Continuous improvements in peripheral nerve blockade (e.g. using ultrasound-guided techniques) and the advantages of regional blocks (such as reduced side effects) compared with administration of systemic opioids, have strengthened postoperative pain management over recent years.1,2 Furthermore, it has been shown that peripheral regional anaesthetic techniques are used increasingly, but only 25% of peripheral nerve blocks are continuous.3,4 However, the duration of analgesia provided by single-shot nerve blocks is limited to 6 to 8 h, even if only long-lasting local anaesthetic drugs (e.g. bupivacaine or ropivacaine) are used. Therefore, in recent years, research has focused on possible perineural or intravenous adjuvants (opioids,5 dexamethasone6–9 and α2-agonists).
Dexmedetomidine (DEX), an α2-agonist with more promising pharmacokinetic and pharmacodynamic characteristics than clonidine,10,11 has been proved to have a positive effect on postoperative pain intensity and to reduce opioid consumption compared with a placebo when administered intravenously.12,13 In rats, DEX blocks the hyperpolarisation-activated cation current,14 and it also causes more a2-selective vasoconstriction than clonidine.15 Thus DEX might enhance the duration of analgesia when used as an adjunct to local anaesthetics. A preclinical safety trial in rats showed no histopathological damage following a high dose of perineural DEX in addition to bupivacaine in sciatic nerve blocks.16 Recently, three meta-analyses investigated the efficacy and safety of perineural DEX in combination with local anaesthetics versus local anaesthetics alone in brachial plexus blocks.17–19 However, all three meta-analyses can be criticised because insufficient data were analysed, the influence of zero events was not investigated and no trial sequential analysis was performed.
The present meta-analysis investigated the efficacy and safety of perineural DEX in combination with local anaesthetics versus local anaesthetics alone or in combination with intravenous DEX in all peripheral nerve blocks. The latter comparison might be relevant, because perineural (compared with intravenous) administration of DEX in combination with local anaesthetics in a sciatic nerve block produced a longer sensory block.20
Materials and methods
Systematic review and meta-analysis were performed according to the criteria of the PRISMA statement21 and the current recommendations of the Cochrane Collaboration.22 The present review was registered within the PROSPERO database after the beginning of the literature search (CRD42016042486).
Search strategy and study selection
Randomised controlled trials (RCTs) were identified in a systematic search of the data sources MEDLINE, CENTRAL and EMBASE (to May 2017) combining indexed and free text terms (‘dexmedetomidine’ or ‘regional analgesia’ or ‘nerve block’). Search terms were modified to the conditions of each database. Additional articles were checked from references of selected studies and ongoing registered trials on ‘www.clinicaltrials.gov’. There were no restrictions due to language or year of publication. Two reviewers (A.S., S.R.) independently reviewed titles and abstracts in order to exclude irrelevant studies. Differences in literature inclusion/exclusion were decided by consensus or discussed with a third review author (C.M.F.).
We included all RCTs investigating the efficacy and safety of perineural DEX combined with local anaesthesia compared with local anaesthetics alone or local anaesthetics in combination with systemic DEX in peripheral nerve blocks. The following criteria for inclusion were used: prospective RCT, and human studies that included adult men or women (≥18 years) undergoing surgery.
Data collection, extraction and quality analysis
Two review authors (S.R., C.M.F.) independently extracted the data using a self-developed standardised data extraction form. If different doses of perineural DEX were investigated, we always decided to extract the results following the highest perineural DEX dose. At each step of data extraction, differences were discussed within the reviewers’ group.
A critical evaluation of study quality was performed by two reviewers (A.S., S.R.) using the Cochrane Collaboration's recommended tool for assessing risk of bias (sequence generation, allocation sequence concealment, blinding of participants and personnel, blinding of outcome assessment, incomplete outcome data, selective outcome reporting, other potential threats to validity).
Definition of primary and secondary outcome parameters
The following parameters were defined as primary outcome parameters: duration of analgesia (h), number of patients with intraoperative bradycardia, number of patients with hypotension. Secondary outcomes were postoperative pain intensity (<2 h, 2 to 24 h, 24 to 48 h after surgery), onset of motor block or sensory block (min), duration of motor block (h), number of patients with opioid-related adverse events (nausea, vomiting, or nausea and vomiting), respiratory depression, pruritus, sedation or urinary retention.
All statistical analyses were performed with the Review Manager (RevMan; Computer program; version 5.3.5; The Nordic Cochrane Centre, The Cochrane Collaboration, Copenhagen, Denmark) or as otherwise indicated.
Measures of treatment effects
For dichotomous data, we calculated the risk ratio with 95% confidence intervals (95% CIs), while for continuous endpoints, we calculated the mean difference with 95% CI. For the outcome ‘postoperative pain’, all results reported on different scales used to measure pain intensity were pooled. We used the standardised mean difference (SMD) as a summary statistic for pain intensity. For dichotomous and continuous data, we used the inverse variance fixed-effect or random-effects model. A number-needed-to-harm (NNH) was calculated for dichotomous outcomes, if more than at least two trials reported patients suffering from that adverse event and the result was significant.
Zero event handling
Computational problems can occur when no events are observed in one or both groups in an individual study.22 We performed a constant continuity correction of 1.0 (0.5 to each cell) for zero total event trials23 (TSA software v0.9.5.10 Beta; Copenhagen Trial Unit) and presented these results as sensitivity analysis.
Trial sequential analysis
For the primary outcomes, we performed a trial sequential analysis on the basis of all ‘low-risk of bias’ trials to calculate the required information size (RIS).24–27 The following assumptions (as defined in the protocol; a priori) were used to calculate the (heterogeneity-adjusted, D2) RIS and to construct the trial sequential monitoring boundaries by using the TSA software v.0.9.5.10 Beta (Copenhagen Trial Unit)28: a risk for a type I error of 5% and for a type II error of 10% (90% power), mean difference = 4 h with SD = 0.5 h (for duration of anaesthesia), RRR (relative risk reduction) = 10% and control event rate (CER) = 18% (for rate of intraoperative bradycardia) and CER = 60% (for rate of intraoperative hypotension). In a posthoc TSA, we calculated the RIS based on the empirical data from the meta-analyses for the outcome ‘intraoperative bradycardia’ (RRR = −115%, CER = 3.65%) and ‘intraoperative hypotension’ (RRR = −173%, CER = 16.05%), as the empirical data deviated strongly from the a-priori assumptions.
