Description of the condition
The sensation of pain is one of the vital functions of the human body's nervous system that enables protection of an injured area while healing and repair occur. All surgical procedures are followed by pain, which may trigger or amplify endocrine and metabolic responses, autonomic reflexes, nausea, ileus and muscle spasm to increase postoperative morbidity and mortality (Desborough 2000). The current armamentarium of analgesic drugs and techniques for the management of postoperative pain continues to grow at a rapid rate. However, effective treatment of acute postsurgical pain still poses unique challenges for practitioners (White 2005). Major shoulder surgery is often associated with severe postoperative pain, greater than after gastrectomy or thoracotomy, especially within the first 48 hours. One of the characteristics of this pain is its dynamic component (exacerbation on movement); therefore moderate pain at rest can become severe during rehabilitation. Optimal postoperative pain treatment is, therefore, helpful in enabling early mobilization and rehabilitation, enhancing recovery and reducing morbidity (Borgeat 2007).
Description of the intervention
Different options are available for the management of postoperative pain after major shoulder surgery. These include enteral or parenteral administration of opioid and non-opioid (including non-steroidal anti-inflammatory drugs) analgesics, infiltration of local anaesthetics into the joint (intra-articular) or around the joint (bursal or sub-bursal) and use of regional analgesia (Borgeat 2007), including interscalene brachial plexus block (ISBPB), through different techniques. Patient-controlled analgesia (PCA) is a technique that allows patients to have autonomy in controlling their pain by timing the administration of small doses of an analgesic or local anaesthetic drug. PCA can be used with both parenteral and regional analgesic techniques (Macintyre 2001).
How the intervention might work
Traditionally, parenteral analgesia, which is used for postoperative pain relief, includes intravenous, transdermal, subcutaneous and intramuscular routes for administration of opioids and non-opioids, depending on the type of surgery. Opioid analgesics are commonly associated with opioid-related adverse effects such as nausea and vomiting, pruritus, sleep disturbance and constipation. Interscalene block provides effective anaesthesia and analgesia for shoulder surgery (Fredrickson 2010a).
Winnie was the first to describe the single-injection technique of interscalene brachial plexus block (Winnie 1970). ISBPB, alone or in combination with general anaesthesia, is a very suitable technique for shoulder surgery. The catheter technique for ISBPB, which was first described by Touminen in 1987 (Tuominen 1987), offers many advantages over single-injection ISBPB. These include early and aggressive mobilization, extension of analgesia for prolonged periods and use of smaller doses of local anaesthetics to prevent motor block and reduce opioid requirements and associated side effects (Bishop 2005; Fredrickson 2010a; Russon 2006).
However, ISBPB is an invasive procedure that may lead to serious complications. In addition to commonly associated risks, such as failure, nerve injury, vascular injury, bleeding, phrenic nerve block and Horner's syndrome, interscalene block is associated with risks of pleural puncture, central neuraxial needle placement, cervical spinal cord damage and permanent paralysis (Benumof 2000). The introduction of ultrasound to locate the brachial plexus has helped to reduce these adverse events.
Why it is important to do this review
This review will attempt to distinguish between the effectiveness of continuous ISBPB and that of parenteral opioid in providing better analgesia with minimal adverse effects in patients undergoing major shoulder surgery. This will help anaesthetists, as well as patients, to choose from the two techniques to obtain better pain control with fewer side effects after major shoulder surgery.
The objective of this review was to compare the analgesic efficacy of continuous interscalene brachial plexus block (ISBPB) with parenteral opioid analgesia for pain relief after major shoulder surgery.
Criteria for considering studies for this review
Types of studies
We included randomized controlled trials (RCTs) that compared continuous ISBPB with any form of parenteral opioid analgesia after major shoulder surgery.
