Pediatric Anesthesia ISSN 1155-5645

RESEARCH REPORT

Analgesic management after thoracotomy for decortication

in children: a retrospective audit of 83 children managed
with a paravertebral infusion-based regime
Tim Murphy, Alan McCheyne & Jacob Karlsson
Department of Paediatric Cardiothoracic Anaesthesia, Freeman Hospital, Newcastle Upon Tyne, UK

What is already known

• A thoracotomy is a painful incision necessitating multimodal postoperative analgesic management including,

preferably, a regional technique.
What this article adds

• A paravertebral infusion of local anesthetic delivered via a surgically placed catheter supplemented with sys-
temic analgesics results in satisfactory pain management for children undergoing decortication via thoracotomy.

Keywords
pain; postoperative; anesthesia; local;
empyema; pleural; pneumonia;
thoracotomy; analgesia
Correspondence
Tim Murphy, Department of Anaesthetics,
Bristol Children’s Hospital, Upper Maudlin
Street, Bristol BS2 8BJ, UK
Email: tim.murphy@uhbristol.nhs.uk
Section Editor: Per-Arne Lonnqvist
Accepted 10 April 2016
doi:10.1111/pan.12921
Summary
Background: It is important that postoperative analgesic management after
thoracotomy is very effective in order to optimize postoperative recovery. A
regional technique such as an epidural or a paravertebral catheter with an
infusion of local anesthetic may be supplemented with systemically adminis-
tered analgesic drugs in order to achieve satisfactory analgesia.
Objective: The objective of this observational study was to evaluate
whether a paravertebral infusion of local anesthetic delivered via a surgi-
cally placed catheter together with systemic analgesics is associated with
low pain scores and satisfactory analgesia after thoracotomy for decortica-
tion in children.
Methods: We performed a retrospective analysis of the notes and charts of 83
children admitted with empyema thoracis and managed with thoracotomy
and decortication. We collected data on the doses of analgesic drugs
(morphine, paracetamol, and ibuprofen) and details of paravertebral infu-
sions, together with postoperative pain scores for the first 48 h after surgery,
or earlier if the paravertebral infusion was stopped within 48 h of surgery.
Poor quality analgesia was defined as a score of 7 or more on the Visual Ana-
log/Smiley Faces Scale (‘VAS/SF’).
Results: A total of 81 children were ASA 1 status and two were ASA 3 status.
Analgesia comprised intravenous morphine at a mean dose of
20 lg
kg 1
h 1, together with oral paracetamol (62.5 mg
kg 1/24 h) and
ibuprofen (14.2 mg
kg 1/24 h). The mean paravertebral bupivacaine dose
was 0.29 mg
kg 1
h 1. Sixty-four patients (77.1%) had good quality analge-
sia, 17 (20.5%) patients had moderate quality analgesia, and only two
patients (2.4%) had poor quality analgesia.

