Received: 17 March 2023 Revised: 19 July 2023 Accepted: 8 August 2023

DOI: 10.1111/aas.14318

RESEARCH ARTICLE

Rebound pain and postoperative pain profile following brachial

plexus block compared to general anaesthesia—An
observational study

Ann-Kristin Schubert 1 | Thomas Wiesmann 1,2 | Christian Volberg 1 |

Jenny Riecke 3 | Alexander Schneider 1 | Hinnerk Wulf 1 | Hanns-Christian Dinges 1

1
Department of Anaesthesiology and Intensive
Care Medicine, University Hospital Marburg, Abstract
Philipps University of Marburg, Marburg,
Background: Regional anaesthesia has the benefit of reducing the need for systemic
Germany
2
Department of Anaesthesiology and Intensive
analgesia and therefore, potentially reducing undesired side effects. With the end of
Care Medicine, Diakoneo Diak Klinikum the sensory nerve block however, many patients report severe pain that requires
Schwäbisch-Hall, Schwäbisch-Hall, Germany
3
therapy with opioids and often compromise the initial opioid sparing effect. This
Department of Clinical Psychology and
Psychotherapy, Philipps University Marburg, study aimed to characterise the postoperative pain profile and the phenomenon of
Marburg, Germany rebound pain after axillary brachial plexus anaesthesia (RA) compared to general
Correspondence anaesthesia (GA).
Ann-Kristin Schubert, Department of Design: Single-centre observational, stratified cohort study.
Anaesthesiology and Intensive Care Medicine,
University Hospital Marburg, Philipps Setting: The study was conducted at University Hospital Marburg from May 2020
University Marburg, 35033 Marburg,
until September 2022.
Germany.
Email: [email protected] Participants: One hundred thirty-two patients receiving elective hand and forearm
surgery were enrolled in this study.
Interventions: Group RA received ultrasound-guided brachial plexus anaesthesia via
the axillary approach with 30 mL of prilocaine 1% and 10 mL ropivacaine 0.2%.
Group GA received balanced or total intravenous general anaesthesia.
Main Outcome Measures: Primary endpoint were integrated pain scores (IPS) within
24 h postoperatively. Secondary endpoints were pain scores (NRS 0–10), morphine
equivalents, patient satisfaction, quality of recovery and opioid-related side effects.
Results: One hundred thirty-two patients were analysed of which 66 patients received
brachial plexus block and 66 patients received general anaesthesia. Following RA sig-
nificantly lower IPS were seen directly after surgery (p < .001) and during the post-
anaesthesia care unit interval (p < .001) but equalised after 3 h at the ward. No over-
shoot in pain scores or increased opioid consumption could be detected. Patient satis-
faction and postoperative recovery were comparable between both groups.
Conclusion: The IPS and NRS was initially lower in the RA group, increased with fad-
ing of the block until equal to the GA group and equal thereafter. Although various
definitions of rebound pain were met during this phase, the opioid sparing effect of
regional anaesthesia was not counteracted by it. The incidence of episodes with

This is an open access article under the terms of the Creative Commons Attribution-NonCommercial-NoDerivs License, which permits use and distribution in any
medium, provided the original work is properly cited, the use is non-commercial and no modifications or adaptations are made.
© 2023 The Authors. Acta Anaesthesiologica Scandinavica published by John Wiley & Sons Ltd on behalf of Acta Anaesthesiologica Scandinavica Foundation.

Acta Anaesthesiol Scand. 2023;1–9. wileyonlinelibrary.com/journal/aas 1

 13996576, 0, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/aas.14318 by Tunisia Hinari NPL, Wiley Online Library on [05/09/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
2 SCHUBERT ET AL.

uncontrolled, severe pain did not differ between groups. We found no clinical impli-
cations of rebound pain in this setting, since the RA group did not show higher pain
scores than the GA group at any time point.
Trial Registration: German Clinical Trials Register (DRKS00021764).

KEYWORDS
brachial plexus block, peripheral nerve block, postoperative pain, rebound pain, regional
anaesthesia

Editorial Comment
In this stratified observational study comparing brachial plexus block to general anaesthesia, a
significant reduction in pain and opioid consumption was observed in the post-anaesthesia care
unit (PACU) and the first 3 h at the ward. Importantly, no signs of rebound pain were observed
in the brachial plexus group at any timepoint. These findings support the use of brachial plexus
block for lower arm surgery and potentially early PACU discharge or bypass as the next step.