Types of participants
We included adult patients (aged over 18 years) undergoing elective major shoulder surgery (shoulder arthroplasty or total shoulder replacement, open rotator cuff repair and internal fixation of shoulder fractures) who received either continuous ISBPB or any form of parenteral opioid analgesia for postoperative pain relief.
We excluded patients undergoing arthroscopic procedures, those who had undergone previous shoulder surgery and those coming in for daycare surgery.
Types of interventions
We included any RCT that compared continuous ISBPB with any form of parenteral opioid analgesia after major shoulder surgery with a minimum duration of follow-up of 12 hours postoperatively ( Table 1).
We included studies with:
- any method of localization of the brachial plexus by an interscalene approach.
- any local anaesthetic, in any concentration with or without any other drug(s), given by the ISBPB technique.
- any form(s) of parenteral (intravenous, intramuscular, subcutaneous) opioid analgesia technique.
- any parenteral opioid analgesic drug(s), given in any dosage.
Types of outcome measures
- The effectiveness of pain relief after major shoulder surgery using continuous ISBPB or parenteral opioid analgesia for the entire follow-up period (minimum 12 hours).
- Complications related to ISBPB including those related to technique (failure, pneumothorax, nerve injury, Horner's syndrome, vascular injury, haematoma, bleeding, catheter dislocation, infection, facial nerve palsy) and those related to drugs (nausea, vomiting, pruritus, metallic taste, motor block, sedation, respiratory depression, central nervous system excitation).
- Complications related to parenteral analgesia (nausea, vomiting, pruritus, sedation, respiratory depression).
- Effectiveness of pain relief at mobilization.
- Supplemental analgesia, if used.
- Time to mobilization.
- Participant satisfaction.
- Length of stay in postanaesthesia care unit (PACU).
- Length of stay in hospital.
Search methods for identification of studies
We searched the following databases.
- The Cochrane Central Register of Controlled Trials (CENTRAL) (2012, Issue 12) (Appendix 1).
- MEDLINE (via PubMed) (1950 to December 2012) (Appendix 2).
- EMBASE (1980 to December 2012) (Appendix 3).
- ISI Web of Science (1954 to December 2012) (Appendix 4).
- CINAHL (EBSCO host) (1982 to December 2012) (Appendix 5).
We imposed no language restrictions.
Searching other resources
We included all relevant studies irrespective of the language of publication. We manually checked the reference lists of relevant studies to identify trials missed by the electronic search strategy.
We contacted primary authors of identified trials to ask for more information, if required.
We searched for ongoing trials on the following websites.
Data collection and analysis
Selection of studies
We (HU and KS) independently reviewed the titles and abstracts identified by the searches. We obtained full copies of potentially relevant trials and assessed all full copies according to the parameters outlined in Criteria for considering studies for this review. We assessed only trials meeting these criteria for methodological quality. No disagreement arose during this process; therefore we did not consult the third review author (FK).
Data extraction and management
Two review authors (HU and KS) independently extracted data using a data extraction form (Appendix 6) modified from one developed by the Cochrane Anaesthesia Review Group (CARG). We resolved discrepancies by discussion. We extracted data (as far as was possible) on the basis of an intention-to-treat (ITT) analysis. We contacted primary investigators to ask for missing data. Two review authors (HU and KS) independently entered all data into Review Manager (RevMan 5.1).
Assessment of risk of bias in included studies
We assessed trial quality using the criteria outlined in the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2011). We observed whether inclusion and exclusion criteria were clearly defined in the text. Criteria included the following.
- Random sequence generation.
- Allocation concealment.
- Blinding of participants, personnel and outcome assessors.
- Incomplete outcome data.
- Selective reporting.
Based on the above, we rated study validity as follows.
- Low risk of bias: if most of the above criteria are met.
- High risk of bias: if most of the above criteria are not met.
- Unclear risk of bias: if criteria are expressed in an unclear manner.
We (HU and KS) resolved any disagreements regarding the assessment by discussion and by coming to consensus.