Heterogeneity, subgroup and sensitivity analyses
If data appeared to be clinically and statistically heterogeneous, we used a random-effect model. For all primary outcomes with significant summary treatment effect (95% CI) in random-effects meta-analysis, we additionally calculated the 95% prediction interval, which is an index of dispersion (based on the standard deviation) that tells us how widely the effects vary across the population.29,30 Furthermore, statistical heterogeneity was assessed with the I2 statistic and was rated as substantial if an I2 value higher than 50% was observed.31 To explore the effects of heterogeneity, the following subgroup analyses were performed for the primary outcome ‘duration of analgesia’, if more than five trials were included for the outcome: doses of perineural DEX (25, 50, 75, 100, 150 μg), type of local anaesthetics used (long-lasting versus short-lasting or mixture of local anaesthetics), addition of perineural adrenaline (1 : 200 000), location of nerve block, guiding method (landmark, nerve stimulation, ultrasound), anaesthetic technique (general anaesthesia combined with regional anaesthesia versus regional anaesthesia alone). If weight-based doses of DEX were used, we calculated the approximate dose according to the average weight mentioned in the trials and rounded up or down. In addition, sensitivity analyses were performed to detect the influence of study quality [high-quality trials (study quality rated as ‘low risk’ or ‘unclear risk’) versus low-quality trials (study quality rated as ‘high risk’)] and the impact of zero total event trials.
Assessment of reporting biases
We created funnel plots for outcomes including more than 10 trials by plotting effect estimates of included trials versus their precision (inverse of the standard error of point estimate). If asymmetry was suggested by visual assessment, we investigated funnel plot asymmetry by performing exploratory analyses (e.g. arcsine test for binary data, Egger's test for continuous data). To adjust for small-study effects, we used Duval and Tweedie's trim and fill method. We performed all statistical tests for publication bias using the R package.
Summary of findings
We used the GRADE (Grades of Recommendation, Assessment, Development and Evaluation) approach to rate the quality of evidence,32–39 associated with the primary outcomes in our review. The exact description can be found within the supplementary file (Supplemental Digital Content 1, http://links.lww.com/EJA/A169).
The literature search identified 273 potential studies at the first stage but only 60 RCTs appeared to satisfied the inclusion criteria according to their title and abstract. Fourteen of these were subsequently excluded (Fig. 1).40–53
All 46 remaining studies were parallel group RCTs (Table 1). 54–99 All studies (3149 patients) investigated the efficacy of perineural administration of DEX and local anaesthetics in comparison with placebo and local anaesthetics. The DEX doses were 25,54–58 50,43,54,55,58–69 75,54,70 10071–80 or 150 μg.81 The other studies administrated DEX according to body weight (0.5, 0.75 or 1 μg kg−1).17,82–97 Two trials (107 patients) investigated the comparison of perineural DEX combined with local anaesthetics versus systemic DEX combined with local anaesthetics.63,98 In two studies, adrenaline (1 : 200 000) was given with the local anaesthetics.86,98 Although patients underwent surgery only under regional anaesthesia in 37 studies, nine studies used a combination of general and regional anaesthesia.54,60,62,81,82,87,95,96,98 Different nonopioid analgesics (paracetamol, traditional NSAIDs) were given in 18 out of 46 trials for rescue analgesia.55,56,60,61,63,64,66,67,72,74,75,78,84,86,89–92 In addition, five trials combined nonopioid analgesics with opioids57,65,83,88,98 and 13 trials reported that patients received only opioids for rescue analgesia.54,62,68,69,71,73,76,77,81,85,87,95,96 Nine authors reported no details about rescue analgesics.43,58,59,70,79,82,93,97,99 Finally, five of the included trials reported the use of a prophylactic drug for postoperative nausea and vomiting (PONV) (serotonin antagonists, steroid or droperidol).2,60,66,81,87
The Cochrane Handbook for Systematic Review of Interventions was consulted to assess risk of bias among RCTs. All RCTs provided clear inclusion and exclusion criteria. The methodological quality assessment is summarised in Supplemental Digital Content 2, http://links.lww.com/EJA/A169. Most trials (29 out of 46) used an adequate randomisation method.54,55,59,62,63,68,69,73,74,76–78,80–83,85–90,92,93,95–99 Two trials reported no details about the randomisation method58,67; they were assessed as being at high risk of bias. Only 17 trials specified explicitly the method of allocation concealment (mainly via sealed envelopes).54,60,65,71,73,78,81–83,85,86,88,90,92,94,98,99 Double blinding (participants and personnel) was used in most RCTs (36 out of 46). Five of the included trials were assessed as having an unclear risk of bias.58,67,87,90,97 Five trials were rated as being at a high risk of detection bias because only participants were blinded91,99 or blinding was unclear.65,66,70 Only 22 trials describe the blinding of outcome assessment. Two trials were rated at a high risk, because blinding was unclear or missing.66,91 Twenty RCTs provided complete outcome data.54,57,60,62,66,67,69,74,76,77,79,81,83,86,88,89,92,94,97,99
Comparison of perineural dexmedetomidine combined with local anaesthetics versus local anaesthetics alone
Thirty-two trials investigated the outcome ‘duration of analgesia’.54–56,58,59,61,63–75,77–79,83,85,87,88,91,92,95–98 Eight of these 32 trials were not included in the meta-analysis,65–67,70–72,74,96 because they did not state that data were reported as mean and standard deviation. The exact definitions of duration of analgesia can be seen in Supplemental Digital Content 3, http://links.lww.com/EJA/A169. Three trials did not define the duration of analgesia explicitly.64,68,78 The random-effects meta-analysis of 24 trials with 1510 participants demonstrated that participants receiving 25 to 100 μg of perineural DEX in combination with local anaesthetics had a 4.87 hours longer duration of analgesia than those treated only with local anaesthetics (95% CI 4.02 to 5.73; P < 0.00001, Fig. 2).