1 | I N T RO DU CT I O N and its impact on the opioid sparing effect of peripheral regional

anaesthesia.
Rebound pain is a clinical phenomenon that is associated with fading
of sensory block after regional anaesthesia.1 Patients affected by this
describe it as a very noticeable and sudden increase in pain that 2 | METHODS
requires immediate analgesic treatment. This raises the question how
much the opioid sparing effect of regional anaesthesia is counteracted 2.1 | Reporting
by the phenomenon of rebound pain and might result in poorer qual-
ity of recovery after surgery. According to the recommendations of the EQUATOR network
Different authors have tried to characterise rebound pain further (Enhancing the Quality and Transparency of Health Research), the guide-
and it has become a topic of interest in recent literature on analgesic line for observational studies STROBE (The Strengthening the Reporting
quality. A clear definition of rebound pain is still missing. Some authors of Observational Studies in Epidemiology Statement)10 was followed.
like Barry, Hamilton and others propose a definition via an increase in The completed checklist can be found in the supplemental content.
pain from NRS <3 to NRS >7,2,3 while Dada et al. define the phenome-
non by the time and interval, in which the pain occurs.4 Yet, most
authors that reported rebound pain in the past have not given a defini- 2.2 | Study approval
tion other than severe pain or a quantifiable increase in pain associated
with fading of the block.5–7 To assess postoperative pain and analgesia Ethical approval for this study (Ethical Committee N 24/20) was pro-
quality, different effect measures have been proposed. Raw pain scores vided by the Ethical Committee at University Hospital Marburg, Mar-
(NRS 0–10) and morphine equivalents (MEQ) are the most common burg, Germany (Chairperson C. Seifart) on 13 March 2020.
measures in pain studies. These endpoints are easily comprehensible;
however, when interpreted separately, they can be misleading because
differences in MEQ consumption can nullify differences in pain scores 2.3 | Study registration
and vice versa. To overcome this and enhance the precision at which
differences in analgesic quality can be detected, composite outcomes This study followed the Declaration of Helsinki and was prospectively
like integrated pain scores (IPS) have been proposed.8 registered on 15 May 2020 in the German Clinical Trials Register
Composite endpoints are particularly useful when investigating (DRKS00021764). Written informed consent was obtained from all
postoperative pain profiles and the phenomenon of rebound pain, subjects participating.
9
which has been conducted by Sort et al. characterising rebound pain
following foot and ankle surgery under spinal anaesthesia versus pop-
liteal sciatic nerve block. 2.4 | Study design
The aim of our observational study is to characterise the pain pro-
files and analgesic quality after brachial plexus anaesthesia compared This single-centre, observational, stratified cohort study was con-
to general anaesthesia for hand and forearm surgery. Furthermore, we ducted at University Hospital Marburg from May 2020 until
wanted to investigate the incidence and relevance of rebound pain September 2022.
 13996576, 0, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/aas.14318 by Tunisia Hinari NPL, Wiley Online Library on [05/09/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
SCHUBERT ET AL. 3