Measures of treatment effect
We planned to express the treatment effect as a pooled risk ratio and 95% confidence interval (CI) for dichotomous data, and as a mean difference and 95% CI for continuous data. We planned to use a fixed-effect model when minimal heterogeneity existed; otherwise a random-effects model was planned. In cases in which such data combination was not possible or was inappropriate (in the presence of significant clinical heterogeneity or for other reasons), we provided a narrative synthesis.
Unit of analysis issues
In the studies included in the review, participants were randomly assigned to either group (ISBPB or parenteral opioid), and a measurement for each outcome from each participant was recorded and analysed. In cases of multiple observations for the same outcome, we computed the effect measure for each individual participant that incorporated all time points.
Dealing with missing data
Intention-to-treat analysis is recommended to minimize bias. No consensus has been reached on how to handle missing data in ITT analysis in systematic reviews (Higgins 2011). We excluded all participants for whom outcome data were missing.
Assessment of heterogeneity
We planned statistical heterogeneity using the Chi
Assessment of reporting biases
We looked at the number of outcomes presented in the methods section of the included studies and confirmed this number in the results section of the trial. We planned to provide a funnel plot to detect publication bias or a difference between smaller and larger studies expressed by asymmetry (Egger 1997). However, as recommended by the Cochrane Handbook for Systematic Reviews of Interventions, Chapter 10 (Higgins 2011), at least 10 studies are needed to create a funnel plot.
We planned to enter the data extracted from the studies into Review Manager 5.1.
We also planned to:
- pool the data from various trials, when appropriate, and perform analysis in both groups;
- record either the means of an event or the number of participants experiencing an event in each group;
- express the treatment effect as a pooled risk ratio and 95% CI for dichotomous data, and as a mean difference and 95% CI for continuous data;
- use a fixed-effect model when minimal heterogeneity existed, otherwise a random-effects model; and
- generate forest plots when only a single study was available for a particular outcome. In situations in which variables were presented differently, we planned to equate each variable by methods prescribed by the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2011).
Subgroup analysis and investigation of heterogeneity
If appropriate data were found, we planned subgroup analysis based on:
- the type of local anaesthetic used;
- the type of analgesic drug used;
- a combination of drugs in either technique;
- any technique for localization of brachial plexus; and
- the type of surgery performed (in cases of multiple surgical procedures, we will subanalyse each procedure).
Heterogeneity was assessed as mentioned in the section Assessment of heterogeneity.
We planned a sensitivity analysis if methodological quality or characteristics of participants in the studies differed significantly and adequate data were available.
Description of studies
Results of the search
We identified 122 abstracts from the 2774 results obtained by searching according to the methods mentioned in Search methods for identification of studies. The details are presented in Figure 1. Of these, 47 abstracts qualified for full-paper analysis after mutual discussion. We accessed all 47 full-text papers, and 45 were excluded because of inappropriate participants or comparison groups. Thus two studies (Hofmann-Kiefer 2008; Wei 2012) qualified to be included in the final analysis, of which one study was written in Chinese and required translation for extraction of data.
|Figure 1. Abstracts search flow diagram.|
We included two studies in the review. The sample size was 147 participants (87 in Hofmann-Kiefer 2008 and 60 in Wei 2012). See Characteristics of included studies for details. Because of the small number of studies included in the review and significant clinical heterogeneity among them, we were unable to perform meta-analysis in this review.
Upon the advice of Dr Mathew Zacharia and Professor Nathan Pace, we have listed in the Excluded studies section all studies that we retrieved as full-text papers from literature searches that did not meet the eligibility criteria for inclusion in the review. Forty-five studies have been excluded, and the reasons for their exclusion are mentioned in the Characteristics of excluded studies. These studies did not include appropriate interventions or appropriate participants, as mentioned in the Criteria for considering studies for this review.