Significant heterogeneity influenced this result (I2 = 100%). However, the 95% prediction interval did not include the line of equality (95% prediction interval 0.52 to 9.23) indicating that perineural DEX will be beneficial and prolong analgesia when applied in at least 95% of individual study settings (Supplemental Digital Content 4, http://links.lww.com/EJA/A169). In consideration of the high statistical heterogeneity (I2 = 100%), we performed subgroup analyses. Heterogeneity was not reduced below an I2 of 50% in any of the subgroups with more than two trials (Supplemental Digital Content 5, http://links.lww.com/EJA/A169). The test for subgroup differences reached statistical significance in case of type of local anaesthetics (P < 0.001), nerve block location (P < 0.00001) and guiding technique (P = 0.02) (Supplemental Digital Content 5, http://links.lww.com/EJA/A169).
The effect estimate remained robust in a sensitivity analysis excluding high risk of bias trials56,58,59,61,64,68,73,78,79,91 (mean difference 4.29 h, 95% CI 3.19 to 5.40, 14 trials, 923 patients).54,55,63,69,75,77,83,85,87,88,92,95,97,98 However, the effect estimate changed when the fixed effect model was applied (mean difference 2.32 h, 95% CI 2.28 to 2.36) (Supplemental Digital Content 4, http://links.lww.com/EJA/A169).
Trial sequential analysis for duration of analgesia with a-priori defined assumptions (which were similar to the empirical result of the meta-analysis) and with only low risk of bias trials (14 trials, n = 923) revealed a RIS of 571 participants, indicating that firm evidence was reached with respect to the sample size (Supplemental Digital Content 6, http://links.lww.com/EJA/A169).
There was a discrepancy between funnel plots using random-effects and fixed-effect models. The linear regression test suggested funnel plot asymmetry (P = 0.0059) and the adjusted effect estimate with trim and fill analysis (with 12 studies added) suggested a smaller mean difference in the duration of analgesia (mean difference 2.02 h, 95% CI 1.15 to 2.88).
We graded the quality of the evidence for the outcome ‘duration of analgesia’ as ‘moderate’ (downgrade for inconsistency).
The other primary outcomes, the most common adverse events following DEX administration, intraoperative bradycardia and hypotension, were reported by 30 (1729 patients)54–56,59–64,66,69,70,72,74–76,78–80,83–86,88,92–95,97,98 and 25 (1369 patients)54,56,59–64,66,69,70,72,73,75,76,78,80,83–86,88,92,97,98 of the included trials, respectively. The meta-analysis showed an approximately three times higher risk of intraoperative bradycardia (risk ratio 2.83; 95% CI 1.50 to 5.33; P = 0.035; I2 = 40%; NNH = 12.5; Fig. 3) and hypotension (risk ratio 3.42; 95% CI 1.24 to 9.48; P = 0.002; I2 = 65% NNH = 25; Fig. 4) with perineural DEX. Meta-analyses including all trials with double zero cell frequencies (continuity correction of 0.5) revealed a relative risk of bradycardia of 2.01 (95% CI 1.27 to 3.19) and for hypotension of 1.49 (95% CI 0.94 to 2.38) with perineural DEX.
The meta-analysis for intraoperative hypotension (without double zero event trials) was influenced by heterogeneity (I2 = 40%; I2 = 65%), which was reduced by inclusion of double zero event trials to 11 and 6%, respectively. The 95% prediction interval crossed the line of equality and the range of effects ranged from benefit to harm for bradycardia (95% prediction interval 0.47 to 17.07) and for hypotension (95% prediction interval 0.19 to 62.95). The subgroup analyses investigating possible reasons for the heterogeneity of intraoperative hypotension showed that the addition of perineural adrenaline might have influenced the result (P < 0.001; I2 = 94.8%) (Supplemental Digital Content 7, http://links.lww.com/EJA/A169).
We analysed only trials that reported patients with intraoperative bradycardia or hypotension requiring active treatment. The pooled analysis of 13 trials55,61,64,68,69,72,74,76,80,92–94,98 revealed a two times higher risk of intraoperative bradycardia requiring treatment following perineural DEX administration in combination with local anaesthetics compared with local anaesthetics alone (risk ratio 2.46; 95% CI 1.22 to 4.96; P = 0.01; I2 = 9%; NNH = 20). In contrast, the pooled risk of six trials62–64,66,88,98 for intraoperative hypotension with the need for active treatment was not significantly different between DEX and local anaesthetics compared with local anaesthetics alone (risk ratio 2.54; 95% CI 0.64 to 10.16; P = 0.19; I2 = 50%, NNH = 16.7).
The effect estimates for bradycardia and hypotension did not remain robust in a sensitivity analysis excluding high-risk trials, with risk ratio 2.16 (95% CI 1.00 to 4.67)54,55,69,72,74–76,80,88,93,94,98 and risk ratio 2.73 (95% CI 0.75 to 10.00),54,63,80,88,98 respectively.
Trial sequential analysis for bradycardia with both a-priori defined assumptions and assumptions based on the empirical meta-analysis (including only low risk of bias trials with events in at least one arm) revealed RIS of 80 032 and 5139 participants, respectively, indicating that insufficient evidence was reached with respect to the sample size. Trial sequential analysis for hypotension resulted in a RIS of 31 510 and 4584 participants with a-priori and empirical assumptions, respectively. In this case, evidence was also insufficient.
Funnel plots and explorative analyses for bradycardia and hypotension did not reveal funnel plot asymmetry.
We graded the quality of the evidence for the outcomes ‘bradycardia’ and ‘hypotension’ as ‘very low’ (downgrade for risk of bias, inconsistency and imprecision).
Seven trials investigated the postoperative pain intensity at the earliest time point (<2 h) (474 patients).57,58,60,89,90,95,98 The meta-analysis showed no significant differences between participants treated with perineural DEX combined with local anaesthetics compared with local anaesthetics alone (SMD −0.16 points; 95% CI −1.73 to 1.41; P = 0.84; I2 = 97%). The same holds true for the other time point reported in six included trials (429 patients)57,60,81,87,90,95 (SMD2–24
h −0.42 points; 95% CI −1.01 to 0.18; P = 0.17; I2 = 87%). Only one trial (60 patients) reported data for the last time point (24 to 48 h) and again showed no significant differences between patients treated with perineural DEX combined with local anaesthetics compared with local anaesthetics alone.83
In all, 34 trials (2058 patients)54–61,63–70,72–75,77–80,83–85,88,89,91–93,97,98 investigated the duration of motor blockade defined as the time interval between the end of the injection and complete recovery of motor function (Supplemental Digital Content 3, http://links.lww.com/EJA/A169). Patients receiving perineural DEX combined with local anaesthetics reported a longer duration of motor blockade than patients treated with local anaesthetics only (mean difference 3.92 h; 95% CI 3.17 to 4.67; P < 0.00001; I2 = 100%). Results of the meta-analyses ‘onset of motor or sensory blockade’ can be found in Supplemental Digital Content 8, http://links.lww.com/EJA/A169.
