Background Adductor canal block (ACB) can provide important analgesic benefits following total knee arthroplasty (TKA), however, the extent to which these benefits can be enhanced or prolonged by a continuous catheter-based infusion compared with a single-shot injection of local anesthetic is unclear.
Objectives This systematic review and meta-analysis (PROSPERO: CRD42021292738) review sought to compare the analgesic effectiveness of single shot to continuous ACB following TKA.
Evidence review We sought randomized trials from the US National Library of Medicine database (MEDLINE), Excerpta Medica database (EMBASE), and Cochrane Database of Systematic Reviews from inception to November 1, 2021, that compared single-shot to continuous ACB in adult patients undergoing TKA. The primary outcomes were (1) area under the curve (AUC) pain severity at rest and (2) cumulative opioid (oral morphine equivalent) consumption during the first 48 hours postoperatively. Secondary outcomes included postoperative pain severity scores up to 48 hours, cumulative opioid consumption at 24 hours, functional recovery, opioid-related side effects, and block-related complications. Risk of bias of included studies was assessed using the Cochrane risk of bias tool. Statistical pooling was conducted using the Hartung-Knapp-Sidik-Jonkman method for random effects. No funding was obtained for this review.
Findings Eleven trials (1185 patients) were included. No differences were observed in rest pain severity (AUC) or cumulative opioid consumption up to 48 hours postoperatively. In addition, no differences were observed in individual postoperative rest pain scores in the recovery room and at 12 and 24 hours, or in cumulative opioid consumption at 24 hours, functional recovery, and opioid-related side effects. Finally, fewer block-related complications were observed with single-shot ACB, with an OR (95% CI) of 0.24 (0.14 to 0.41) (p=0.002).
Conclusions Our results suggest that continuous catheter-based ACB does not enhance or prolong the analgesic benefits when compared with single-shot ACB for TKA over the first 48 hours postoperatively. Overall, the results of our meta-analysis do not support the routine use of continuous ACB for postoperative analgesia after TKA.
- Lower Extremity
- Pain, Postoperative
- Analgesics, Opioid
Statistics from Altmetric.com
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.
Adductor canal block (ACB) is one of the most effective analgesic techniques in the setting of total knee arthroplasty (TKA) with important benefits including reduced pain severity and opioid consumption, as well as improved functional recovery compared with traditional pain management strategies. Both single-shot and continuous catheter-based ACB are commonly used as part of a multimodal analgesia regimen following TKA,1 but each technique is associated with unique strengths and limitations; namely, the technical ease, safety, and reliability of a single-shot ACB may be offset by its limited duration of analgesia (and possibly rebound pain) when compared with a continuous catheter-based technique. Indeed, the extent to which the clinically important analgesic benefits of single-shot ACB can be enhanced or prolonged by a continuous catheter-based infusion is unclear as evidence from randomized trials comparing single-shot and continuous ACB has so far been mixed.2–5 Further, previous reviews2–5 on the topic have had mixed results and have paid little attention to the day-after knee arthroplasty (24–48 hours). Therefore, we undertook this systematic review and meta-analysis to compare the analgesic and functional outcomes of single-shot versus continuous ACB for adult patients undergoing TKA. Specifically, we aimed to identify which technique offered the greatest analgesic efficacy, as evaluated by pain severity at rest and opioid consumption during the first 48 hours postoperatively. Secondary outcomes included functional recovery, opioid-related side effects, and block-related complications.
This manuscript was prepared in adherence with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses statement guidelines.6 We sought randomized controlled trials that evaluated analgesic outcomes in adult patients receiving single-shot ACB compared with continuous ACB alone in patients undergoing TKA. The study protocol was registered with the international prospective register of systematic reviews (PROSPERO): CRD42021292738. The study protocol was then modified to extend the outcome assessment time up to 48 hours instead of limiting it to one postoperative day, to allow capturing any differences between the two comparators that would otherwise be missed. The protocol was not published in a peer-reviewed journal.