2.5 | Sample size calculation blocks were ultrasound guided (S-Series; Sonosite, Bothell,
Washington, USA) and performed by anaesthetists with basic experi-
The sample size calculations for this observational study were based ence in regional anaesthesia. Nerve stimulation (HNS12; B. Braun, set
on preliminary studies that examined pain levels expressed as IPS and at 2 Hz, 0.1-ms pulse duration, target current 0.5 to 1.0 mA) was used
the incidence of rebound pain after peripheral nerve block.1,9,11 The for additional verification of final tip position.
goal was to detect a 30% difference in IPS, pain score and/or con- Peripheral nerve block was conducted with 30 mL 1% prilocaine
sumption of morphine equivalents, which resembles a NRS difference to which 10 mL 0.2% ropivacaine were added for longer block dura-
of 1.5–2 depending on the mean. Assuming a power of 80% in a two- tion, according to the local standard.
tailed test at a level of significance of 0.05, a total number of
90 patients were calculated. To account for potential drop-outs, a
total of 110 patients were required. 2.8 | General anaesthesia
Sample size calculation was performed using G*Power
software.12 General anaesthesia was induced using intravenous propofol 2–
3 mg
kg 1
and intravenous fentanyl 2–3 μg
kg 1
according to local
standards. Insertion of either a laryngeal mask or an endotracheal tube
2.6 | Eligibility after intravenous rocuronium 0.5–0.6 mg
kg 1
was at the discretion
of the anaesthetist in charge. Balanced or total intravenous anaesthe-
Adult patients ASA status I-IV undergoing elective surgery on the sia was maintained with desflurane or propofol (target bispectral index
hand, wrist and forearm under either brachial plexus block or general values between 35 and 55) and intermittent fentanyl as required.
anaesthesia were eligible for this observational study.
Exclusion criteria were age under 18 years, inability, or unwilling-
ness to give consent, pregnancy, presence of infection at injection 2.9 | Postoperative pain therapy
site, presence of pre-existing central and peripheral neurological dis-
ease, pre-existing coagulopathy, and known or suspected allergy to Standard basic postoperative analgesia protocol included ibuprofen
local anaesthetics. and metamizole or paracetamol as basic analgesia as follows: ibupro-
fen 600 mg QDS (four times daily), metamizole 1000 mg QDS or para-
cetamol 1000 mg QDS. For rescue analgesia patients received
2.6.1 | Recruitment oxycodone + naloxone combination and/or intravenous morphine or
piritramide on demand, following local standard operating procedures
In absence of severe cardiopulmonary comorbidity, peripheral regional for postoperative pain.
anaesthesia and general anaesthesia are considered equally viable for
surgery of the hand and forearm at our institution. Written informed
consent to the individually preferred anaesthesia was obtained from 2.10 | Data collection
patients before informing them about the study. After written
informed consent to also participate in the observation was obtained, Baseline data and patient characteristics (age, sex, weight, ASA physi-
the patients were enrolled. Towards the end of reaching the sample cal status) were collected pre-operatively.
size with this intention-to-treat recruitment manner, the demographi- To examine the full timespan in which rebound pain is likely to
cal differences between groups were too inhomogeneous to answer occur after fading of peripheral regional anaesthesia, 24 h postopera-
the research question. An amendment was made to the hospital ethics tively were chosen for assessment.1,9
committee to recruit 30 additional patients and to switch to a strati- Patients were asked on current pain score 24 h postoperatively
fied recruitment to even out the demographic differences. During on a 0–10 numeric rating scale (NRS) and cumulative opioid and non-
stratified recruitment, patients still chose their individually preferred opioid consumption were reported.
type of anaesthesia, but were only asked to participate in the study, if For longitudinal pain assessment patients registered pain scores
they happened to be of the underrepresented gender and had chosen on a NRS from 0 to 10 for the first 24 h postoperatively at 3-h
the underrepresented type of anaesthesia. intervals.
At 24 h postoperatively patients answered questionnaires on sat-
isfaction, quality of recovery (German QoR-15 and Global Surgical
2.7 | Peripheral nerve block Recovery Index, GSR) on a 0–100 NRS. Opioid-related side effects
were assessed using the German version of the opioid-related symp-
We used an axillary approach to the brachial plexus for regional tom distress scale (OR-SDS).
anaesthesia group. For missing data in longitudinal repeated measures, the last
In the induction area, standard monitoring was applied, and observation was carried forward. For missing data in single measures,
patients were placed in the supine position. All peripheral nerve the respective case was excluded from calculations.
 13996576, 0, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/aas.14318 by Tunisia Hinari NPL, Wiley Online Library on [05/09/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
4 SCHUBERT ET AL.

2.11 | Follow-up 2.13 | Secondary endpoints

All patients received a telephone call on postoperative Day 7 with a Secondary endpoints were:
questionnaire on actual pain scores, functional status and potential
adverse events. For long-time follow-up assessment patients were • Mean pain scores on a numeric rating scale (0–10) registered
contacted by phone 3 months after surgery to obtain actual pain within the first 24 h in 3-h intervals
score, satisfaction and quality of recovery (QoR-15, GSR). Patients not • Cumulative morphine equivalent consumption 24 h
available for telephone interview were called each day for up to postoperatively
5 days and were then excluded from analysis as lost to follow-up. • Patient satisfaction, adverse events and opioid-related site-effects
(OR-SDS), quality of postoperative recovery (GSR)