The most common adverse events were postoperative nausea (PON), postoperative vomiting (POV), PONV and respiratory depression. No patient with respiratory depression was reported in 18 trials (1095 patients),54,59,63,64,66,69,73,75,76,78,79,83–86,88,92,93 but the meta-analyses focusing on PON (13 trials; 785 patients),59–61,63,72,73,75,78,79,82,92,95,96 POV (12 trials; 734 patients)59–61,63,72,73,75,78,79,92,95,96 and PONV (11 trials; 586 patients)64,66,70,71,76,78,82,83,86,88,98 showed no significant difference between patients treated with perineural DEX combined with local anaesthetics versus local anaesthetics (Supplemental Digital Content 9, http://links.lww.com/EJA/A169). The same holds true for the other meta-analyses investigating the risk for pruritus (six trials; 323 patients)66,70–72,78,83 and sedation (nine trials; 591 patients64,66,69,70,78,79,83,86,89 (Supplemental Digital Content 9, http://links.lww.com/EJA/A169). No patient with urinary retention was reported within three trials including 167 participants.72,78,96
Comparison of perineural dexmedetomidine combined with local anaesthetics versus systemic DEX combined with local anaesthetics
Because only two trials (107 patients) investigated the comparison ‘perineural DEX combined with LA versus systemic DEX combined with LA’, meta-analyses could be performed only for a low number of outcomes.63,98
The included trials defined the duration of analgesia as ‘time to first analgesic request’63 or ‘time to first pain’.98 The pooled analysis showed that there was a significant difference between the duration of analgesia between 53 participants treated with less than 0.05 mg of perineural DEX combined with local anaesthetics compared with 54 patients receiving local anaesthetics only (mean difference 0.98 h; 95% CI −0.12 to - 2.08; P = 0.08; I2 = 0%). Due to the limited number of included trials, sensitivity analyses and funnel plot analysis were not performed.
Both trials reported data for bradycardia and hypotension, but there were no differences between the risk for intraoperative bradycardia (risk ratio 0.47; 95% CI 0.11 to 1.97; P = 0.3; I2 = 0%) or hypotension (risk ratio 0.77; 95% CI 0.5 to 1.18; P = 0.23; I2 = 0%) between groups.63,98 Only one trial reported the more specific outcome ‘number of patients with intraoperative bradycardia, respectively hypotension requiring treatment’ and again showed no significant difference between the groups.98
The two included trials investigated only the outcome duration of motor blockade, defined as return to normal or presurgical strength in the arm,98 and the recovery of complete motor power of the hand and forearm.63 The meta-analysis showed a significantly longer duration of motor blockade in patients receiving perineural DEX combined with local anaesthetics compared with systemic DEX combined with local anaesthetics (mean difference 2.19 h; 95% CI 0.67 to 03.7; P = 0.005; I2 = 0%).
The present meta-analysis indicates that use of DEX as an adjuvant to long-lasting local anaesthetics significantly prolongs the duration of analgesia by around 5 h following peripheral regional anaesthesia, but is associated with three times higher risks of intraoperative bradycardia and hypotension. However, the meta-analysis demonstrated that the duration of analgesia and the risks of intraoperative bradycardia or hypotension might not differ between patients treated with perineural or systemic DEX combined with local anaesthetics.
Efficacy of perineural dexmedetomidine in peripheral nerve blocks
The present meta-analysis demonstrated that the duration of analgesia is prolonged by around 5 h, if 25 to 150 μg DEX is added perineurally. In contrast to other meta-analyses,17–19 we excluded trials if data were not explicitly reported as mean and SD or 95% CI. However, we showed for the first time within the TSA (accounting for heterogeneity) that the evidence is powered sufficiently by a large number of trials so that RCTs should not investigate the duration of analgesia following perineural DEX combined with local anaesthetics compared with local anaesthetics alone in the future.
The subgroup analyses demonstrated that the duration of analgesia might be longer if DEX is used in combination with long-lasting local anaesthetics (instead of short-lasting local anaesthetics or a local anaesthetic mixture), for supraclavicular or femoral nerve blocks (compared with interscalene, infraclavicular, axillary nerve blocks, transversus abdominis plane blocks and eye blocks) or blocks that were performed guided by ultrasound (compared with nerve stimulation or a landmark technique). We could not demonstrate a DEX dose-related difference for the duration of analgesia, in contrast to another meta-analysis.17 However, a RCT in volunteers not only showed a significant dose-dependent increase in mean duration of sensory block but also dose-dependent prolonged sedation following perineural DEX use combined with ropivacaine 0.2%.92,100 These authors concluded that 100 μg of perineural DEX might offer the best risk–benefit profile. Another clinical study showed that the addition of perineural DEX decreased the median effective dose of lignocaine required for an obturator nerve block in patients undergoing resection of bladder tumours.42,50 However, the latter results should be confirmed in more common nerve blocks (e.g. femoral or axillary nerve blocks).
We showed a significant prolongation of motor blockade following the perineural administration of DEX. This difference (∼4 h) might be clinically relevant if it delays discharge following outpatient surgery. Nevertheless, those results were influenced by significant heterogeneity and should be viewed with caution.
We investigated for the first time the comparison of perineural versus intravenous DEX administration in combination with local anaesthetics, but there was no significant difference for the duration of analgesia. This is confirmed by a recent RCT in volunteers,20 which showed that DEX as an adjuvant might have a peripheral action, but the observed difference might show no clinical relevance. However, our results were limited by the small number of trials, and there is a need for more RCTs investigating this aspect in the future.
Safety of perineural dexmedetomidine in peripheral nerve blocks
Because perineural administration of DEX is an off-label use, reporting of possible adverse events and an appropriate risk–benefit analysis are crucial before using DEX as an adjuvant in everyday regional anaesthesia. We showed that the risk for intraoperative hypotension or bradycardia is around three times higher in patients treated with perineural DEX compared with placebo. The results for intraoperative bradycardia were comparable with other systematic reviews.17,19
However, because one-third of the included trials reported zero events in treatment and control groups, we investigated this fact in a specific sensitivity analysis and revealed only around a two times higher risk of intraoperative bradycardia, while the risk for intraoperative hypotension was not significantly different between treatment and control groups. This result was confirmed in the sensitivity analysis, including only trials that specifically stated the number of patients with intraoperative bradycardia and hypotension with the need for active treatment. However, the TSA clearly showed that the overall number of patients analysed is too low to obtain a clear view of the evidence.