Randomized trials that randomized adult patients (>18 years old) undergoing unilateral or bilateral TKA to either single-shot ACB or continuous ACB were considered for inclusion. All types of continuous based catheter techniques were included (ie, continuous infusion and intermittent bolus).7 8 We excluded those studies that performed surgery on additional areas other than the knee joint. To account for variety of clinical practice in this patient population, no restriction was placed on (1) trials that incorporated perineural adjuncts (ie, dexamethasone, epinephrine, clonidine, or dexmedetomidine) in the block solution and (2) trials that performed additional regional anesthetic modalities alongside the ACB (ie, infiltration or peripheral nerve block), as long as the block was performed in both intervention and comparator groups. Trials were considered irrespective of whether general anesthesia or neuraxial anesthesia was used. All English and non-English trials were considered for inclusion. Non-English studies were translated using Google online translator.
A systematic search strategy was created by TS for the US National Library of Medicine database (MEDLINE), Excerpta Medica database (EMBASE), and Cochrane Database of Systematic Reviews from inception to November 1, 2021. The strategy contained key words related to the following: ACB, TKA, and analgesia. In addition, the reference lists of potentially eligible citations were manually searched to identify additional trials that fulfilled inclusion criteria. The full search strategy for all databases can be viewed in online supplemental appendix A. All references were stored using EndNote V.20.4. We also electronically searched the US clinical trial registry for additional citations (www.clinicaltrials.gov) in addition to Google Scholar and ResearchGate using the keywords described above. Finally, published abstracts of the following international conferences were also queried: American Society of Anesthesiologists 2011–2021, American Society of Regional Anesthesia and Pain Medicine 2013–2020, and International Anesthesia Research Society 2013–2021.
Selection of included studies
Two reviewers (NH and SZ) screened titles and abstracts generated by the literature search independently. All potentially eligible citations had their full-text versions retrieved and evaluated for inclusion. Disagreements on inclusion of potentially eligible full text were discussed until a consensus was reached. If a final decision for inclusion could not be agreed on by discussion, a third reviewer (FA) made the final decision.
A reviewer (NH) created a standardized data extraction form on Microsoft Excel. Data extraction was carried out by NH and verified by SZ. The data extraction form collected information regarding the following: year of publication; type of surgery; primary outcome of study; ACB technique (including dose, volume, adjuvants, and use of single-shot or catheter technique); anesthetic technique; preoperative, intraoperative and postoperative analgesic regimen; rest pain scores at all time points; analgesic consumption at all reported time intervals; time-to-first analgesic request; postanesthesia care unit (PACU) length of stay; hospital length of stay; time of block performance; postoperative function; opioid-related side effects; and block-related complications. Data reported in graphical form were extracted with graph digitizing software (Engauge Digitizer V.12.1, USA).
Assessment of risk of bias
The risk of bias of the included trials was assessed using the Cochrane Collaboration Risk of Bias 2.0 tool9 by two independent reviewers (NH and SZ). Specifically, risk of bias for each study was rated as having a low risk, some concern, or high risk of bias for the following: bias due to deviations from the intended intervention, bias due to missing outcome data; bias in measurement of the outcome; bias in selection of the reported result; and overall risk of bias.
The overall strength of evidence for each statistically pooled outcome was also assessed using the guidelines created by the Grades of Recommendation, Assessment, Development and Evaluation (GRADE) working group.10 11 Strength of evidence was then classified as being of high-quality (⊕⊕⊕⊕), moderate-quality (⊕⊕⊕⊖), low-quality (⊕⊕⊖⊖), or very low-quality (⊕⊖⊖⊖) evidence.
Primary and secondary outcomes
As the continuous ACB is purported to extend analgesia beyond the first 24 hours postoperatively, the coprimary outcomes evaluated in this systematic review and meta-analysis were the 0–48 hours interval (1) area under the curve (AUC) of postoperative pain severity at rest (measured by the Numeric Rating Scale (NRS)) and (2) postoperative oral morphine equivalent consumption.
Secondary outcomes included AUC of pain at rest between 24 and 48 hours; oral morphine equivalent consumption in PACU and during the 0–24 hours and 24–48 housr intervals; rest pain severity scores in PACU, and at 6, 12, 24, and 48 hours; time-to-first analgesic request (hours); functional recovery; time to PACU discharge (hours); time to hospital discharge (days); block procedure time (minutes); postoperative opioid-related side effects (excessive sedation, postoperative nausea and vomiting, respiratory depression, pruritus, urinary retention, or constipation); and block-related complications (block failure, nerve injury, hematoma, catheter dislodgement, or local anesthetic systemic toxicity).
