Journal of Clinical Anesthesia 42 (2017) 6976

Contents lists available at ScienceDirect

Journal of Clinical Anesthesia

Original Contribution

Programmed intermittent peripheral nerve local anesthetic bolus

compared with continuous infusions for postoperative analgesia: A
systematic review and meta-analysis
Matthew A. Chong, MD ,1, Yongjun Wang, BScPharm 2, Shalini Dhir, MD 3, Cheng Lin, MD 4
Department of Anesthesia and Perioperative Medicine, Western University, London, Ontario, Canada

a r t i c l e i n f o a b s t r a c t

Article history: Study objective and background: The role of the programmed intermittent bolus (PIB) technique for infusion of
Received 29 July 2017 local anesthetics in continuous peripheral nerve blockade (CPNB) remains to be elucidated. Randomized
Received in revised form 13 August 2017 controlled trials (RCTs) on PIB versus continuous infusion for CPNB have demonstrated conicting results and
Accepted 15 August 2017 no systematic review or meta-analysis currently exists. We aimed to delineate via systematic review with
Available online xxxx
meta-analysis if there is any analgesic benet to performing PIB versus continuous infusion for CPNB.
Design: We conducted a systematic review and random-effects meta-analysis of RCTs.
Keywords:
Anesthetics, local
Data sources: We searched Medline, Embase, and the Cochrane Library without language restriction from
Regional anesthesia inception to 2-May-2017.
Eligibility criteria: Included RCTs had to compare PIB to continuous infusion in adult surgical patients receiving
any upper or lower limb CPNB for postoperative analgesia. VAS pain scores were the primary outcome. The
Cochrane Risk of Bias Tool with GRADE methodology was utilized to assess evidence quality.
Results: Nine RCTs (448 patients) met the inclusion criteria. Two studies performed upper limb blocks and
the rest lower limb blocks. Five RCTs activated the CPNB with long-acting local anesthetic and only ve
used multi-modal analgesia. PIB modestly reduced VAS pain scores at 6 h ( 14.2 mm; 95%CI 23.5 mm to
5.0 mm; I2 = 82.5%; p = 0.003) and 12 h ( 9.9 mm; 95%CI 14.4 mm to 5.4 mm; I2 = 12.4%; p b
0.001), but not at later time points. There were no other meaningful differences in the rest of the outcomes,
apart from more residual motor block with PIB (OR 4.27; 95% CI 1.0816.9; p = 0.04; NNTH = 8). GRADE scoring
ranged from low to very low.
Conclusions: The existing evidence demonstrates that PIB does not meaningfully reduce VAS pain scores in CPNB.
This systematic review provides important information about the limitations of existing studies. Future studies
should reect contemporary practice and focus on more painful operations.
2017 Elsevier Inc. All rights reserved.

1. Introduction

Continuous infusions of local anesthetics are widely utilized in re-

gional anesthesia to provide catheter-based postoperative analgesia
Financial disclosures and conicts of interest: This research did not receive any
specic grant from funding agencies in the public, commercial, or not-for-prot sectors.
after painful surgeries [1]. Recently, there has been high interest in a
The authors do not have any conicts of interest. novel infusion strategy termed the programmed intermittent bolus
Corresponding author at: Department of Anesthesia & Perioperative Medicine, (PIB) technique. PIB differs from continuous infusion insofar as the
University Hospital (London Health Sciences Centre), 339 Windermere Road, C3-108, hourly block volume is given as a bolus, rather than infused continuous-
London, Ontario N6A 5A5, Canada.
ly. Such bolus administration is thought to result in better spread of local
E-mail addresses: matthew.chong@lhsc.on.ca (M.A. Chong),
Shalini.Dhir@sjhc.london.on.ca (S. Dhir). anesthetic around the targeted nerves [2]. Indeed, benets of PIB have
1
Contribution: MC conceived the idea for the meta-analysis, designed the search strat- been demonstrated for labor analgesia, where PIB results in better
egy, selected study articles, performed the risk of bias assessment, extracted the data, per- patient satisfaction and less local anesthetic consumption [3].
formed the data analysis, created the gures/tables, and prepared the manuscript. Despite the clinical effectiveness of PIB in labor analgesia, the evi-
2
Contribution: YW assisted with the data extraction and manuscript preparation.
3
Contribution: SD reviewed and edited the manuscript.
dence for this technique in regional anesthesia is unclear. Numerous
4
Contribution: CL selected study articles, extracted data, performed the risk of bias as- randomized controlled trials (RCTs) have been performed and they
sessment, and revised the nal manuscript. have demonstrated conicting results as to whether PIB is superior to