Interestingly, there was no significant difference in the risks of intraoperative bradycardia or hypotension in the meta-analyses of the two trials which compared perineural versus systemic DEX in combination with local anaesthetics for peripheral nerve blockade.
Taken together, according to the GRADE approach, the overall quality of evidence for the risk of intraoperative bradycardia or hypotension was rated as very low (Supplemental Digital Content 10, http://links.lww.com/EJA/A169). There is an urgent need for large RCTs focusing on these adverse events before perineural DEX is used in every patient receiving regional anaesthesia.
Other adverse events were reported in only a small number of included trials and there were no significant differences in the risks of PON, POV, PONV, respiratory depression, pruritus, sedation and urinary retention between patients treated with perineural DEX or placebo.
Finally, an important issue of perineural adjuvants is the risk of neurotoxicity. After an injection at the sciatic nerves of rats, several preclinical studies suggested that the combination of perineural DEX with bupivacaine or ropivacaine did not cause more significant histopathological changes compared with administration of ropivacaine93,101 or bupivacaine16,94,95,102,103 alone. Moreover, these studies revealed that perineural DEX reduced the local anaesthetic induced inflammatory effects.94,102 Therefore, the use of perineural DEX might be protective against local anaesthetics induced nerve damage. Within our meta-analysis, no trial reported any case of possible nerve damage, but the duration of the follow-up was no longer than 2 weeks.
The most important limitation is the large heterogeneity of the meta-analysis investigating the primary outcomes ‘duration of analgesia’ and ‘intraoperative hypotension’ for the comparison of perineural DEX in combination with local anaesthetics versus local anaesthetics alone. Accordingly, we performed a number of subgroup and sensitivity analyses and identified a number of possible influencing factors (e.g. nerve block location), which should be studied more closely in the future. In contrast to the previously published meta-analyses,17–19 we calculated the 95% prediction interval for all primary outcomes, in order to obtain a better view of treatment effects, and showed less pronounced effects. Finally, we cannot exclude that different definitions describing primary and secondary outcomes might have influenced our results, but there is currently no ‘gold standard’ for the description of regional anaesthesia outcomes.96,104
In conclusion, the present meta-analysis of 46 RCTs demonstrated that perineural DEX in combination with local anaesthetics led to a longer duration of analgesia compared with the use of local anaesthetics alone (moderate quality evidence). However, it was associated with increased risks of intraoperative bradycardia or hypotension (both very low quality evidence). In contrast, there was no significant difference between perineural or systemic application of DEX in combination with local anaesthetics. As the perineural injection of DEX is currently off-label in peripheral regional anaesthesia, there is a need for trials focusing on adverse events following perineural DEX to better define patients at a high risk of intraoperative bradycardia or hypotension.
Acknowledgements relating to this article
Assistance with the study: none.
Financial support and sponsorship: none.
Conflicts of interest: none.
Presentation: preliminary data from this study were presented as a poster presentation at the Congress of the German Society of Anaesthesiology and Intensive Care Medicine ‘DAC 2017’, 3 to 5 May 2017, Nürnberg, Germany.
1. Chan E-Y, Fransen M, Parker DA, et al. Femoral nerve blocks for acute postoperative pain after knee replacement surgery. Cochrane Database Syst Rev
2. Abdallah FW, Macfarlane AJR, Brull R. The requisites of needle-to-nerve proximity for ultrasound-guided regional anesthesia: a scoping review of the evidence. Reg Anesth Pain Med
3. Memtsoudis SG, Poeran J, Zubizarreta N, et al. Anesthetic care for orthopedic patients: is there a potential for differences in care? Anesthesiology
4. Sites BD, Barrington MJ, Davis M. Using an international clinical registry of regional anesthesia to identify targets for quality improvement. Reg Anesth Pain Med
5. Schnabel A, Reichl SU, Zahn PK, et al. Efficacy and safety of buprenorphine in peripheral nerve blocks: a meta-analysis of randomised controlled trials. Eur J Anaesthesiol
6. Huynh TM, Marret E, Bonnet F. Combination of dexamethasone and local anaesthetic solution in peripheral nerve blocks: a meta-analysis of randomised controlled trials. Eur J Anaesthesiol
7. Choi S, Rodseth R, McCartney CJL. Effects of dexamethasone as a local anaesthetic adjuvant for brachial plexus block: a systematic review and meta-analysis of randomized trials. Br J Anaesth
8. Rasmussen SB, Saied NN, Bowens C, et al. Duration of upper and lower extremity peripheral nerve blockade is prolonged with dexamethasone when added to ropivacaine: a retrospective database analysis. Pain Med
9. Abdallah FW, Johnson J, Chan V, et al. Intravenous dexamethasone and perineural dexamethasone similarly prolong the duration of analgesia after supraclavicular brachial plexus block: a randomized, triple-arm, double-blind, placebo-controlled trial. Reg Anesth Pain Med
10. Fairbanks CA, Kitto KF, Nguyen HO, et al. Clonidine and dexmedetomidine produce antinociceptive synergy in mouse spinal cord. Anesthesiology
11. Gil DW, Cheevers CV, Kedzie KM, et al. Alpha-1-adrenergic receptor agonist activity of clinical alpha-adrenergic receptor agonists interferes with alpha-2-mediated analgesia. Anesthesiology
12. Jessen Lundorf L, Korvenius Nedergaard H, Møller AM. Perioperative dexmedetomidine for acute pain after abdominal surgery in adults. Cochrane Database Syst Rev
13. Schnabel A, Meyer-Frießem CH, Reichl SU, et al. Is intraoperative dexmedetomidine a new option for postoperative pain treatment? A meta-analysis of randomized controlled trials. Pain
14. Brummett CM, Hong EK, Janda AM, et al. Perineural dexmedetomidine added to ropivacaine for sciatic nerve block in rats prolongs the duration of analgesia by blocking the hyperpolarization-activated cation current. Anesthesiology
15. Masuki S, Dinenno FA, Joyner MJ, Eisenach JH. Selective alpha2-adrenergic properties of dexmedetomidine over clonidine in the human forearm. J Appl Physiol
16. Brummett CM, Norat MA, Palmisano JM, Lydic R. Perineural administration of dexmedetomidine in combination with bupivacaine enhances sensory and motor blockade in sciatic nerve block without inducing neurotoxicity in rat. Anesthesiology
17. Vorobeichik L, Brull R, Abdallah FW. Evidence basis for using perineural dexmedetomidine to enhance the quality of brachial plexus nerve blocks: a systematic review and meta-analysis of randomized controlled trials. Br J Anaesth
18. Ping Y, Ye Q, Wang W, et al. Dexmedetomidine as an adjuvant to local anesthetics in brachial plexus blocks: a meta-analysis of randomized controlled trials. Medicine (Baltimore)
19. Hussain N, Grzywacz VP, Ferreri CA, et al. Investigating the efficacy of dexmedetomidine as an adjuvant to local anesthesia in brachial plexus block: a systematic review and meta-analysis of 18 randomized controlled trials. Reg Anesth Pain Med
20. Andersen JH, Grevstad U, Siegel H, et al. Does dexmedetomidine have a perineural mechanism of action when used as an adjuvant to ropivacaine? A paired, blinded, randomized trial in healthy volunteers. Anesthesiology
21. Moher D, Liberati A, Tetzlaff J, Altman DG. PRISMA Group. Preferred reporting items for systematic reviews and meta-analyses: the PRISMA statement. BMJ
22. Higgins J, Green S, editors. Handbook for systematic reviews of interventions version 5.1.0 [updated March 2011]. Cochrane Collab; 2011. http://www.cochrane-handbook.org
. [Accessed 10 July 2018].