Measurement of outcome data
All postoperative pain score data were converted to an equivalent score on 0–10 cm NRS, with 0 cm corresponding to no pain and 10 cm corresponding to worst pain imaginable. All analgesic consumption data were converted to equivalent doses of oral morphine in milligrams. For postsurgical functional recovery, based on prior work,12 we expected a variety of measurements to be used. Therefore, we pooled the various measures according to type of data reported (see below). Finally, all time-to-event data related to duration of PACU stay were converted to hours, duration of hospital stay was converted to days, and block procedure time was converted to minutes.
The mean and SD were extracted for all continuous outcome data. When these were not available, the median and IQR, median and range, or mean and 95% CI were used to approximate these values.13 The median was used to approximate the mean in situations when statistical conversions could not be made.14 Dichotomous outcome data related to severity of total knee pain and adverse events were converted to overall incidence numbers.
For pain data that was presented at time points that differed from our prespecified intervals (PACU, 6, 12, 24, and 48 hours), we included this data only if it was within 2 hours of the predefined time point. For studies that only reported overall analgesic consumption between 0 and 48 hours time interval, we extrapolated the 0–24 and 24–48 hours data using the percent distribution of the weighted means from studies that did provide data on these prespecified intervals.
In anticipation of the diversity of scales used to assess postoperative function, we a priori planned to report: (1) the weighted mean difference (WMD) of all studies that used the same continuous scale; (2) the log(OR) if trials reported a blend of continuous outcomes, binary outcomes, and used different scales that measured the same theme; and (3) an OR if all trials reported only binary outcomes. If scenario (ii) was applicable, the conversion to log(OR) from a standardized mean difference (SMD) was done using the formula log(OR) = SMD (π/√3).15 16
Statistical pooling was conducted for outcomes with data from at least three included studies. For all outcome data, pooling was done using the Hartung-Knapp-Sidik-Jonkman (HKSJ) method for random effects.17 For the coprimary outcomes of this review (AUC0-48 pain score and 0–48 hours opioid consumption) a WMD or SMD and HKSJ 95% CI was calculated. The decision to calculate a WMD or SMD was based off the clinical variability observed across included studies. A two-tailed p<0.025 was designated as the threshold for statistical significance for the two primary outcomes.
A WMD or SMD with a HKSJ 95% CI or an OR with HKSJ 95% CI was calculated for all continuous and dichotomous secondary outcomes, respectively. A two-tailed p<0.05 was also designated as the threshold for statistical significance.
For functional recovery, either a WMD, log(OR), or an OR with HKSJ 95% CI was calculated, depending on the nature of reported data (as described above).
Finally, additional post hoc analysis was conducted by calculating a 95% prediction interval18 for the two coprimary outcomes of this review. This provides a range of estimates of effects that facilitate powering a future randomized on the topic when taking into account heterogeneity.
The interpretation of all pain scores presented in this review was based on the minimal clinically important difference (MCID).20 For a single time point, the MCID of NRS acute postoperative pain has been reported to be 2.26 cm.21 For the 0–48 hours AUC analysis, this was extrapolated to 9.04 cm.hour (four time intervals, 0–6, 6–12, 12–24, and 24–48 hours). Postoperative opioid consumption was interpreted in the context of an MCID equivalent to 30 mg of oral morphine in the first 24 hours postsurgery.
Assessment of inconsistency
The extent of statistical inconsistency was evaluated for all pooled outcomes by calculation of an I2 statistic, with significant inconsistency considered to be present with I2 values >50%0.13 When significant inconsistency was present in the coprimary outcomes of this review, a priori sensitivity analysis was performed by sequential exclusion based on: (1) using a continuous ACB infusion for less than 48 hours; (2) use of nerve block adjuncts (ie, dexamethasone or epinephrine); (3) use of an opioid-based (unimodal) analgesic techniques; (4) timing of single-shot and continuous ACB (preoperative vs postoperative). Post hoc sensitivity analysis was performed by sequential exclusion of studies that did not perform additional peri or local infiltration analgesia. Sensitivity analysis was only carried out for those outcomes with three or more studies.