http://dx.doi.org/10.1016/j.jclinane.2017.08.018
0952-8180/ 2017 Elsevier Inc. All rights reserved.
70 M.A. Chong et al. / Journal of Clinical Anesthesia 42 (2017) 6976
continuous infusion in the setting of continuous peripheral nerve block-
ade (CPNB) for postoperative analgesia [4,5]. The elucidation of whether
or not PIB is of benet is important: on one hand, PIB may result in bet-
ter patient outcomes, as suggested by the benets in the labor analgesia
literature; on the other hand, PIB requires specialized pumps that are
capable of giving the automated boluses and, therefore, may be costly
to adopt [6]. Finally, we are not aware of any existing systematic review
and meta-analysis that synthesizes the evidence for PIB in regional
anesthesia. Therefore, given this gap in the literature and clinical equi-
poise, we designed and conducted a systematic review and meta-anal-
ysis of RCTs to elucidate the effect of PIB on patient centered outcomes
in CPNB for postoperative analgesia. Based on the benets in the labor
analgesia literature, we hypothesized that PIB would result in better
analgesia with less local anesthetic and rescue opioid consumption.
2. Methods
This systematic review and meta-analysis complies with the PRISMA
statement [7]. Our institutional research ethics board does not require
approval for systematic reviews and meta-analyses, as no data is being
collected from patients. The clinical question, study inclusion criteria,
outcomes, and analysis plan were dened a priori.
2.1. Literature search
Ovid Medline, Embase, and the Cochrane Library were searched
from inception to 2 May 2017 without language restriction for RCTs
meeting the following inclusion criteria.
2.2. Study selection criteria
2.2.1. Population
Studies had to recruit adult surgical patients receiving any CPNB
with catheter placement. Neuraxial blockade was not considered part
of the scope of the review. Furthermore, we did not include studies
recruiting healthy volunteers, such as subjects who volunteer to have
a nerve catheter placed without undergoing surgery.
2.2.2. Intervention and control
To be included, the trial had to compare PIB to continuous infusion
for delivery of the local anesthetic solution. The dosing of the local
anesthetic had to be mathematically similar (e.g. 5 mL per hour in the
continuous arm versus 10 mL every 2 h in the PIB arm was considered
acceptable). We excluded studies where the local anesthetic solution
volume or dose differed between study arms or where the patient was
relied upon to administer the intervention boluses (e.g. the PIB was
not truly programmed).
2.2.3. Outcomes
The primary outcome was visual analogue scale (VAS) pain scores
on a 100 mm scale. Secondary outcomes included opioid and local anes-
thetic consumption, patient satisfaction, rescue analgesia requirement,
side effects (e.g. nausea and vomiting), and block-related complications.
2.2.4. Search strategy
The full search strategy is available in the Appendix (Supplemental
Digital Content 1, which lists search terms for each database). The
search utilized a comprehensive combination of medical subject head-
ing (MeSH) terms, free-text terms, and corresponding synonyms. The
reference lists of included articles were manually searched for addition-
al studies. We also queried the clinicaltrials.gov database for additional
and on-going studies, but did not seek unpublished data.
2.3. Article screening and data extraction
All titles, abstracts, and full texts (where required to assess the study
for inclusion) were reviewed in duplicate by MC and CL. Any disagree-
ments were resolved by consensus with SD. Data from included studies
were extracted independently onto standardized forms by MC, YW, or
CL. Extracted data included important baseline demographic informa-
tion of each study, information regarding assessment of the risk of
bias of the study, and the pre-specied outcomes. To facilitate meta-
analysis, medians, IQR, and range values were approximated into
means and their corresponding standard deviation using methods
suggested by the Cochrane Library [8]. Where necessary (e.g. data
values not reported in text and only within graphs), numerical data
were extracted from graphs by digital measurement. We attempted to
contact principal investigators of included studies for additional infor-
mation, where necessary.
2.4. Risk of bias evaluation
The Cochrane Risk of Bias Tool was utilized to appraise each included
study's risk of bias by MC and CL and all discrepancies were resolved by
consensus [9]. We considered studies to be at low risk of bias if they
scored 3 or higher on these criteria: (1) appropriately generated
the randomization sequence, (2) appropriate allocation concealment,
(3) blinded study personnel and participants, and (4) blinded outcome