722 © 2016 John Wiley & Sons Ltd

Pediatric Anesthesia 26 (2016) 722–726
T. Murphy et al.
Conclusion: Analgesic outcomes with this regimen appear to be very satisfac-
tory. It compares favorably with an epidural-based regimen.
Introduction
Pneumonia is a common illness in the pediatric popula-
tion with an estimated incidence of 34–40 cases per
1,000 children annually in preschool children in Europe
and North America (1). A pneumonic illness becomes
complicated if there is lung necrosis, lung abscess, para-
pneumonic effusion, or empyema thoracis (‘empyema’).
Common etiological organisms include Streptococcus,
Staphylococcus, and Haemophilus species. Fungal infec-
tions are rare, but are more common in immunocom-
promised patients. There are three stages of empyema
formation. Stage I is exudative, stage II fibrinopurulent
(with the development of loculations) followed by stage
III (organizational) (2). A prospective study in the north
east of England carried out from 2001 to 2002 showed
that 89% of 711 children presenting to hospital with
pneumonia were admitted; 96% received antibiotics,
70% intravenously. Twenty percent had lobar chest
X-ray (CXR) changes, 3% empyema, and 4% required
intensive care (3). The incidence of empyema in children
has increased noticeably since the early 1990s, although
the reasons for this are not entirely clear (4,5). In the
setting of complicated pneumonia, there are a variety of
management options, including antibiotics alone, pleu-
ral aspiration, thoracostomy tube drainage, intrapleural
fibrinolytics, thoracoscopic interventions or thoraco-
tomy, and decortication. There is no consensus on opti-
mal management within the United Kingdom (2,6,7) or
elsewhere.
Our institution serves as a tertiary referral center for
children with complicated pneumonia and clinically sig-
nificant pleural collections of fluid, pus, or both, or
where a bronchopleural fistula has developed. Patients
are managed with early primary thoracotomy and
decortication: in the lateral position, a limited postero-
lateral muscle-sparing thoracotomy incision is made,
and the thorax is entered in the fifth intercostal space.
This permits removal of pleural fluid followed by the
‘peeling off’ of any fibrinopurulent material that adheres
to both the visceral pleura and sometimes the inter-lobar
fissures. At the end of the operation, an intercostal chest
tube is left in situ.
The aim of this study was to establish whether an
infusion of local anesthetic via a surgically placed par-
avertebral catheter, supplemented with an intravenous
morphine infusion and simple oral analgesics (paraceta-
mol and ibuprofen), provided satisfactory postoperative
pain relief after this surgical procedure. Our primary
© 2016 John Wiley & Sons Ltd
Pediatric Anesthesia 26 (2016) 722–726
Analgesic management after thoracotomy
outcome measure was the patient’s pain score (using the
VAS/SF scale) recorded at regular intervals by the nurs-
ing/pain management staff in the standard pain docu-
mentation used for postoperative recording of these
infusions in our institution.
Methods
After receiving appropriate institutional approval, we
searched the theater records for the period March 2008
to November 2014 and identified 136 children who
underwent thoracotomy and decortication for
empyema. We obtained 91 sets of notes for review and
excluded five patients who were not managed with a par-
avertebral infusion and three in whom data were incom-
plete. This analysis therefore included 83 children for
whom demographic data, intraoperative management,
and postoperative analgesic use and pain scores for the
first 48 h after the surgical procedure were available.
Patients were generally managed with a standard
anesthetic regime. Sedating premedicants were usually
avoided as the majority of children were acutely unwell
and were unlikely to tolerate further respiratory depres-
sion. Following preoxygenation (where tolerated), anes-
thesia was induced with either an intravenous or
uncommonly an inhalational technique. The trachea
was intubated and both lungs ventilated in a conven-
tional manner as it is generally easier to peel off the
adherent material on the surface of the infected lung if
that lung remains ventilated. Anesthesia was maintained
with isoflurane or sevoflurane, an intravenous opiate
(fentanyl/morphine) together with paracetamol and a
nonsteroidal anti-inflammatory drug (‘NSAID’) unless
the patient had received a recent preoperative dose. In
our series of 83 patients, five (6%) received intercostal
nerve blocks before commencement of surgery.
At the end of the surgical procedure, the surgeon
placed a 16-gauge paravertebral catheter (19-gauge in
patients weighing <10 kg) under direct vision. A small
pocket of parietal pleura was stripped away into
which the tip of a catheter is placed. The catheter is
subsequently passed through the thoracic wall and
secured. The entrance to the pocket is fashioned to be
as small as possible in order to restrict the loss of