We conducted additional post hoc sensitivity on the primary outcomes by sequentially removing each study from the pooled analysis once (ie, influence analysis). We further conducted a post hoc outlier analysis by excluding only those studies that had an individual 95% CI outside the pooled HKSJ 95% CI.
Assessment of small study effects
The risk of small study effects was assessed for all statistically pooled outcomes by creating a funnel plot.22 We created additional post hoc Doi plots and quantitatively assessed for small-study effects by the Luis Furuya–Kanamori (LFK) index. The LFK index suggests possible small-study effects and plot asymmetry when values exceed ±1.00.23
Funnel and forest plots were created using Review Manager Software (RevMan V.5.2; Nordic Cochrane Center, Cochrane Collaboration). Risk of bias plots were created using the robvis tool.24 Doi plots and LFK index was calculated using MetaXL V.5.3 (EpiGear International, Sunrise Beach, Australia)
The search strategy identified 1705 citations, and after exclusion of duplicates, a total of 569 were screened based on title and abstract alone. From these, 515 citations were excluded and 54 had their full-text versions retrieved for eligibility. An additional 43 full-text citations were subsequently excluded due to ineligible comparator/intervention (n=24); ineligible study design (n=17); and duplicate citation (n=2) (online supplemental appendix D). Two additional trials25 26 were identified by clinical trials registry hand search; but data were unavailable for these trials. Thus, a total of 11 randomized trials27–37 were included from the full-text search strategy. Of the 11 trials included, 6 were prospectively registered,28 29 31 33 35 36 and the remaining 5 were either completed or published prior to 2019.27 30 32 34 37 The full-flow diagram for study inclusion can be viewed in figure 1.
The characteristics and outcomes of the 11 trials included in this review are presented in tables 1 and 2.27–37 Studies were conducted across a variety of centers worldwide, including Canada, China, Denmark, India, South Korea, Turkey, and USA. None of the trials were industry sponsored. All trials included patients with an average age greater than 60 years and the majority had a higher percentage of female participants. Across all trials, there were a total of 1185 patients of which 582 received single-shot ACB, and 578 received continuous ACB. Six of these studies performed the surgery using neuraxial anesthesia,27 30 31 33–35 3 studies used a general anesthetic,29 32 37 and 2 used either spinal or general anesthesia.28 36 Ten of the trials assessed rest pain scores at early (<24 hours) or late (>24 hours) time intervals.27–29 31–37 Six of trials assessed opioid consumption in the early postoperative period (<24 hours).28 29 31 32 34 37 Opioid-related adverse events were evaluated by eight trials,27 29 31 32 34–37 while block-related complications were assessed by nine.27 29–32 34–37
The single-shot and continuous ACB techniques and analgesic regimens used by the included trials are described in table 3. All of the blocks were completed under ultrasound guidance,27–37 but only five assessed block success.28 33 36 37 Timing of single-shot ACB was preoperative in five studies,28 29 35–37 while six trials performed it postoperatively.27 30–34 For continuous ACB, the catheter was activated preoperatively in one study29 postoperatively in nine,27 28 30–34 36 37 and one study did not specify the time.35 The duration of catheter infusion was 24 hours in two studies,27 37 48 hours in seven30–36; either 24 or 48 hours in one28; and one study did not specify the duration.29 One study performed an additional sciatic nerve block in all patients,36 and eight studies performed periarticular/local infiltration analgesia alongside ACB.28–31 33 34 36 37 Finally, preincisional preventive analgesia was provided to patients in eight studies28 30–33 35–37; and postoperative analgesia was multimodal in ten studies27 28 30–37 and opioid-based in one.29
The risk of bias assessment for included studies can be viewed in figure 2. Briefly, all studies were classified as either having some concerns or a low risk of bias. Six studies were preregistered with clinical trial registries.28 29 31 33 35 36 . The majority of studies had a low risk of bias arising from the randomization process, deviations from the intended intervention, missing outcome data, and in the measurement of outcome data.