2.12 | Primary endpoint Opioid-related side effects were reported as incidence of clini-
cally meaningful events (CME) assessed with the German OR-SDS as
Primary endpoint were differences in the IPS over the course of 24 h proposed by Chan et al.13
postoperatively, measured in 3-h intervals. IPS were calculated from Patient satisfaction was assessed on a 0–10 NRS (very dissatis-
reported pain scores and cumulative morphine equivalents (MEQ) tak- fied to very satisfied).
ing into account any higher MEQ requirements to achieve comparable
analgesia (equi-analgesia) as proposed by Silverman et al.8
IPS were calculated as the deviation from mean rank in pain score 2.14 | Statistical analysis
added to deviation from mean rank in MEQ using Microsoft Excel,
resulting in an individual score for each participant between 200% Statistical analysis was performed using SPSS version 27.0 (SPSS Inc.,
and +200% (the higher the pain score and the higher the opioid con- Chicago, Illinois, USA) and Microsoft Excel for Mac Version 16.43
sumption, the higher the IPS-Score). (Microsoft Corporation, Redmond, Washington, USA).

Enrollment Patients with forearm/hand

surgery from May 2020 to
January 2021 (n= 195)
and from May 2022 to
September 2022 (n=46)

Excluded (n= 101)

¨ Declined to participate
¨ Did not meet inclusion criteria
¨ Other reasons

Allocation

Patients enrolled in the study

(n=140)

Allocated to regional anesthesia (n= 74) Allocated to general anesthesia (n= 66)
¨ Conversion to general anesthesia due to
insufficient block (n= 6)
¨ Local infiltration instead of regional
anesthesia (n=1)
¨ Regional and general anesthesia (n=1)

Analysis

Analysed for primary endpoint (n= 66) Analysed (n= 66)

¨ Secondary endpoint (n= 65)

FIGURE 1 CONSORT-flow chart of the inclusion process.

 13996576, 0, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/aas.14318 by Tunisia Hinari NPL, Wiley Online Library on [05/09/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
SCHUBERT ET AL. 5

Patient characteristics are presented as mean (SD) or number (%) 3 | RE SU LT S

unless otherwise stated as a standard normal distribution was
assumed for patient characteristics. 3.1 | Enrolment
For pain scores, MEQ and integrated scores standard normal dis-
tribution was tested via Normal-Q-Q-Plots. p < .05 was considered Over the duration of recruitment, a total number of 241 patients were
statistically significant. scheduled for elective surgery on the hand, wrist, or forearm.
One hundred ten patients were recruited via convenience sam-
pling method. The patient characteristics of the cohorts were inhomo-
TABLE 1 Patient characteristics. geneous regarding cohort size and sex distribution. As the general

General
anaesthesia group was underpowered based on the sample size calcu-
Axillary brachial anaesthesia, lation, an amendment was applied to proceed with stratified recruit-
plexus block, n = 66 n = 66 ment of 30 additional patients.
Age in years, mean ± SD 40 ± 16 40 ± 16 In total, 140 patients met the inclusion criteria and gave written
BMI (kg m 2
), mean 27 ± 7 26 ± 4 informed consent to be included in the study. Eight patients were
± SD excluded from data analysis: one patient due to requiring surgical infil-
Sex (male/female) 46/20 51/15 tration instead of regional anaesthesia and one patient because of
ASA physical status 41/18/7 31/31/4 receiving a supplemental supraclavicular block in combination with
(n = I/II/III) general anaesthesia and six patients due to insufficient nerve block
Duration of surgery in 48 ± 29 56 ± 45 requiring conversion to general anaesthesia. One hundred thirty-two
minutes, mean ± SD patients were included in final data analysis for primary endpoint.
Type of surgery The regional anaesthesia group had 66 patients; the general anaes-
Osteosynthesis 35 39 thesia group had 66 patients. The inclusion process is shown in Figure 1.
Soft tissue surgery 24 18
Arthroscopy 0 2
Removal of 7 7 3.2 | Patient characteristics
osteosynthesis
Use of intraoperative 56 60 Patient characteristics, details of anaesthesia techniques and surgical
Pneumatic tourniquet procedures of included patients are presented in Table 1.
Note: Patient characteristics by study group. Data are presented as mean Both groups were comparable regarding age, body mass index
± SD (standard deviation) and n. (BMI), ASA physical status, duration and type of surgery.

F I G U R E 2 Integrated pain scores

(mean ± SD) over 24 h after upper limb
surgery per group. General compared with
regional anaesthesia.
 13996576, 0, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/aas.14318 by Tunisia Hinari NPL, Wiley Online Library on [05/09/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
6 SCHUBERT ET AL.