23. Sweeting MJ, Sutton AJ, Lambert PC. What to add to nothing? Use and avoidance of continuity corrections in meta-analysis of sparse data. Stat Med
24. Brok J, Thorlund K, Gluud C, Wetterslev J. Trial sequential analysis reveals insufficient information size and potentially false positive results in many meta-analyses. J Clin Epidemiol
25. Thorlund K, Devereaux PJ, Wetterslev J, et al. Can trial sequential monitoring boundaries reduce spurious inferences from meta-analyses? Int J Epidemiol
26. Wetterslev J, Thorlund K, Brok J, Gluud C. Trial sequential analysis may establish when firm evidence is reached in cumulative meta-analysis. J Clin Epidemiol
27. Wetterslev J, Thorlund K, Brok J, Gluud C. Estimating required information size by quantifying diversity in random-effects model meta-analyses. BMC Med Res Methodol
28. Copenhagen Trial Unit. TSA Software. 2017. http://www.ctu.dk/tsa/
. [Accessed 10 July 2018].
29. Borenstein M, Higgins JPT, Hedges LV, Rothstein HR. Basics of meta-analysis: I(2) is not an absolute measure of heterogeneity. Res Synth Methods
30. IntHout J, Ioannidis JPA, Rovers MM, Goeman JJ. Plea for routinely presenting prediction intervals in meta-analysis. BMJ Open
31. Higgins JPT, Thompson SG. Quantifying heterogeneity in a meta-analysis. Stat Med
32. Guyatt G, Oxman AD, Akl EA, et al. GRADE guidelines: 1. Introduction-GRADE evidence profiles and summary of findings tables. J Clin Epidemiol
33. Guyatt GH, Oxman AD, Kunz R, et al. GRADE guidelines: 2. Framing the question and deciding on important outcomes. J Clin Epidemiol
34. Guyatt GH, Oxman AD, Vist G, et al. GRADE guidelines: 4. Rating the quality of evidence-study limitations (risk of bias). J Clin Epidemiol
35. Guyatt GH, Oxman AD, Kunz R, et al. GRADE guidelines: 7. Rating the quality of evidence-inconsistency. J Clin Epidemiol
36. Guyatt GH, Oxman AD, Kunz R, et al. GRADE guidelines: 8. Rating the quality of evidence-indirectness. J Clin Epidemiol
37. Guyatt GH, Oxman AD, Kunz R, et al. GRADE guidelines 6. Rating the quality of evidence-imprecision. J Clin Epidemiol
38. Guyatt GH, Oxman AD, Montori V, et al. GRADE guidelines: 5. Rating the quality of evidence-publication bias. J Clin Epidemiol
39. Balshem H, Helfand M, Schünemann HJ, et al. GRADE guidelines: 3. Rating the quality of evidence. J Clin Epidemiol
40. Kaur H, Singh G, Rani S, et al. Effect of dexmedetomidine as an adjuvant to levobupivacaine in supraclavicular brachial plexus block: a randomized double-blind prospective study. J Anaesthesiol Clin Pharmacol
41. Nema N, Badgaiyan H, Raskaran S, et al. Effect of addition of dexmedetomidine to ropivacaine hydrochloride (0.75%) in brachial plexus block through supraclavicular route in upper limb surgeries: a clinical comparative study. J Evol Med Dent Sci
42. Sharma B, Rupal S, Swami AC, Lata S. Effect of addition of dexmedetomidine to ropivacaine 0.2% for femoral nerve block in patients undergoing unilateral total knee replacement: a randomised double-blind study. Indian J Anaesth
43. Luan H, Zhang X, Feng J, et al. Effect of dexmedetomidine added to ropivacaine on ultrasound-guided transversus abdominis plane block for postoperative analgesia after abdominal hysterectomy surgery: a prospective randomized controlled trial. Minerva Anestesiol
44. Kang H. The effect of dexmedetomidine added to preemptive ropivacaine infiltration on postoperative pain after inguinal herniorrhaphy: a prospective, randomized, double-blind, placebo-controlled study. Eur Surg Acta Chir Austriaca
45. Mohamed SA, Fares KM, Mohamed AA, Alieldin NH. Dexmedetomidine as an adjunctive analgesic with bupivacaine in paravertebral analgesia for breast cancer surgery. Pain Physician
46. Mohta M, Kalra B, Sethi AK, Kaur N. Efficacy of dexmedetomidine as an adjuvant in paravertebral block in breast cancer surgery. J Anesth
47. Obayah GM, Refaie A, Aboushanab O, et al. Addition of dexmedetomidine to bupivacaine for greater palatine nerve block prolongs postoperative analgesia after cleft palate repair. Eur J Anaesthesiol
48. Hanoura SE, Elsayed MM, Abdullah AA, et al. Dexmedetomidine improves the outcome of a bupivacaine brachial plexus axillary block: a prospective comparative study. Ain-Shams J Anaesthesiol
49. Rancourt M-PM, Albert NT, Côté M, et al. Posterior tibial nerve sensory blockade duration prolonged by adding dexmedetomidine to ropivacaine. Anesth Analg