Analgesic consumption between 0 and 48 hours
Eight studies27–29 31–33 36 37 inclusive of 711 patients (Single-shot ACB: 387, Continuous ACB: 324) reported 0–48 hours analgesic consumption. Overall, no significant differences were observed between the two groups for this outcome (SMD (HKSJ 95% CI) of 0.28 (–0.47 to 1.03) (p=0.41, I2=85%)) (figure 3), and the 95% prediction interval was found to be −1.30 to 1.86. This analysis was characterized by substantial inconsistency (I2=85%), justifying sensitivity analysis. The sensitivity analysis conducted maintained a robust treatment effect and successfully resolved inconsistency, which was attributed to (1) use of opioid-based postoperative analgesic regimen29 (SMD (HKSJ 95% CI) of 0.00 (–0.13 to 0.13) (p=1.0, I2=0%)); and (2) preoperative block timing28 29 36 37 (SMD (HKSJ 95% CI) of −0.07 (–0.21 to 0.07) (p=0.24, I2=0%)). The other factors examined (using continuous ACB for less than 48 hours, including nerve block adjuncts, and not using local infiltration analgesia) did not influence inconsistency. Post hoc influence and outlier analysis revealed that the original estimate of effect remained robust to the exclusion of each of the studes, as well as the one the one study29 that had a 95% CI outside the HKSJ 95% CI for the estimate of effect (−0.47 to 1.03); heterogeneity resolved (I2=0%) with removal of this specific study. The risk of small study effects was low (online supplemental appendix B), and additional post hoc assessment of small study effects further revealed an LFK index of 1.86, suggestive of minor asymmetry in the Doi plot (online supplemental appendix C). Thus, the GRADE strength of evidence was high in the pooled estimate (table 4).
AUC rest pain scores between 0 and 48 hours
Four studies27 29 32 33 inclusive of 325 patients (single-shot ACB: 161, continuous ACB: 164) reported rest pain scores at all time points between 0 and 48 hours. Overall, no significant differences were observed between the two groups for this outcome (SMD (HKSJ 95% CI) of 2.18 (–2.38 to 6.74) (p=0.22, I2=96%)), (figure 4) and the 95% prediction interval was found to be −9.43 to 13.79. This analysis was characterized by substantial inconsistency (I2=96%), but sensitivity analysis failed to resolve inconsistency for this outcome. Post hoc influence analysis revealed that the original estimate of effect remained robust, but inconsistency remained unresolved, to the exclusion of each of the studies; post hoc outlier analysis was not conducted since none of the studies had a 95% CI outside the boundaries of the HKSJ 95% CI for the estimate of effect (2.38 to 6.74). The risk of small study effects was low with visual inspection of the funnel plot (online supplemental appendix B); however, additional post hoc assessment of small study effects further revealed an LFK index of 3.11, suggestive of major asymmetry in the Doi plot (online supplemental appendix C). Thus, the GRADE strength of evidence was low due to unresolved inconsistency and risk of small study effects in the pooled estimate (table 4).
Analgesic consumption in PACU
One study29 reported this outcome, and qualitatively no difference between the two groups was observed.
Analgesic consumption between 0 and 24 hours
Nine studies27–29 31–34 36 37 inclusive of 796 patients (single-shot ACB: 426, continuous ACB: 370) reported 0–24 hours analgesic consumption. Overall, no significant difference was observed between the two groups for this outcome (SMD (HKSJ 95% CI) of 0.25 (–0.40 to 0.90) (p=0.39, I2=85%)) (table 5). The risk of small study effects was low (online supplemental appendix B) and the GRADE strength of evidence was moderate due to inconsistency in the pooled estimate (table 4).
Analgesic consumption between 24 and 48 hours
Eight studies27–29 31–33 36 37 inclusive of 711 patients (single-shot ACB: 387, continuous ACB: 324) reported 24–48 hours analgesic consumption. Overall, no significant difference was observed between the two groups for this outcome (SMD (HKSJ 95% CI) of 0.12 (–0.19 to 0.43)(p=0.40, I2=52%)) (table 5). The risk of small study effects was low (online supplemental appendix B) and the GRADE strength of evidence was moderate due to inconsistency in the pooled estimate (table 4).