Risk of bias in included studies
The overall quality of each study was evaluated according to the methodology mentioned in Assessment of risk of bias in included studies. The different bias domains are presented in Characteristics of included studies. A graph and a summary of the risk of bias of included studies are presented in Figure 2 and Figure 3.
|Figure 2. Risk of bias graph: review authors' judgements about each risk of bias item presented as percentages across all included studies.|
|Figure 3. Risk of bias summary: review authors' judgements about each risk of bias item for each included study.|
Effects of interventions
Hofmann-Kiefer 2008 enrolled 87 participants. In all, 17 participants were excluded for reasons mentioned in Characteristics of included studies, and they were removed from the analysis. A total of 16 participants received continuous ISBPB as postoperative analgesia, and one was given parenteral opioids. The authors of the study used a nerve stimulator to locate the brachial plexus.
Wei 2012 enrolled 60 participants, all of whom completed the study. Four participants were excluded according to exclusion criteria. Investigators used ultrasound technique to perform brachial plexus block.
Effectiveness of pain relief
The effectiveness of pain relief was assessed by visual analogue scale (VAS) (Hofmann-Kiefer 2008) and numerical rating pain scale (NRPS) (Wei 2012). Study authors reported observations at different time points (Hofmann-Kiefer 2008 reported at four time points and Wei 2012 at three time points). Because only two studies qualified for inclusion in the review and because significant clinical heterogeneity was present among the studies, we did not perform the meta-analysis; hence, the results are reported in a narrative form.
One hundred thirty participants received continuous ISBPB (66) or parenteral opioid analgesia (64). Hofmann-Kiefer 2008 followed participants for 72 hours postoperatively, whereas Wei 2012 followed participants for 48 hours postoperatively.
Hofmann-Kiefer 2008 reported pain scores at six, 24, 48 and 72 hours after surgery, and pain scores were significantly lower at six, 24 and 72 hours in the ISBPB group than in the group given parenteral opioids. However, pain scores were similar at 48 hours among the two groups. In the Wei 2012 study, pain scores were reported at arrival to the PACU and 24 and 48 hours after surgery; scores were significantly lower in the ISBPB group at all time points.
Incomplete block did not occur in any participant; therefore, no participants were excluded before surgery. Postoperatively, 17 participants were withdrawn (Hofmann-Kiefer 2008) because of one or another adverse event (see Risk of bias in included studies) and were not included in the analysis. Of these, 16 were from the ISBPB group (one for no C5 block, nine for catheter dislocations, one for Horner's syndrome, four for diaphragmatic paresis leading to dyspnoea and one on the participant's choice) and one from the parenteral opioid group (pruritus). Four participants were excluded from the Wei 2012 study, according to exclusion criteria.
Of the secondary outcomes listed in the Types of outcome measures, only the following have been found to be reported in the included studies. The remaining are not reported at all or are mentioned only in narrative rather than numerical form.
Hofmann-Kiefer 2008 mentions that VAS scores were lower in the ISBPB group during physiotherapy at day two (P = 0.016) but were similar at day three. Investigators used piritramide boluses (3.75 to 7.5 mg intravenously) when the pain score exceeded 40 on a scale of 100. Six of 36 participants (16.6%) in the ISBPB group and 11 of 34 (32.3%) in the parenteral opioid group required boluses of piritramide; this was statistically significant (P < 0.01). Regarding nausea and/or vomiting, two of 36 (5.5%) participants in the continuous ISBPB group versus nine of 34 (26.4%) in the parenteral opioid group developed nausea and/or vomiting during the study; this again was statistically significant. No catheter infections were reported.
Wei 2012 looked at participant satisfaction, which was significantly greater (P < 0.01) in the ISBPB group. This study also reported nausea and/or vomiting as an adverse event that was significantly more frequent in the parenteral opioid group (one of 30 (3.3%) in the ISBPB group vs eight of 30 (26.6%) in the parenteral opioid group). The drug used for postoperative analgesia was sufentanil.