50. Lu Y, Sun J, Zhuang X, et al. Perineural dexmedetomidine as an adjuvant reduces the median effective concentration of lidocaine for obturator nerve blocking: a double-blinded randomized controlled trial. PLoS One
51. Rutkowska K, Knapik P, Misiolek H. The effect of dexmedetomidine sedation on brachial plexus block in patients with end-stage renal disease. Eur J Anaesthesiol
52. Yektaş A, Gümüş F, Alagol A. Dexmedetomidine and propofol infusion on sedation characteristics in patients undergoing sciatic nerve block in combination with femoral nerve block via anterior approach. Braz J Anesthesiol
53. Helal S, Eskandr A, Gaballah K, Gaarour I. Effects of perineural administration of dexmedetomidine in combination with bupivacaine in a femoral-sciatic nerve block. Saudi J Anaesth
54. Abdulatif M, Fawzy M, Nassar H, et al. The effects of perineural dexmedetomidine on the pharmacodynamic profile of femoral nerve block: a dose-finding randomised, controlled, double-blind study. Anaesthesia
55. Channabasappa S, Shetty V, Dharmappa S, Sarma J. Efficacy and safety of dexmedetomidine as an additive to local anesthetics in peribulbar block for cataract surgery. Anesth Essays Res
56. Gandhi R, Shah A, Patel I. Use of dexmedetomidine along with bupivacaine for brachial plexus block. Natl J Med Res
57. Ghali AM, Shabana AM, El Btarny AM. The effect of low-dose dexmedetomidine as an adjuvant to levobupivacaine in patients undergoing vitreoretinal surgery under sub-tenon's block anesthesia. Anesth Analg
58. Saraf DS, Batra DM, Agarwal DS, et al. Dexmedetomidine as adjuvant to ropivacaine in brachial plexus block-a dose response study. Indian J Res
59. Arun S. Effect of dexmedetomidine as an adjuvant to 0.75% ropivacaine in axillary brachial plexus block for forearm and hand surgeries. Int J Biomed Res
60. Bengisun ZK, Ekmekçi P, Akan B, et al. The effect of adding dexmedetomidine to levobupivacaine for interscalene block for postoperative pain management after arthroscopic shoulder surgery. Clin J Pain
61. Dar FA, Najar MR, Jan N. Dexmedetomidine added to ropivacaine prolongs axillary brachial plexus block. Int J Biomed Adv Res
62. Gurajala I, Thipparampall AK, Durga P, Gopinath R. Effect of perineural dexmedetomidine on the quality of supraclavicular brachial plexus block with 0.5% ropivacaine and its interaction with general anaesthesia. Indian J Anaesth
63. Kathuria S, Gupta S, Dhawan I. Dexmedetomidine as an adjuvant to ropivacaine in supraclavicular brachial plexus block. Saudi J Anaesth
64. Khade DAR, Makwana DJC, Jethva DNK, et al. Evaluation of effect of dexmeditomidine as an adjuvant to bupivacaine in supraclavicular brachial plexus block. Natl J Integr Res Med
65. Kumar N. Comparative study between 0.25% bupivacaine with 8 mg dexamethasone and 0.25% bupivacaine with 50 (g dexmedetomidine as an adjuvant for interscalene brachial plexus block: prospective clinical study. J Evol Med Dent Sci
66. Manohar P, Prakash M. Comparison of the effects of fentanyl and dexmedetomidine in supraclavicular brachial plexus block achieved with 0.5% bupivacaine in Karpaga Vinayaga Medical College and Hospital, Maduranthagam. J Med Sci Clin Res
67. Patki YS, Bengali R, Patil T, et al. Efficacy of dexmedetomidine as an adjuvant to 0.5% ropivacaine in supraclavicular brachial plexus block for postoperative analgesia. Int J Sci Res
68. Rashmi HD, Komala HK. Effect of dexmedetomidine as an adjuvant to 0.75% ropivacaine in interscalene brachial plexus block using nerve stimulator: a prospective, randomized double-blind study. Anesth Essays Res
69. Tiwari P, Manish J, Rastogi B, et al. A comparative clinical study of perineural administration of 0.75% ropivacaine with and without dexmedetomidine in upper limb surgery by ultrasound guided single injection supraclavicular brachial plexus block. Glob Anesth Perioper Med
70. Li J, Halaszynski TM. Dexmedetomidine improves supraclavicular brachial plexus block quality and duration of analgesia when used together with 0.5% ropivacaine: single center randomized control trial. 2015. http://asaabstracts.com/strands/asaabstracts/abstract.htm?year=2015&index=1&absnum=4257
. [Accessed 12 December 2017].
71. Abdelaal W, Metry AA, Refaat M, et al. Comparative study between levobupivacaine versus levobupivacaine plus dexmedetomidine for transversus abdominis plane block ‘TAP’ in postoperative pain management after abdominoplasty. J Anesthesiol Crit Care Med 2015; 2: 1–7.