F I G U R E 3 Mean pain scores (mean

± SD) over 24 h after upper limb surgery
on a 0–10 numeric rating scale (NRS) per
group. General compared with regional
anaesthesia.

Normal Distribution of the data was verified via Q-Q Plot for pri- interval was seen in 45 patients with regional anaesthesia, and
mary endpoint (Supplemental Digital Content 1). 30 patients following general anaesthesia.6 Transitions from well-
controlled pain (≤3) to uncontrolled and strong pain (≥7) were seen in
13 (RA) and 13 (GA) patients.2
3.3 | Primary endpoint

3.3.1 | Integrated pain scores 3.4.3 | Postoperative cumulative morphine

equivalents
The IPS following Silverman et al. are shown in Figure 2. Statistically
significant differences between the groups were seen at the PACU Postoperative cumulative morphine equivalents per group are shown
( p < .001) and at transfer to the ward (Figure 2, ward 0 h, p < .001). in Figure 4. Cumulative morphine consumption at PACU was signifi-
During the first 6 h at the ward, the IPS of the groups had equalised. cantly lower in RA group (0.15 mg, SD 0.97) compared with GA group
The mean IPS over 24 h was 8.87 (SD 48.06) in the RA group and (4.62 mg, SD 5.63) ( p < .001). Cumulative morphine consumption at
8.87 (SD 67.18) in the GA group (not significant, p = .08). ward was comparable between the RA group (2.54 mg, SD 8.08) and
the GA group (2.20 mg, SD 5.83).

3.4 | Secondary endpoints

3.4.4 | Quality of recovery and patient
3.4.1 | Pain scores at group level satisfaction

Mean pain scores over 24 h are shown in Figure 3. Significant differ- Global surgery recovery rated on a 0–10 NRS was 6.1 (SD 2.1) in the
ences for pain scores were seen directly postoperative during the RA group and 5.7 (2.4) in the GA group showing no statistical
PACU interval ( p < .001) and for timepoint 0 h at the ward (p < .001) significance.
in favour of the RA group. Quality of recovery assessed on the QoR-15 was 113.3 (SD 15.1)
in RA group compared with 109.8 (19.0) in GA group without statisti-
cal significance.
3.4.2 | Individual pain scores Patient satisfaction on a 0–10 NRS was comparable in either
group (mean score of 8.6 [SD 2.1] in the RA group and 8.3 [SD 2.8] in
According to the two most common definitions of rebound pain,2,6 a the GA group).
quantifiable increase in pain (>2) during the postoperative time Results for secondary endpoints are reported in Table 2.
 13996576, 0, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/aas.14318 by Tunisia Hinari NPL, Wiley Online Library on [05/09/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
SCHUBERT ET AL. 7

F I G U R E 4 Postoperative morphine
equivalents (mean ± SD) for separate
timepoints (PACU, ward) per group.
General compared with regional
anaesthesia.

TABLE 2 Secondary endpoints.

Regional General
anaesthesia anaesthesia Statistical
Secondary endpoints (n = 66) (n = 66) p-Value test
Cumulative postoperative morphine i.v. equivalents 24 h; mg 3±9 7 ± 10 .01* t-test
Pain score AUC 0–24 h 7.3 ± 4.5 8.3 ± 5.6 .25 t-test
Quality of recovery, (QoR-15) 113 ± 15 110 ± 19 .24 t-test
Opioid adverse effects, CME ≥ 1 16 (24) 14 (21) .34 χ2
Patient satisfaction (NRS 0–10) 8.6 ± 2.1 8.3 ± 2.8 .6 t-test
Would recommend type of anaesthesia (NRS 0–10) 8.95 ± 2.0 8.82 ± 2.5 .79 t-test
Global surgery recovery (NRS 0–10) 6.1 ± 2.1 5.7 ± 2.4 .28 t-test

Note: Summary of secondary endpoints. Data presented as mean ± SD or number (n) with percentage (%).
*p < .05.