Summary of main results
The primary focus of this review was to look at clinically and statistically significant differences in postoperative pain management following major shoulder surgery when continuous ISBPB is compared with parenteral opioids. We also looked at adverse effects related to both techniques, use of supplemental analgesia, participant satisfaction and length of hospital stay.
We included two randomized controlled trials (Hofmann-Kiefer 2008; Wei 2012) consisting of 147 participants, of whom 17 were excluded from analysis—all from the Hofmann-Kiefer 2008 study. Of these 17, 16 had complications related to continuous ISBPB (one no C5 block, nine catheter dislocation, one Horner's syndrome, four diaphragmatic paresis leading to dyspnoea and one poor choice by the participant) and one had complications related to parenteral opioid analgesia (pruritus). Therefore 130 participants were included in the analysis (66 in the continuous ISBPB group and 64 in the parenteral opioid group). The included studies carried a high risk of bias, mainly for the reasons of absence of allocation concealment, lack of blinding and large numbers of dropouts. Both mentioned primary outcomes and several secondary outcomes. A nerve stimulator was used to locate the brachial plexus in all participants in the Hofmann-Kiefer 2008 study, and Wei 2012 used ultrasound to perform brachial plexus block, which may be the reason why no complications related to ISBPB were reported in this study. Because of the small number of studies included in the review, high risk of bias and significant clinical heterogeneity, no concrete conclusions can be drawn. However, following is the summary reported by these two studies.
Effectiveness of pain relief was assessed at different time points for up to 72 hours by Hofmann-Kiefer 2008, and up to 48 hours by Wei 2012. Hofmann-Kiefer 2008 used VAS and Wei 2012 used NRPS to evaluate postoperative pain. Pain scores were significantly lower in the ISBPB group at six, 24 and 72 hours, whereas they were similar at 48 hours in the Hofmann-Kiefer 2008 study. Pain scores were significantly lower in the ISBPB group, as reported by Wei 2012, at all time points (upon arrival to PACU and 24 and 48 hours postoperatively).
Supplemental analgesia was used in the Hofmann-Kiefer 2008 study; the drug used was piritramide, whereas Wei 2012 used sufentanil as supplemental analgesia. The number of boluses was significantly greater in the parenteral opioid group. Adverse effects mentioned included nausea and vomiting, which was significantly more frequent in the parenteral opioid group, as reported by both studies (Hofmann-Kiefer 2008; Wei 2012).
Overall completeness and applicability of evidence
Both studies included in this review had participants, interventions and outcomes that were appropriate to our objectives and outcomes, both primary and secondary. The main issues with the applicability of this evidence include the small number of studies, the high risk of bias in these studies and the presence of significant clinical heterogeneity. Another important point is the method used to perform brachial plexus block. It is evident that ISBPB-related complications were significantly more frequent when a nerve stimulator was used when compared with ultrasound technique. Another aspect worth mentioning is the usual current practice of using ultrasound for performing brachial plexus block; one study in this review did not use ultrasound and may not represent usual current practice.
Quality of the evidence
Generally, the quality of evidence included in our review was limited by a high risk of bias of the included studies, as evaluated from the information provided by the papers. The included studies were prospective randomized clinical trials in which randomization was done by computer or by random number tables. These studies were not placebo-controlled, and sham catheters were not used for ethical reasons, thus making the studies unblinded. These studies were heterogeneous clinically, as different methods were used to perform ISBPB. We were unable to retrieve the protocol and thus were unable to compare published outcomes versus proposed ones.
Potential biases in the review process
We strictly followed the protocol to perform the review and could think of no source of potential bias in the review process.
Agreements and disagreements with other studies or reviews
We found only one review article, which is actually a critical appraisal of current techniques (Fredrickson 2010a), that reviewed the postoperative analgesic techniques used for shoulder surgery. The results of this review are similar to those reported in our review. This review also does not include a meta-analysis.