72. Agarwal S, Aggarwal R, Gupta P. Dexmedetomidine prolongs the effect of bupivacaine in supraclavicular brachial plexus block. J Anaesthesiol Clin Pharmacol
73. Biswas S, Das RK, Mukherjee G, Ghose T. Dexmedetomidine an adjuvant to levobupivacaine in supraclavicular brachial plexus block: a randomized double blind prospective study. Ethiop J Health Sci
74. Das A, Majumdar S, Halder S, et al. Effect of dexmedetomidine as adjuvant in ropivacaine-induced supraclavicular brachial plexus block: a prospective, double-blinded and randomized controlled study. Saudi J Anaesth
75. Esmaoglu A, Yegenoglu F, Akin A, Turk CY. Dexmedetomidine added to levobupivacaine prolongs axillary brachial plexus block. Anesth Analg
76. Lee MJ, Koo DJ, Choi YS, et al. Dexamethasone or dexmedetomidine as local anesthetic adjuvants for ultrasound-guided axillary brachial plexus blocks with nerve stimulation. Korean J Pain
77. Mirkheshti A, Saadatniaki A, Salimi A, et al. Effects of dexmedetomidine versus ketorolac as local anesthetic adjuvants on the onset and duration of infraclavicular brachial plexus block. Anesthesiol Pain Med
78. Singh AP, Mahindra M, Gupta R, Bajwa SJS. Dexmedetomidine as an adjuvant to levobupivacaine in supraclavicular brachial plexus block: a novel anesthetic approach. Anesth Essays Res
79. Tandon N, Gupta M, Agrawal J, et al. Supraclavicular brachial plexus block, levobupivacaine, clonidine, dexmedetomidine. J Evol Med Dent Sci
80. Zhang Y, Wang C-S, Shi J-H, et al. Perineural administration of dexmedetomidine in combination with ropivacaine prolongs axillary brachial plexus block. Int J Clin Exp Med
81. Fritsch G, Danninger T, Allerberger K, et al. Dexmedetomidine added to ropivacaine extends the duration of interscalene brachial plexus blocks for elective shoulder surgery when compared with ropivacaine alone: a single-center, prospective, triple-blind, randomized controlled trial. Reg Anesth Pain Med
82. Almarakbi W, Kaki AM. Addition of dexmedetomidine to bupivacaine in transversus abdominis plane block potentiates postoperative pain relief among abdominal hysterectomy patients: a prospective randomized controlled trial. Saudi J Anaesth
83. Ammar A, Mahmoud KM. Ultrasound-guided single injection infraclavicular brachial plexus block using bupivacaine alone or combined with dexmedetomidine for pain control in upper limb surgery: a prospective randomized controlled trial. Saudi J Anaesth
84. Anandh K, Bhupal JP, Grewal TK, et al. Dexmedetomidine prolongs the effect of 0.5% isobaric levobupivacaine in axillary brachial plexus block. J Evol Med Dent Sci
85. Bangera A, Manasa M, Krishna P. Comparison of effects of ropivacaine with and without dexmedetomidine in axillary brachial plexus block: a prospective randomized double-blinded clinical trial. Saudi J Anaesth
86. Bharti N, Sardana DK, Bala I. The analgesic efficacy of dexmedetomidine as an adjunct to local anesthetics in supraclavicular brachial plexus block: a randomized controlled trial. Anesth Analg
87. Bisht S, Sreeramalu, Sunitha K, et al. Use of dexmedetomidine as adjuvant in oblique subcostal transversus abdominis plane block. J Evol Med Dent Sci
88. Chinnappa J, Shivanna S, Pujari VS, Anandaswamy TC. Efficacy of dexmedetomidine with ropivacaine in supraclavicular brachial plexus block for upper limb surgeries. J Anaesthesiol Clin Pharmacol
89. Das B, Lakshmegowda M, Sharma M, et al. Supraclavicular brachial plexus block using ropivacaine alone or combined with dexmedetomidine for upper limb surgery: a prospective, randomized, double-blinded, comparative study. Rev Esp Anestesiol Reanim
90. Ding W, Li W, Zeng X, et al. Effect of adding dexmedetomidine to ropivacaine on ultrasound-guided dual transversus abdominis plane block after gastrectomy. J Gastrointest Surg
91. Dixit A, Singhal S, Neema C, Sanwatsarkar S. An evaluation of the addition of dexmedetomidine to levobupivacaine for supraclavicular brachial plexus block in upper limb orthopaedic surgeries. Asian Pac J Health Sci
92. Kaygusuz K, Kol IO, Duger C, et al. Effects of adding dexmedetomidine to levobupivacaine in axillary brachial plexus block. Curr Ther Res
93. Kwon Y, Hwang SM, Lee JJ, Kim JH. The effect of dexmedetomidine as an adjuvant to ropivacaine on the bispectral index for supraclavicular brachial plexus block. Korean J Anesthesiol
94. Lin Y-N, Li Q, Yang R-M, et al. Addition of dexmedetomidine to ropivacaine improves cervical plexus block. Acta Anaesthesiol Taiwan
95. Rai P, Singh Negi D, Singh SK, et al. Effect of addition of dexmedetomidine to ropivacaine in transverses abdominis plane block on postoperative pain in lower segment caesarean section: a randomized controlled trial. J Dental and Med Sciences
96. Santosh B, Mehandale S. Does dexmedetomidine improve analgesia of superficial cervical plexus block for thyroid surgery? Indian J Anaesth
97. Song J-H, Shim HY, Lee TJ, et al. Comparison of dexmedetomidine and epinephrine as an adjuvant to 1% mepivacaine in brachial plexus block. Korean J Anesthesiol
98. Abdallah FW, Dwyer T, Chan VWS, et al. IV and perineural dexmedetomidine similarly prolong the duration of analgesia after interscalene brachial plexus block: a randomized, three-arm, triple-masked, placebo-controlled trial. Anesthesiology
99. Khandaitkar S, Kolte V, Shenoi SR, Budhraja N. A clinical study to determine the efficacy of 7 ppm dexmedetomidine as an adjuvant to 2% lignocaine in infraorbital nerve block. Br J Oral Maxillofac Surg
100. Keplinger M, Marhofer P, Kettner SC, et al. A pharmacodynamic evaluation of dexmedetomidine as an additive drug to ropivacaine for peripheral nerve blockade: a randomised, triple-blind, controlled study in volunteers. Eur J Anaesthesiol
101. Brummett CM, Padda AK, Amodeo FS, et al. Perineural dexmedetomidine added to ropivacaine causes a dose-dependent increase in the duration of thermal antinociception in sciatic nerve block in rat. Anesthesiology
102. Tüfek A, Kaya S, Tokgöz O, et al. The protective effect of dexmedetomidine on bupivacaine-induced sciatic nerve inflammation is mediated by mast cells. Clin Invest Med
103. Memari E, Hosseinian M-A, Mirkheshti A, et al. Comparison of histopathological effects of perineural administration of bupivacaine and bupivacaine-dexmedetomidine in rat sciatic nerve. Exp Toxicol Pathol
104. Abdallah FW, Brull R. The definition of block ‘success’ in the contemporary literature: are we speaking the same language? Reg Anesth Pain Med
* Both Alexander Schnabel and Sylvia U. Reichl contributed equally to this study.