assessors, and (5) reported data completely. Furthermore, the study had
to demonstrate no signicant selective reporting bias or other source of
bias [9]. Finally, GRADE methodology was utilized to provide an overall
appraisal of the quality of evidence underlying each outcome [10].
2.5. Statistical analysis
Statistical analysis was carried out in Stata (Version 13.1 by
StataCorp, College Station, Texas). Standard summary measures were
generated with the weighted mean difference (WMD) or standardized
mean difference (SMD) for continuous data and odds ratios (ORs) for
binary data, with their corresponding 95% condence intervals (CIs)
and an = 0.05. All analyses were carried out using a random-effects
model. For multiple comparisons over time (e.g. VAS pain score data),
a Bonferroni correction was applied. Furthermore, to account for
heterogeneity in reporting of pain scores, we converted 10-point and
ordinal pain scales to a 100 mm VAS. Only one study utilized an ordinal
scale (3 points) that had to be converted in this manner [11]. The I2
statistic was utilized to quantify heterogeneity. We interpreted an I2
value of 025% as low heterogeneity, 2550% as moderate heterogene-
ity, and N50% as high heterogeneity. The continuity correction was
utilized for zero event studies [12]. A funnel plot was constructed for
the primary outcome and Egger's regression performed to assess for
statistical evidence of publication bias.
The pre-specied subgroup analyses included block technique
(ultrasound-guided versus nerve stimulator versus landmark), type of
local anesthetic (short-acting versus intermediate-acting versus long-
acting), use of concurrent multimodal analgesia (e.g. acetaminophen
and/or non-steroidal anti-inammatory agents), study quality (high
versus low risk of bias), block location (upper versus lower limb),
usage of patient-controlled demand boluses (patient-controlled region-
al analgesia [PCRA]), and type of catheter (e.g. single versus multi-
orice).
3. Results
3.1. Literature search and study selection
Our search strategy initially captured 226 citations and 9 RCTs (448
patients) ultimately met the inclusion criteria (Fig. 1, PRISMA Flow
Chart). Notable study exclusions included one RCT where the patients
 M.A. Chong et al. / Journal of Clinical Anesthesia 42 (2017) 6976
Fig. 1. PRISMA owsheet describing the study selection process. Three databases (MEDLINE, EMBASE, and Cochrane) were searched for relevant articles and screened against our inclusion
criteria.
were relied upon to administer the intermittent boluses (e.g. the inter-
vention boluses in that study were not programmed) [13].
3.2. Study characteristics
The baseline characteristics of the included studies are listed in Table
1 and outcomes reported by each study are shown in a supplementary
table (Supplemental Digital Content 2). For block placement, ultrasound
guidance [5,1416] and neurostimulation [4,11,17,18] were used in 4
studies each, and placement technique was not reported by the last
study [19]. Seven studies activated the nerve block with a bolus dose
of local anesthetic: two studies used mepivacaine [4,17] and 5 utilized
a long-acting local anesthetic (Table 1) [5,11,15,16,18]. In two studies,
the block activation solution was not reported [14,19]. All studies uti-
lized ropivacaine or levobupivacaine for the maintenance dosing of
the CPNB. Certain studies focused on less painful surgeriessuch as hal-
lux valgus surgerywith a full listing of details available in Table 1 [4,14,
17]. Interestingly, only 4 studies employed PCRA boluses [4,5,14,19],
with the rest utilizing rescue analgesia via other means (e.g. IV opioid
71
patient-controlled analgesia [PCA]) [11,1518]. Finally, only 5 studies
reported administration of multi-modal analgesia [4,5,1517]. We con-
sidered 6 of 9 studies to be at low risk of bias (Fig. 2, Risk of Bias Assess-
ment) [4,5,11,1517].
All studies could be pooled for meta-analysis except for one, where
the authors did not report the number of patients assigned to each
arm [14]. This small study of 55 patients scheduled for hallux valgus sur-
gery found marginally decreased local anesthetic consumption favoring
PIB, but otherwise no other differences in pain scores, patient satisfac-
tion, or side effects [14].
3.3. Primary outcome: VAS pain scores
Patients receiving PIB had lower pain scores in the early postopera-
tive period (Fig. 3). Notably, the p-value threshold for this analysis
(given the repeated comparisons) was set via Bonferroni correction at
p b 0.007 (initial = 0.05 divided by 7 comparisons). Specically, PIB
patients experienced less pain at 6 h ( 14.2 mm; 95% CI 23.5 mm
to 5.0 mm; p = 0.003; I2 = 82.5%), 12 h ( 9.9 mm; 95% CI
 72
Table 1
Baseline characteristics of included studies.