Implications for practice
Based on the evidence generated by this review, no reasonable conclusions can be drawn. Because of the presence of significant clinical heterogeneity among the two included studies, meta-analysis could not be performed. Current practice is to perform single-shot ISBPB as well as placement of ISBPB catheters by using ultrasound assistance; for this reason, the study by Hofmann-Kiefer 2008 is unlikely to be performed again.
Implications for research
Further randomized controlled trials with low risk of bias are needed to generate good-quality evidence on the topic. Future research should focus on blinding studies to make them stronger and using ultrasound to perform ISBPB.
We would like to thank Mathew Zacharias (content editor); Cathal Walsh (statistical editor); Nathan Pace (co-ordinating editor), Jean Pierre Estebe, Jeff Swenson and Thomas Fichtner Bendtsen (peer reviewers) and Janet Wale (consumer editor) for help and editorial advice provided during the preparation of this systematic review. We would also like to thank Karen Havhannisyan (trials search Co-ordinator, CARG) for developing the search strategies and running them and for helping us to acquire the full text of several papers, and Jane Cracknell (managing editor, CARG) for co-ordinating this review.
We are grateful to Dr Xiaoli Ge for extracting data from the Chinese study (Wei 2012).
Data and analyses
This review has no analyses.
Appendix 1. CENTRAL search strategy
Appendix 2. MEDLINE (via PubMed) search strategy
Appendix 3. EMBASE search strategy
Appendix 4. ISI Web of Science search strategy
Appendix 5. CINAHL (EBSCO host) search strategy
Appendix 6. Data extraction form
Continuous interscalene blockade versus parenteral analgesia for postoperative pain relief after major shoulder surgery
Quality of concealment of random allocation
Comment on treatment
Contributions of authors
Conceiving of the review: Hameed Ullah (HU).
Co-ordinating the review: HU.
Undertaking manual searches: HU, Khalid Samad (KS).
Screening search results: HU, KS.
Organizing retrieval of papers: HU, KS.
Screening retrieved papers against inclusion criteria: HU, KS.
Appraising quality of papers: HU, KS.
Abstracting data from papers: HU.
Writing to authors of papers to ask for additional information: KS.
Providing additional data about papers: HU, KS.
Obtaining and screening data on unpublished studies: HU, KS.
Managing data for the review: HU, KS.
Entering data into Review Manager (RevMan 5.1): HU.
Handling RevMan statistical data: HU, KS.
Performing other statistical analyses not using RevMan: HU, KS.
Entering data double: data entered by person one, HU; data entered by person two, KS.
Interpreting data: HU, KS, Fauzia A Khan (FK).
Making statistical inferences: HU, KS, FK.
Writing the review: HU.
Securing funding for the review: not applicable.
Performing previous work that was the foundation of the present study: HU.
Servind as guarantor for the review (one author): HU.
Taking responsibility for reading and checking the review before submission: KS, FK,
Declarations of interest
Hameed Ullah: none known.
Khalid Samad: none known.
Fauzia A Khan: none known.
Sources of support
- Department of Anaesthesiology, Aga Khan University, Karachi, Pakistan.
- No sources of support supplied
Differences between protocol and review
- The title has undergone a minor change. "Continuous interscalene blockade" has been changed from the protocol (Ullah 2008) to "Continuous interscalene brachial plexus block" in the review.
- In the review, we searched CENTRAL, MEDLINE, EMBASE, CINAHL and ISI Web of Science, whereas we mentioned only the first three databases in the protocol.
- The risk of bias tool has been updated and the methods amended to reflect this.
Medical Subject Headings (MeSH)
*Brachial Plexus; Analgesics, Opioid [*administration & dosage]; Anesthesia, Local [*methods]; Nerve Block [*methods]; Pain, Postoperative [*therapy]; Pirinitramide [administration & dosage]; Randomized Controlled Trials as Topic; Shoulder [*surgery]; Sufentanil [administration & dosage]
MeSH check words