3.4.5 | Opioid-related side effects (OR-SDS) serious adverse events occurred during the follow-up period (see Sup-
plemental Digital Content 3; Tables 1 and 2).
Opioid-related side effects were reported as critical meaningful event
(CME) on the German OR-SDS. Side effects with at least one CME
were comparable and observed by 16 patients (24.24%) in the RA 4 | DI SCU SSION
Group compared to 14 patients (21.21%) in the GA group (Table 2).
Opioid-related adverse event rates are presented in Supplemental In the present study, we examined the pain profile in a multimodal
Digital Content 2. analgesia environment following hand and forearm surgery under gen-
eral versus brachial plexus anaesthesia. Our goal was to quantify the
pain associated with fading of the plexus block, which has previously
3.5 | Follow-up been described as rebound pain.
From a methodological point of view, the chosen endpoints (IPS,
The 7-day and 3-months follow-up of the patients showed no statisti- pain scores and MEQs) are adequate tools to detect differences in the
cally significant differences in pain chronification or pain profiles. No quality of analgesia between study populations in a comprehensible
 13996576, 0, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/aas.14318 by Tunisia Hinari NPL, Wiley Online Library on [05/09/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
8 SCHUBERT ET AL.

fashion, while reporting the raw pain scores and the total morphine We initially recruited patients by convenience sampling, however,
equivalents maintains the transparency and traceability for the as the resulting study groups would be too inhomogeneous to answer
clinician. the research question. The GA group included primarily women and
The patient population of young adults with predominantly bony was underpowered in size and the RA group was larger than required
surgery in the ambulatory setting receiving a single injection brachial and included primarily men. We decided to proceed with an observa-
plexus anaesthesia is at the highest risk of developing rebound pain tional but stratified recruitment regarding group size and sex distribu-
2,14,15
according to the existing literature. tion. The resulting groups were therefore homogeneous regarding
Our study revealed a statistically significant difference in IPS and these aspects. Since the enrolment to the study was still only offered
pain scores during the PACU time interval and during the first 3 h at to patients that had already decided on their type of anaesthesia, the
the ward in favour of the regional anaesthesia group. During the first study design remained entirely observational in this process and nei-
6 h, the regional anaesthesia group showed an increase in pain scores ther of the groups was biased by this. A demographic difference that
of >2 on the NRS, after which it levelled with the general anaesthesia we did not stratify remains in the distribution of surgical procedures
group (Figure 3). (Table 1). More patients in the GA group received an osteosynthesis,
Regarding the patient-centred endpoints, patients rated both which is expected to be rather painful, while soft tissue surgery, which
anaesthesia techniques equally and highly satisfactory. is considered less painful was slightly underrepresented in this group.
The initial difference in pain scores could in fact be described as This inhomogeneity is a limitation and with a randomised study
rebound pain following the definition of Williams et al.,6 since a differ- design, the GA group may have had slightly less pain than we
ence of >2 can be assumed to be quantifiable according to the com- observed here.
16,17
mon definitions of Cepeda and Farrar. If we assume that demographic differences affect the occurrence
However, apart from this definition being matched, we could not of the phenomenon rebound pain, it is important to look into the
see any of the clinical implications that have brought the phenomenon intention-to-treat distribution of risk factors or groups at risk for
of rebound pain to such a high interest in the scientific debate. The higher sensation of pain. These distributions may differ across other
pain score did not overshoot the GA group at any time point and institutions and populations and may explain, why rebound pain or
the initial morphine sparing effect of the RA was not counteracted by other phenomenons are noticeable in a clinical setting or not. If popu-
this at all, which can clearly be seen from the IPS curve (Figure 2) and lations at risk for an unwanted event tend to already choose the
is the reason why the IPS is the ideal tool to detect rebound pain. favourable intervention by themselves, the effect of the ‘worse’ inter-
If we cannot detect rebound pain at group level in the pain scores vention may be masked in the daily routine. If an effect only occurs in
and morphine equivalents, the question remains, whether it is a rele- the setting of controlled trials, it may drive a scientific debate on
vant problem for individual patients. Regarding the definition of Barry something that cannot have more clinical implications than to aim for
et al.,2 describing the transition of well-controlled to uncontrolled, the demographic distribution that is already natural.
severe pain, rebound pain occurred exactly 13 times in each of our Multiple testing was the case for the families of hypotheses
groups. The term rebound pain which implies an overshoot in pain regarding IPS and raw pain scores. However, this was not considered
and related to fading of regional anaesthesia, is therefore not suitable problematic because the described differences were detected with
to describe our observations on brachial plexus anaesthesia. high statistical significance, which would hold against a Bonferroni
The scientific debate on rebound pain originated from the differ- correction of for 10 tests and a new level of statistical significance
ences between general and regional anaesthesia. Unfortunately, only of p < .005.
few studies have compared the pain profiles of these two further.
Galos et al.14 for example could not detect clinically relevant differ-
ences in the pain profiles following brachial plexus RA. 5 | CONC LU SION
Other recent studies that have described postoperative pain pro-
files but only compared different approaches of regional anaesthesia, for Statistically significant differences in IPS were observed within the
example, the addition of adjuvants as dexamethasone or ketamine, con- early postoperative period. The IPS following brachial plexus block
cluded that the incidence of rebound pain is 50%–70% in peripheral levelled the general anaesthesia group after 6 h at ward, without
nerve blocks of brachial, lumbar and sacral plexus and that it has strong counteracting the initial opioid sparing effect. According to various
implications for clinical practice.2,9,15,18–20 Yet, they did not compare it definitions, rebound pain could be detected in this trial, however, the
to general anaesthesia anymore. But since episodes of uncontrolled pain clinical implications of it were minor. Uncontrolled, severe pain
occur during recovery from general anaesthesia as well, a study design occurred equally often in either group.
that does not include this control group may not be suitable to estimate
the true incidence of rebound pain. And therefore, it is arguable that AUTHOR CONTRIBU TIONS
these studies actually describe clinically relevant rebound pain. Concept and design: Hanns-Christian Dinges, Ann-Kristin Schubert,
Limitations of the present study are the observational design, Jenny Rieke, Christian Volberg, Thomas Wiesmann and Hinnerk Wulf.
which of course provides a lower quality of evidence than a random- Study planning: Hanns-Christian Dinges, Ann-Kristin Schubert, Jenny
ised trial would do. Rieke, Christian Volberg, Thomas Wiesmann and Hinnerk Wulf.
 13996576, 0, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/aas.14318 by Tunisia Hinari NPL, Wiley Online Library on [05/09/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
SCHUBERT ET AL. 9