Demographic details Block details Type of local Other intervention details

anesthetic
Trial Year Total (n) Type of surgery Type of block Block and catheter Block activation solution Infusion rate for control Rescue analgesia Catheter Use of

M.A. Chong et al. / Journal of Clinical Anesthesia 42 (2017) 6976

placement technique group (bolus volume orices MMA
(U/S or NS) and interval for PIB
group)

De Meyer (Abstract) 2016 55 Hallux valgus surgery Subparaneural U/S Levobupivacaine NR 5 mL per hour (9.8 mL PCRA (6 mL per 30 min) NR NR
[14] sciatic (0.125%) per 2 h)
Hamdani [5] 2014 101 Major shoulder surgery Interscalene U/S Ropivacaine (0.2%) 20 mL of ropivacaine 4 mL per hour (4 mL PCRA (5 mL per 30 min) Single Yes
0.25% and lidocaine 0.5% per hour)
Hillegass [18] 2013 45 Total knee arthroplasty Femoral Both Ropivacaine (0.2%) 20 mL of ropivacaine 0.5% 10.1 mL per hour (5 mL IV PCA Single NR
per 30 min)
Mezzatesta [11] 1997 20 Upper limb reconstructive Axillary NS Bupivacaine (0.25%) 2 mg/kg of bupivacaine 0.25 mg/kg per hour IM opioids Multiple NR
microsurgery 0.5% with epinephrine (0.25 mg/kg per hour)
Short (Abstract) [19] 2016 23 Major foot and ankle Popliteal sciatic NR Ropivacaine (0.2%) NR 5 mL per hour (10 mL PCRA (volume NR) NR NR
surgery per 2 h)
Taboada [17] 2008 44 Hallux valgus surgery Popliteal sciatic NS Levobupivacaine 30 mL of mepivacaine 1.5% 5 mL per hour (5 mL Oral opioids Single Yes
(0.125%) per hour)
Taboada [4] 2009 50 Hallux valgus surgery Popliteal sciatic NS Levobupivacaine 30 mL of mepivacaine 1.5% 5 mL per hour (5 mL PCRA (3 mL per 15 min, Single Yes
(0.125%) per hour) max 6 mL of PCRA
per hour)
Thapa [15] 2017 50 ACL reconstruction Adductor canal U/S Ropivacaine (0.5%) 15 mL of ropivacaine 0.5% 2.5 mL per hour IV PCA NR Yes
(15 mL per 6 h)
Wang [16] 2016 60 Total hip arthroplasty Fascia iliaca U/S Ropivacaine (0.2%) 40 mL of ropivacaine 0.2% 10 mL per hour Oral opioids Unclear Yes
(10 mL per hour)