AUC rest pain scores between 24 and 48 hours
Eight studies27 29 31–36 inclusive of 761 patients (single-shot ACB: 408, continuous ACB: 353) reported 24 and 48 hours rest pain scores, which allowed for AUC analysis. Overall, no significant difference was observed between the two groups for this outcome (SMD (HKSJ 95% CI) of 0.75 (–0.09 to 1.59) (p=0.07, I2=89%)) (table 5). The risk of small study effects was low (online supplemental appendix B) and the GRADE strength of evidence was moderate due to inconsistency in the pooled estimate (table 4).
Rest pain scores at individual time points
Rest pain scores was assessed by four studies27 29 32 33 in PACU (single-shot ACB: 161, continuous ACB: 164); six studies27 29 32–36 at 6 hours (single-shot ACB: 288, continuous ACB: 296); five studies27 31 32 34 36 at 12 hours (single-shot ACB: 279, continuous ACB: 225); nine studies27 29 31–37 at 24 hours (single-shot ACB: 433, continuous ACB: 376); and nine studies27–29 31–36 at 48 hours (single-shot ACB: 461, continuous ACB: 395). Overall, no significant difference was observed between the two groups for this outcome in PACU (p=0.12) and at 12 (p=0.21) and 24 hours (p=0.14) postoperatively. However, continuous ACB seemed to improve rest pain at 6 and 48 hours by an SMD of (HKSJ 95% CI) of 1.52 (0.05 to 2.99)(p=0.04, I2=96%) and 1.19 (0.56 to 2.32) (p=0.04, I2=93%), respectively (table 5). Nonetheless, these difference did not reach the preset threshold for clinical importance (2.26 cm). The risk of small study effects was low for each time point (online supplemental appendix B) and the GRADE strength of evidence was moderate for PACU, 12, 24, and 48 hours due to inconsistency, but low for 6 hours due to inconsistency and imprecision in the pooled estimates (table 4).
Attempting to explore the potential cause of inconsistency of results between the 6 and 48 hours rest pain assessments and the remaining time points (PACU, 12, and 24 hours), we resorted to sensitivity analysis. Our sensitivity analysis suggested that the treatment effect at 6 and 48 hours was driven by differences in block timing and was attributed to potentially biased comparisons between continuous ACB activated with postoperatively versus preoperative single-shot ACB. This difference was rendered non-significant at 6 hours (SMD (HKSJ 95% CI) of 1.31 (–0.28 to 2.90) (p=0.08, I2=96%) and 48 hours (SMD (HKSJ 95% CI) of 0.88 (–0.07 to 1.83) (p=0.06, I2=93%) when studies that performed preoperative single-shot ACB were excluded28 29 36 from analyses, and the comparisons were limited to postoperative blocks.
Time-to-first analgesic request
Two studies31 36 inclusive of 236 patients (single-shot ACB: 150, continuous ACB: 86) reported this outcome. Qualitatively both groups did not observe a significant difference between the two groups.
Ten studies27–30 32–37 inclusive of 776 patients (single-shot ACB: 386, continuous ACB: 390) reported this outcome. Studies assessed function using a variety of tools and the decision to include a measurement for pooling was based on overall instrument objectivity. Data was pooled for the (1) timed up and go test27 28 33–35; (2) ability to ambulate for 40 meters30; (3) knee range of motion32 37; (4) ability to perform straight leg test36; and (5) Medical Research Council Scale of Quadriceps Strength.29 Further, since both continuous and dichotomous scales that measured the same theme were reported, a log(OR) (HKSJ 95% CI) was calculated. Overall, no difference was observed between the two groups for this outcome (log(OR) (HKSJ 95% CI) of 0.04 (–0.79 to 0.89)(p=0.91, I2=86%)) (table 5). The risk of small study effects (online supplemental appendix B) and GRADE strength of evidence was low due to inconsistency and indirectness (table 4).
Time to PACU discharge
One study28 reported this outcome, and qualitatively no difference between the two groups was observed.
Time to hospital discharge
Eight studies27 28 30–32 34 36 37 inclusive of 715 patients (single-shot ACB: 386, continuous ACB: 329) reported this outcome. Overall, no difference was observed between the two groups for this outcome (SMD (HKSJ 95% CI) of 0.21 (0.00 to 0.42) (p=0.05, I2=22%)) (table 5). The risk of small study effects was low (online supplemental appendix B), and the GRADE strength of evidence was high (table 4).