Conduction and data acquisition: Alexander Schneider. Statistical analy- 8. Silverman DG, O'Connor TZ, Brull SJ. Integrated assessment of pain
sis: Hanns-Christian Dinges and Ann-Kristin Schubert. Data interpreta- scores and rescue morphine use during studies of analgesic efficacy.
Anesth Analg. 1993;77(1):168-170.
tion: Hanns-Christian Dinges, Ann-Kristin Schubert, Thomas
9. Sort R, Brorson S, Gögenur I, et al. Peripheral nerve block anaesthesia
Wiesmann and Hinnerk Wulf. Drafting of the manuscript: Ann-Kristin and postoperative pain in acute ankle fracture surgery: the AnAnkle
Schubert and Hanns-Christian Dinges. Revision and approval of the randomised trial. Br J Anaesth. 2021;126(4):881-888.
manuscript: Hanns-Christian Dinges, Ann-Kristin Schubert, Thomas 10. von Elm E, Altman DG, Egger M, Pocock SJ, Gøtzsche PC,
Vandenbroucke JP. The Strengthening the Reporting of Observa-
Wiesmann and Hinnerk Wulf.
tional Studies in Epidemiology (STROBE) statement: guidelines for
reporting observational studies. Lancet. 2007;370(9596):1453-1457.
ACKNOWLEDGEMEN TS 11. Sort R, Brorson S, Gogenur I, Moller AM. AnAnkle Trial study proto-
The authors thank all clinical and research staff members of the col: a randomised trial comparing pain profiles after peripheral nerve
block or spinal anaesthesia for ankle fracture surgery. BMJ Open.
Department of Anaesthesiology and Intensive Care Medicine at
2017;7(5):e016001.
Philipps-University Marburg for their assistance with patient recruit- 12. Faul F, Erdfelder E, Lang AG, Buchner A. G*Power 3: a flexible statis-
ment and premedication. tical power analysis program for the social, behavioral, and biomedical
sciences. Behav Res Methods. 2007;39(2):175-191.
13. Chan KS, Chen WH, Gan TJ, et al. Development and validation of a
FUND ING INFORMATION
composite score based on clinically meaningful events for the opioid-
This study was not funded by any external source. Only institutional
related symptom distress scale. Qual Life Res. 2009;18(10):1331-
funds were used in this project. 1340.
14. Galos DK, Taormina DP, Crespo A, et al. Does brachial plexus block-
CONF LICT OF IN TE RE ST ST AT E MENT ade result in improved pain scores after distal radius fracture fixation?
A randomized trial. Clin Orthop Relat Res. 2016;474(5):1247-1254.
Thomas Wiesmann has received honoraria for lectures from Pajunk,
15. Holmberg A, Hassellund SS, Draegni T, et al. Analgesic effect of
Vygon, Teva ratiopharm. Hinnerk Wulf has received honoraria for lec- intravenous dexamethasone after volar plate surgery for distal radius
tures from Sintetica and Grünenthal. All other authors declare that fracture with brachial plexus block anaesthesia: a prospective,
they have no conflict of interest. double-blind randomised clinical trial. Anaesthesia. 2020;75(11):1448-
1460.
16. Cepeda MS, Africano JM, Polo R, Alcala R, Carr DB. What decline in
DATA AVAI LAB ILITY S TATEMENT pain intensity is meaningful to patients with acute pain? Pain. 2003;