U/S = ultrasound guided; NS = neurostimulation guided; IV = intravenous; IM = intramuscular; PCA = patient-controlled analgesia; NR = not reported (or insufcient information provided to make a determination); ACL = anterior cruciate
ligament; PCRA = patient-controlled regional anesthesia (on-demand boluses of local anesthetic given through the nerve block catheter); MMA = multi-modal analgesia. Conference abstracts are denoted with (Abstract) after the principal
investigator's last name in the trial column. The rest of the papers were published in peer-reviewed journals.
 M.A. Chong et al. / Journal of Clinical Anesthesia 42 (2017) 6976
Fig. 2. Risk of Bias Assessment. For each assessed domain of the Cochrane Risk of Bias Tool,
green indicates that the study had low risk of bias and red indicates a high risk of bias.
Clear cells indicate that denitive assessment of that domain was not possible due to
insufcient information reported. (For interpretation of the references to color in this
gure legend, the reader is referred to the web version of this article.)
14.4 mm to 5.4 mm; p b 0.001; I2 = 12.4%), and for the overall
worst pain score reported ( 13.3 mm; 95% CI 19.7 mm to
6.9 mm; p b 0.001; I2 = 61.2%). There were no statistically signicant
differences in the rest of the pain scores at 24 h, 36 h, 48 h, or among
studies only reporting an average postoperative pain score (Fig. 3).
With regard to subgroup analysis, much of the pre-specied analy-
ses were underpowered due to the small number of RCTs and heteroge-
neous reporting of pain scores. A full summary of subgroup analyses
that could be performed is listed in tabular form in Supplemental Digital
Content 3. For example, it was possible to perform subgroup analysis by
type of rescue analgesia (PCRA as rescue analgesia versus rescue analge-
sia from other sources, such as oral opioids). Interestingly, the benet of
PIB was more pronounced in the latter group of studies: VAS pain scores
at 6 h (20.3 mm; 95% CI 33.5 mm to 7.0 mm; p = 0.003; N = 3)
and 12 h (10.5 mm; 95% CI 14.4 mm to 6.6 mm; p b 0.001; N = 5)
73
[11,1518]. The rest of the subgroup analyses that could be conducted
were largely concordant with the overall pooled values (Supplemental
Digital Content 3).
3.4. Secondary outcomes
All secondary outcomes and their corresponding subgroup analyses
are summarized in tabular form in Supplemental Digital Content 4.
3.4.1. Patient satisfaction
No difference was detected in patient satisfaction between groups
(10 point scale WMD 0.40; 95% CI 0.12 to 0.91; p = 0.13), with only
4 studies reporting this outcome in an extractable fashion [15,16,18,
19]. Furthermore, one study utilized an ordinal scale for this outcome,
with no difference between groups [4].
3.4.2. Rescue analgesia requirement and drug consumption
The proportion of patients requiring rescue analgesia was not differ-
ent between groups (OR 0.51; 95% CI 0.221.16; p = 0.11; N = 6). With
regards to local anesthetic consumption, we were unable to perform
meta-analysis due to the heterogeneity in the reporting of this outcome
between trials [5,11]. Nonetheless, to comment on the trials individual-
ly, only two trials demonstrated lower local anesthetic consumption in
the PIB group compared to continuous infusion, although the magni-
tude of reduction was small (b 1 mL/h of study solution in one study
[4] and b 10% reduction in the other study) [14]. Similarly, there was
no difference among opioid consumption between study arms in the 4
RCTs that reported that outcome (12.6 mg of oral morphine equiva-
lents; 29.6 mg to 4.42 mg; p = 0.15) [5,15,18,19].
3.4.3. Side effects
The incidence of nausea and vomiting was no different between
groups (OR 0.88; 95% CI 0.126.2; p = 0.89; N = 2) [15,18]. Other opi-
oid-related side effects such as pruritus and respiratory depression were
only reported in an extractable fashion by a minority of studies, preclud-
ing meta-analysis [18,19]. Within all individual trials, events were infre-
quent and there was no signicant difference between study arms.
Prolonged motor block was reported by 3 studies that demonstrated a
higher chance of residual motor block with the PIB technique (OR
4.27; 95% CI 1.0816.9; p = 0.04; NNTH = 8) [4,17,19]. However,
there was no difference in permanent block-related complications; the
only reported incident among all the reviewed RCTs was in one patient
who developed a cervical abscess after interscalene block [5].
3.5. Publication bias and on-going studies
Egger's regression was performed on the worst reported VAS pain
score data and the result was not signicant (p = 0.052). However,
the constructed funnel plot was quite asymmetric (Supplemental
Digital Content 5: funnel plot for VAS pain scores). Given these conict-
ing results, the presence of publication bias cannot be excluded. With
regard to additional studies of PIB in regional anesthesia, 3 trials are cur-
rently in-progress or pending publication according to the clinicaltrials.
gov database (Table 2).
4. Discussion
4.1. Summary of ndings and interpretation
This systematic review and meta-analysis of RCTs compared 9 trials
of PIB to continuous infusion of local anesthetics during CPNB for
postoperative analgesia. The main ndings are listed in the Summary
of Findings Table (Table 3). Our analysis demonstrates that VAS pain
scores are only modestly reduced at 6 h ( 14.2 mm; 95% CI
23.5 mm to 5.0 mm) and 12 h ( 9.9 mm; 95% CI 14.4 mm to
5.4 mm)with no difference at later time points. This pain score
74 M.A. Chong et al. / Journal of Clinical Anesthesia 42 (2017) 6976

Fig. 3. Forest plot of VAS pain scores over time. Data are presented by time point with a Bonferroni correction applied for multiple comparisons.

reduction would not be considered by most to be clinically meaningful data (I2 = 82.5%), and the fact that not all studies reported pain scores
and there were no differences in VAS pain scores at other time points for each time point (Fig. 3). Finally, nearly half the studies utilized
[20,21]. Other caveats include the small number of studies and patients PCRA boluses, which may have obscured any difference between PIB
included in some comparisons, high statistical heterogeneity in the 6 h versus continuous infusion (Table 1). Taken altogether, we do not

Table 2
On-going trials listed on clinicaltrials.gov.