Block procedure time
One study37 reported this outcome in patients that received a single-shot ACB (no sham catheter) versus continuous ACB. This study reported a faster procedure time for single-shot ACB.
Postoperative opioid-related side effects
Eight studies27 29 31 32 34–37 inclusive of 712 patients (single-shot ACB: 384, continuous ACB: 328) reported this outcome. Overall, no difference (OR (HKSJ 95% CI) of 0.79 (0.50 to 1.25) (p=0.23, I2=0%)) was observed between the two groups in the odds of developing postoperative nausea or vomiting (table 5). One study32 also assessed degree of drowsiness and found no difference between the two groups. The risk of small study effects was low (online supplemental appendix B), and the GRADE strength of evidence was high (table 4).
Ten studies27–32 34–37 inclusive of 884 patients (single-shot ACB: 469, continuous ACB: 415) reported this outcome. Overall, the OR (HKSJ 95% CI) of experiencing block-related complications was 0.24 times lower (0.14 to 0.41) (p=0.002, I2=0%) in the single-shot ACB versus continuous ACB group (table 5). Reported complications in the continuous ACB group included catheter dislodgement (n=6 patients),30 32 36 37 block failure (n=1 patient),28 and femoral vascular injury (n=1 patient). The risk of small study effects was low (online supplemental appendix B), and the GRADE strength of evidence was high (table 4).
Overall, the results of our review suggest that single-shot ACB provides equivalent analgesic and functional benefits when compared with continuous ACB for TKA over the first 48 hours postoperatively. Specifically, continuous-catheter-based ACB appears to provide no benefit for postoperative analgesic consumption, AUC of rest pain severity scores, or functional recovery. Further, continuous ACB does not appear to provide statistically significant improvements in rest pain severity at any time point evaluated. Finally, while no differences were present in the risk opioid-related side effects, continuous ACB was found to be associated with a higher rate of block-related complications, mainly catheter dislodgement. Taken together, the results of our meta-analysis do not support the routine use of continuous catheter-based ACB over single-shot ACB for postoperative analgesia after TKA. Nonetheless, continuous ACB may still have benefit in patients with opioid tolerance or pre-existing chronic pain; such patients are routinely excluded from most studies.
Implications for research
Controlling postoperative pain has been a central focus of rehabilitation following TKA. While regional anesthetic modalities have evolved for this surgery,38 the ACB remains a commonly used technique due to its analgesic benefits and motor sparing properties. Previous reviews2–5 on the topic have been mixed and limited due to several reasons. First, little attention has been paid to the day-after knee arthroplasty (24–48 hours), despite its impact on early effective physiotherapy and postoperative function. In fact, all prior reviews2–5 have focused on the first-postoperative day, and thus some have concluded superiority of the continuous ACB. The consideration that the observed superiority is transient, limited to the first postoperative day, and not apparent with physiotherapy is noticeably absent. Second, none of the prior reviews2–5 evaluated the clinical significance of differences between the two modalities using context specific thresholds. Finally, all prior reviews2–5 have largely overlooked the impact of each modality on functional outcomes, which is imperative for adequate postoperative recovery.
Implications for practice
ACB is thought to target the saphenous nerve and nerves to the vastus medialis, thereby covering the anteromedial aspect of the knee and preserving quadriceps function.38 39 However, much debate remains regarding the benefit of a single-shot versus continuous catheter, and as our review suggests, the analgesic differences may be minimal. Due to the relatively short duration of single-shot nerve blocks, continuous catheter-based techniques emerged as a theoretically superior analgesic option. However, as they are technically more challenging and are associated with higher failure rates,40 41 their role may be limited. Typical analgesic regimens for TKA are now increasingly multimodal and incorporate preoperative medications such as acetaminophen or COX-inhibitors, local infiltration or peripheral nerve blockade, and neuraxial anesthesia to promote analgesia and rehabilitation.38 42 Overall, our review suggests that a multimodal analgesic regimen alongside single-shot ACB, appears to provide equivalent pain control to continuous catheter-based ACB.