The data was collected during the study period and is presented in 105(1–2):151-157.
the manuscript. 17. Farrar JT, Young JP Jr, LaMoreaux L, Werth JL, Poole MR. Clinical
importance of changes in chronic pain intensity measured on an
11-point numerical pain rating scale. Pain. 2001;94(2):149-158.
ORCID 18. Fang J, Shi Y, Du F, et al. The effect of perineural dexamethasone on
Ann-Kristin Schubert https://orcid.org/0000-0001-6712-3812 rebound pain after ropivacaine single-injection nerve block: a ran-
domized controlled trial. BMC Anesthesiol. 2021;21(1):47.
19. Morita S, Oizumi N, Suenaga N, Yoshioka C, Yamane S, Tanaka Y.
RE FE R ENC E S
Dexamethasone added to levobupivacaine prolongs the duration of
1. Sort R, Brorson S, Gogenur I, Nielsen JK, Moller AM. Rebound pain interscalene brachial plexus block and decreases rebound pain after
following peripheral nerve block anaesthesia in acute ankle fracture arthroscopic rotator cuff repair. J Shoulder Elbow Surg. 2020;29(9):
surgery: an exploratory pilot study. Acta Anaesthesiol Scand. 2019; 1751-1757.
63(3):396-402. 20. Woo JH, Lee HJ, Oh HW, Lee JW, Baik HJ, Kim YJ. Perineural dexa-
2. Barry GS, Bailey JG, Sardinha J, Brousseau P, Uppal V. Factors associ- methasone reduces rebound pain after ropivacaine single injection
ated with rebound pain after peripheral nerve block for ambulatory interscalene block for arthroscopic shoulder surgery: a randomized
surgery. Br J Anaesth. 2021;126(4):862-871. controlled trial. Reg Anesth Pain Med. 2021;46(11):965-970.
3. Hamilton DL. Rebound pain: distinct pain phenomenon or nonentity?
Br J Anaesth. 2021;126(4):761-763.
4. Dada O, Gonzalez Zacarias A, Ongaigui C, et al. Does rebound pain
SUPPORTING INF ORMATION
after peripheral nerve block for orthopedic surgery impact postopera-
tive analgesia and opioid consumption? A narrative review. Int J Envi- Additional supporting information can be found online in the Support-
ron Res Public Health. 2019;16(18):3257. ing Information section at the end of this article.
5. Lavand'homme P. Rebound pain after regional anesthesia in the
ambulatory patient. Curr Opin Anaesthesiol. 2018;31(6):679-684.
6. Williams BA, Bottegal MT, Kentor ML, Irrgang JJ, Williams JP. How to cite this article: Schubert A-K, Wiesmann T,
Rebound pain scores as a function of femoral nerve block duration Volberg C, et al. Rebound pain and postoperative pain profile
after anterior cruciate ligament reconstruction: retrospective analysis
following brachial plexus block compared to general
of a prospective, randomized clinical trial. Reg Anesth Pain Med. 2007;
32(3):186-192. anaesthesia—An observational study. Acta Anaesthesiol Scand.
7. Williams BA. Forecast for perineural analgesia procedures for ambula- 2023;1‐9. doi:10.1111/aas.14318
tory surgery of the knee, foot, and ankle: applying patient-centered
paradigm shifts. Int Anesthesiol Clin. 2012;50(1):126-142.