ClinicalTrials.gov identier Type of block Type of surgery Study statusa Estimated date of completiona

NCT02707874 Popliteal sciatic Major ankle Recruiting June 2018

NCT02589288 Adductor canal Anterior cruciate ligament reconstruction Recruiting October 2017
NCT02539628 Adductor canal Total knee arthroplasty Completed September 2016
a
As per clinicaltrials.gov record.
 M.A. Chong et al. / Journal of Clinical Anesthesia 42 (2017) 6976
Table 3
Summary of ndings table.
Outcomes Illustrative comparative risks (95% CI)
Assumed risk (standard care group) Corresponding risk reduction in PIB group
(95% CI)
VAS pain scores The mean VAS pain scores ranged VAS pain scores were 14.2 mm
at 6 h from 10 to 30 in the control group (23.5 mm to 5.0 mm) lower in the
PIB group
VAS pain scores The mean VAS pain scores ranged VAS pain scores were 9.9 mm
at 12 h from 10 to 30 in the control group (14.4 mm to 5.4 mm) lower in the
PIB group
Prolonged motor 53 per 1000 patients 140 (4 to 432) events avoided per 1000
blockade patients
a
High risk of bias due to unclear method of allocation concealment for the majority of studies and deciencies in the reporting of randomization sequence generation technique.
b
Very high statistical heterogeneity.
believe that the current evidence supports the PIB technique to have
any meaningful analgesic benet in CPNB.
The secondary outcomes also reected the lack of analgesic benet
for PIB, with no differences in side effects, opioid consumption, or rescue
analgesia requirement. The only exception was increased rates of resid-
ual motor blockade with PIB (OR 4.27; 95% CI 1.0816.9; p = 0.04;
NNTH = 8). Although no study reported permanent sequelae from
this residual motor blockade, it is possible that such prolonged motor
block may hinder postoperative mobilization and impact patient satis-
faction. Finally, despite our initial hypothesis that PIB decreased local
anesthetic consumption, this was only demonstrated in two studies,
with the remainder showing no meaningful difference [5,11].
4.2. Study limitations
The 9 RCTs that were included in this review only comprised 448
patients. Given this small overall sample size, the majority of our pre-
specied subgroup analyses were underpowered. We await with great
interest the completion and publication of several on-going studies in-
vestigating PIB in regional anesthesia (NCT02707874, NCT02589288,
and NCT02539628; Table 2). Although these trials do overlap with the
existing literature in terms of patient population and type of nerve
block being studied, the additional sample size will be informative
given the small size and number of existing studies.
Regarding the heterogeneity of the included trials, signicant clinical
differences exist between existing studies. For example, although we
had hoped to present subgroup analysis by type of block and block loca-
tion to account for some of this heterogeneity, there were insufcient
studies to do so. Such subgroup analyses would be of high interest,
given that optimal CPNB dosing is known to vary by these anatomic fac-
tors [22,23]. Additionally, PCRA and IV PCA as rescue analgesia are
known to provide different analgesic proles [24]. Taken together,
these variations among the selected trials make the interpretation of
the pooled results more challenging.
To comment on block placement, nearly half of the studies utilized
neurostimulation, which some authors suggest leads to a slightly higher
failure rate compared to ultrasound guidance [25]. Fortunately, this did
not result in a signicant amount of failed blocks among the studies
reviewed. Nonetheless, block and catheter placement method remains
an important consideration, as certain studies demonstrate that block
quality is improved when ultrasound guidance is utilized [26]. Unfortu-
nately, due to the paucity of studies, our pre-specied subgroup analysis
by block placement method was not possible.
Finally, with regard to the generalizability of the results, only 5 of 9
studies utilized multi-modal analgesia [4,5,1517]. Given that multi-
modal analgesia has been demonstrated to decrease pain intensity,
we suspect that its absence from some trials likely exaggerated the