For clinical practice, our results should also be interpreted in light of the GRADE recommendations. Of note, while high certainty in evidence was present for analgesic consumption between 0 and 48 hours, the same cannot be said for AUC pain during the same timeframe, whereby certainty was lower. Our concerns for this outcome lie in the limited assessment of pain scores throughout the 0–48 hours time period. This is especially important to assess for future trails as catheter-based techniques should ideally provide optimal pain control beyond the first postoperative day to be considered a worthy intervention for the anesthesiologist.
Strengths and limitations
Our meta-analysis comes with several strengths. First, our comprehensive search strategy was able to identify both English and non-English citations to be included this review. This enabled us to include 11 trials in the meta-analysis and provide estimates of effects for several clinically meaningful outcomes. Second, as a result of the larger number of studies included, we were able to focus our analysis on day 1 and day 2 analgesic outcomes after TKA. Third, we were able to interpret the results of our meta-analysis using context-specific thresholds of MCID for TKA. Fourth, our results were interpreted in light of the HKSJ method for pooling, which improves external validity in the setting of substantial inconsistency. Fifth, we interpreted our primary outcomes in light of the 95% prediction interval, which appear to suggest that future studies may identify a statistically significant and clinically important difference in AUC of rest pain between single-shot and continuous ACB in the 0–48 hours interval; but are unlikely to detect a significant difference in analgesic consumption during the same timeframe. Finally, we were able to successfully resolve the inconsistency in the coprimary outcome of analgesic consumption at 24–48 hours by exclusion of the only study that used a unimodal29 opioid-based analgesic regimen.
Our review also comes with limitations that are worth noting. First, because of observed clinical variability in the included trials, we calculated an SMD instead of an MD. This may create difficulties in the overall interpretation of our results. Second, although inconsistency was reduced with sensitivity analysis in the coprimary outcome of AUC pain scores, we were still unable to fully resolve its occurrence. Similarly, other outcomes in our review were also noted to have substantial inconsistency. In relation to this, and inherent to meta-analyses, our analysis is also at risk of Simpson’s paradox43 whereby associations observed in subgroup analyses may not be present in the primary analysis. Source studies reflected some variations in clinical practices, particularly in the timing, doses and infusion modes of examined interventions, as well as in the consistency of assessment of sensory block onset. Third, trials included in this meta-analysis varied regarding their duration of ACB infusion ranging from 24 to 72 hours. While many studies used a 48-hour infusion, those that used a shorter duration may have affected the results which were not identified with sensitivity analysis. Fourth, while an important outcome, we were unable to assess the impact of continuous ACB on the incidence of chronic postsurgical knee pain. Finally, we were unable to assess the impact of preoperative continuous ACB infusion, and it is possible that beginning the local anesthetic infusion prior to surgery may improve postoperative analgesic outcomes.
Our results suggest that the benefits of using continuous ACB are mixed and may largely be clinically unimportant over the first 48 hours postoperatively. Specifically, continuous techniques appear to be no different than single-shot ACB for analgesia, AUC pain at rest, functional outcomes, and opioid-related side effect outcomes. Further, the continuous ACB appears to be associated with a higher rate of block related complications such as catheter dislodgement. Taken together, the results of our meta-analysis do not support the routine use of continuous catheter-based ACB for postoperative analgesia after TKA. Nonetheless, continuous ACB may still have benefit in patients with opioid tolerance or pre-existing chronic pain.
Patient consent for publication
Twitter @colinjmccartney, @Faraj_RegAnesth
Contributors All authors contributed to this manuscript to warrant authorship.
Funding The authors have not declared a specific grant for this research from any funding agency in the public, commercial or not-for-profit sectors.
Competing interests None declared.
Provenance and peer review Not commissioned; externally peer reviewed.
Supplemental material This content has been supplied by the author(s). It has not been vetted by BMJ Publishing Group Limited (BMJ) and may not have been peer-reviewed. Any opinions or recommendations discussed are solely those of the author(s) and are not endorsed by BMJ. BMJ disclaims all liability and responsibility arising from any reliance placed on the content. Where the content includes any translated material, BMJ does not warrant the accuracy and reliability of the translations (including but not limited to local regulations, clinical guidelines, terminology, drug names and drug dosages), and is not responsible for any error and/or omissions arising from translation and adaptation or otherwise.