; Chi2 = 46.25, df = 6 (P

Femoral nerve blocks for acute postoperative pain after knee replacement surgery.

Peripheral nerve blocks for ambulatory surgery.

Ultrasound-Guided Percutaneous Peripheral Nerve Stimulation for Postoperative Analgesia: Could Neurostimulation Replace Continuous Peripheral Nerve Blocks?

Inferior alveolar nerve blocks for postoperative pain control after mandibular distraction with osteotomies in a neonate.

Perioperative pain control after total knee arthroplasty: An evidence based review of the role of peripheral nerve blocks.

Effect of saphenous nerve block for postoperative pain on knee surgery: a meta-analysis.

Peripheral nerve blocks versus general anesthesia for total knee replacement in elderly patients on the postoperative quality of recovery.

Inpatient falls after total knee arthroplasty: the role of anesthesia type and peripheral nerve blocks.

Continuous low-dose 3-in-1 nerve blockade for postoperative pain relief after total knee replacement.

Dose-finding methodology for peripheral nerve blocks.

Continuous interscalene brachial plexus block versus parenteral analgesia for postoperative pain relief after major shoulder surgery.

Comparison of ketorolac and morphine for postoperative pain after major surgery.

Effects of Interscalene Nerve Block for Postoperative Pain Management in Patients after Shoulder Surgery.

Comparison of Transcutaneous Electrical Nerve Stimulation and Parasternal Block for Postoperative Pain Management after Cardiac Surgery.

Postoperative pain after mandibular third-molar surgery.

Nerve conduits for peripheral nerve surgery.

Continuous transversus abdominis plane (TAP) blocks for postoperative pain control after hernia surgery: a randomized, triple-masked, placebo-controlled study.

Postoperative pain following hospital discharge after knee replacement surgery: a patient survey.

Perioperative pregabalin for attenuation of postoperative pain after eyelid surgery.

Opioid therapy for knee osteoarthritis and postoperative persistent pain after knee arthroplasty.

Analgesia for pelvic limb surgery. A review of peripheral nerve blocks and the extradural technique.

Use of transcutaneous electrical nerve stimulation in the control of postoperative chest pain after cardiac surgery.

Relieving Pain After Arthroscopic Knee Surgery: Ultrasound-Guided Femoral Nerve Block or Adductor Canal Block?

Transcutaneous electrical nerve stimulation for postoperative pain control after total knee arthroplasty: A meta-analysis of randomized controlled trials.