effect of the PIB on pain scoreswhich was already very modest in our
analysis [1,27].
75
Relative effect (95% CI) Number of subjects Quality of the
(number of studies) evidence (GRADE)
14.2 mm (23.5 mm to 227 (5)
5.0 mm) Very lowa,b
9.9 mm (14.4 mm to 236 (5)
5.4 mm) Very lowa,b
OR 4.27 (1.0816.9) 117 (3)
Lowa
4.3. Conclusion
Although PIB has seen much success for labor analgesia, our system-
atic review and meta-analysis did not demonstrate clinically meaning-
ful analgesic benet in CPNB across the current evidence base [3].
Unfortunately, as we have delineated, the published RCTs of PIB are
fraught with limitations and have poor external validity. Nonetheless,
given the premium cost of PIB-capable infusion pumps, the lack of
meaningful benet in the existing literature serves as a word of caution
for widespread adoption of the PIB technique for CPNB [6].
Our systematic review provides important information to inform the
conduct of future studies, given that we have identied several repeated
limitations in the existing trials of PIB versus continuous infusion in
CPNB. We suggest that future studies involving PIB be adequately
powered to detect meaningful differences in important patient
outcomes, such as VAS pain scores or patient satisfaction. Furthermore,
future trials should reect contemporary anesthetic practice and utilize
techniques shown to improve analgesia, such as multi-modal analgesia
and PCRA boluses [1,27]. Finally, given the very limited pain score
reduction demonstrated herein, it may be most worthwhile to focus
any further trials on only the most painful of surgeries.
Supplementary data to this article can be found online at http://dx.
doi.org/10.1016/j.jclinane.2017.08.018.
Acknowledgements
Dr. Chong is grateful to the Medical Evidence, Decision Integrity,
Clinical Impact (MEDICI) Centre at Western University (London, Ontario,
Canada) for providing access to the statistical analysis software and to Dr.
Nicolas M. Berbenetz for reviewing the nal manuscript. This research
did not receive any specic grant from funding agencies in the public,
commercial, or not-for-prot sectors.
References
[1] Singelyn FJ, Seguy S, Gouverneur JM. Interscalene brachial plexus analgesia after
open shoulder surgery: continuous versus patient-controlled infusion. Anesth
Analg 1999;89(5):121620.
[2] Wong CA, Ratliff JT, Sullivan JT, Scavone BM, Toledo P, McCarthy RJ. A randomized
comparison of programmed intermittent epidural bolus with continuous epidural
infusion for labor analgesia. Anesth Analg 2006;102(3):9049.
[3] George RB, Allen TK, Habib AS. Intermittent epidural bolus compared with continu-
ous epidural infusions for labor analgesia: a systematic review and meta-analysis.
Anesth Analg 2013;116(1):13344.
[4] Taboada M, Rodriguez J, Bermudez M, Amor M, Ulloa B, Aneiros F, et al. Comparison
of continuous infusion versus automated bolus for postoperative patient-controlled
analgesia with popliteal sciatic nerve catheters. Anesthesiology 2009;110:1504.
[5] Hamdani M, Chassot O, Fournier R. Ultrasound-guided continuous interscalene
block: the inuence of local anesthetic background delivery method on postopera-
tive analgesia after shoulder surgery: a randomized trial. Reg Anesth Pain Med
2014;39(5):38793.
[6] McKenzie CP, Cobb B, Riley ET, Carvalho B. Programmed intermittent epidural bolus-
es for maintenance of labor analgesia: an impact study. Int J Obstet Anesth 2016;26:
328.
76 M.A. Chong et al. / Journal of Clinical Anesthesia 42 (2017) 6976
[7] Liberati A, Altman DG, Tetzlaff J, Mulrow C, Gotzsche PC, Ioannidis JP, et al. The
PRISMA statement for reporting systematic reviews and meta-analyses of studies
that evaluate health care interventions: explanation and elaboration. J Clin
Epidemiol 2009;62(10):e134.
[8] Higgins J, Green S. Cochrane handbook for systematic reviews of interventions. The
Cochrane Collaboration; 2011.
[9] Corbett MS, Higgins JP, Woolacott NF. Assessing baseline imbalance in randomised
trials: implications for the Cochrane risk of bias tool. Res Synth Methods 2014;
5(1):7985.
[10] Guyatt GH, Oxman AD, Vist GE, Kunz R, Falck-Ytter Y, Alonso-Coello P, et al. GRADE:
an emerging consensus on rating quality of evidence and strength of recommenda-
tions. BMJ 2008;336(7650):9246.
[11] Mezzatesta JP, Scott DA, Schweitzer SA, Selander DE. Continuous axillary brachial
plexus block for postoperative pain relief. Intermittent bolus versus continuous infu-
sion. Reg Anesth 1997;22:35762.
[12] Friedrich JO, Adhikari NK, Beyene J. Inclusion of zero total event trials in meta-anal-
yses maintains analytic consistency and incorporates all available data. BMC Med
Res Methodol 2007;7:5.
[13] Lee HM, Choi WJ, Choi KT. Comparison of intermittent versus continuous
interscalene brachial plexus block for postoperative analgesia. Anesth Pain Med
2010;5:1114.
[14] Meyer A, Breebaart M, Schepens T, Slappendel R. Continuous infusion versus auto-
mated intermittent bolus of levobupivacaine 0,125% in ultrasound guided
subparaneural sciatic nerve catheters. a prospective double blind randomised study.
Regional Anesthesia and Pain Medicine. Conference: 35th Annual European Society
of Regional Anaesthesia and Pain Therapy Congress, ESRA, 41. ; 2016. p. e4950.
[15] Thapa D, Ahuja V, Verma P, Gombar S, Gupta R, Dhiman D. Post-operative analgesia
using intermittent vs. continuous adductor canal block technique: a randomized
controlled trial. Acta Anaesthesiol Scand 2017;60:137985.
[16] Wang N, Li M, Geng J, Chen X, Guo X. A clinical study of the efcacy of automated
intermittent boluses for continuous fascia iliaca block. [Chinese]. Nat Med J China
2016;96:17504.
[17] Taboada M, Rodriguez J, Bermudez M, Valino C, Ulloa B, Aneiros F, et al. A new au-
tomated bolus technique for continuous popliteal block: a prospective, randomized
comparison with a continuous infusion technique. Anesth Analg 2008;107(4):
14337.
[18] Hillegass MG, Field LC, Stewart SR, Borckardt JJ, Dong L, Kotlowski PE, et al. The ef-
cacy of automated intermittent boluses for continuous femoral nerve block: a pro-
spective, randomized comparison to continuous infusions. J Clin Anesth 2013;25:
2818.
[19] Short A, Chin KJ, Ghosh SM, Jin R, Chan V. A comparison of programmed intermittent
bolus versus continuous infusion local anesthetic regimens for continuous popliteal
nerve block. Regional Anesthesia and Pain Medicine. Conference: 41st Annual
Regional Anesthesiology and Acute Pain Medicine Meeting of the American Society
of Regional Anesthesia and Pain Medicine, ASRA, 41. ; 2016.
[20] Farrar JT, Berlin JA, Strom BL. Clinically important changes in acute pain outcome
measures: a validation study. J Pain Symptom Manag 2003;25(5):40611.
[21] Farrar JT, Portenoy RK, Berlin JA, Kinman JL, Strom BL. Dening the clinically important
difference in pain outcome measures. Pain 2000;88(3):28794.
[22] Madison SJ, Monahan AM, Agarwal RR, Furnish TJ, Mascha EJ, Xu Z, et al. A random-
ized, triple-masked, active-controlled investigation of the relative effects of dose,
concentration, and infusion rate for continuous popliteal-sciatic nerve blocks in
volunteers. Br J Anaesth 2015;114(1):1219.
[23] Ilfeld BM, Le LT, Ramjohn J, Loland VJ, Wadhwa AN, Gerancher JC, et al. The effects of
local anesthetic concentration and dose on continuous infraclavicular nerve blocks:
a multicenter, randomized, observer-masked, controlled study. Anesth Analg 2009;
108(1):34550.
[24] Borgeat A, Perschak H, Bird P, Hodler J, Gerber C. Patient-controlled interscalene an-
algesia with ropivacaine 0.2% versus patient-controlled intravenous analgesia after
major shoulder surgery: effects on diaphragmatic and respiratory function. Anesthe-
siology 2000;92(1):1028.
[25] Abrahams MS, Aziz MF, Fu RF, Horn JL. Ultrasound guidance compared with electri-
cal neurostimulation for peripheral nerve block: a systematic review and meta-
analysis of randomized controlled trials. Br J Anaesth 2009;102(3):40817.
[26] Williams SR, Chouinard P, Arcand G, Harris P, Ruel M, Boudreault D, et al. Ultrasound
guidance speeds execution and improves the quality of supraclavicular block.
Anesth Analg 2003;97(5):151823.
[27] Elia N, Lysakowski C, Tramer MR. Does multimodal analgesia with acetaminophen,
nonsteroidal antiinammatory drugs, or selective cyclooxygenase-2 inhibitors and
patient-controlled analgesia morphine offer advantages over morphine alone?
Meta-analyses of randomized trials. Anesthesiology 2005;103(6):1296304.