infusate into the pleural cavity. Usually, prior to clo-
sure of the thorax, a loading dose of local anesthetic
is given via the paravertebral catheter, the surgeon
confirming visually that there is satisfactory delivery
of local anesthetic into the paravertebral space. Mean
723
Analgesic management after thoracotomy
duration of surgery was 90 min (range 45–170). A
paravertebral infusion of 0.1% racemic bupivacaine
was subsequently commenced, most commonly after
the child has been extubated. Patients also received
intravenous morphine, delivered either as a patient-
or nurse-controlled infusion; it may have a back-
ground, background-bolus, or bolus only facility at
the discretion of the anesthetist. Morphine and bupi-
vacaine infusions continued usually for a maximum of
48–72 h but may be discontinued earlier if the inter-
costal chest tube is removed, the paravertebral bupi-
vacaine and parenteral morphine being discontinued
shortly thereafter. Paracetamol was given to every
patient. NSAIDs were given to every patient other
than one with a significant acute kidney injury.
Members of the nursing staff and pain team reviewed
patients regularly in the postoperative period. Observa-
tions and pain scores were entered on standardized
charts for patients on opiate and local anesthetic infu-
sions. We used a Numerical Rating Scale/Visual Analog
Scale (‘VAS’) or Faces (‘SF’) tool for pain assessment
(8,9). In the setting of postthoracic surgical pain, assess-
ment may be complicated by short-lived acute exacerba-
tions of pain that may accompany coughing, chest
physiotherapy, or drain removal.
We used the commonly accepted definitions of sever-
ity of pain (none/minimal, moderate, and high) that cor-
respond respectively to a pain score of 0–3, 4–6, and 7
or above using the VAS/SF.
Results
A total of 36 children (43%) were female. A total of 81
children were ASA 1 status and two, with severe cere-
bral palsy, were ASA 3. The median patient age was
42 months (range 5–200 months). The median weight
was 16 kg (range 6–78.6 kg). Eleven children were
admitted to the intensive care unit postoperatively
(Tables 1–3).
We did not observe any complications, including local
anesthetic systemic toxicity, relating to this technique.
Table 1 Perioperative drug doses
Mean loading dose of bupivacaine via paravertebral cathetera (mg
kg 1)
Mean duration of paravertebral infusion (h)
Mean paravertebral bupivacaine dose (mg
kg 1
h 1)
Mean duration of morphine infusion (h)
Mean morphine dose(lg
kg 1
h 1)
Mean paracetamol dose (mg
kg 1/24 h)
Mean ibuprofen dose (mg
kg 1/24 h)
a
A total of 81 patients (98%) received a loading dose of bupivacaine via the paravertebral catheter.
724
T. Murphy et al.
Table 2 Pain outcomes
Patients with severe pain (VAS/SF score > 7) 2 (2.4%)
Patients with moderate pain (VAS/SF 4–6) 17 (20.5%)
Patients with no/minimal pain (VAS/SF 0–3) 64 (77.1%)
The paravertebral catheter of one patient became
dislodged and was removed after only 2.3 h.
Discussion
Our results indicate that the number of patients with
severe pain, 2 (2.4%), was low. Seventeen patients
(20.5%) experienced moderate pain. The remainder of
patients had no or minimal pain. This is comparable to
another regional technique (10).
Regional techniques are associated with a recognized
failure rate with a corresponding requirement for rescue
analgesia: one mixed study reported by Lonnqvist et al.
(11) which included 48 patients aged 15 years or less
identified a failure rate of 6.2% for paravertebral blocks
in children. A direct comparison with these patients is
difficult as the failure rate in children was ‘arbitrarily
defined’ as a requirement for the equivalent of
16.7 lg
kg 1
h 1 of morphine; the paravertebral block
was provided by a single-shot technique rather than an
infusion; and the type of thoracic surgery may have been
different to our patients. Nevertheless, the mean mor-
phine consumption in our patients was 20 lg
kg 1
h 1.
This could reflect the fact that either thoracotomy or
decortication for empyema is a more painful procedure
than the surgical procedures in Lonnqvist’s series, or the
paravertebral infusions in our patients were ineffective
or, possibly, that we administered more morphine than
was necessary.
Thoracic epidurals are perhaps perceived to be the
gold standard against which other regional techniques
may be compared. In a retrospective, 10-year audit of
the use of thoracic epidurals for children undergoing tho-
racotomy and decortication, Kotze et al. concluded that
‘the quality of analgesia from a well-managed epidural
Range Interquartile range
1.37 0.37–3.13 0.96–1.93
39.3 2.3–221 23.6–45.2
0.29 0.11–0.50 0.20–0.36
38.2 11–120 22.5–47.0
20.0 5.2–83 10.8–26.4
62.5 24.2–85 53.1–76.0
14.2 4.3–30 10.1–18.6
© 2016 John Wiley & Sons Ltd
Pediatric Anesthesia 26 (2016) 722–726
T. Murphy et al.
Table 3 Additional analysis of postoperative morphine consumption
Number of patients
Mean morphine dose (lg
kg 1
h 1)a
Standard deviation
Mean paravertebral bupivacaine dose (mg
kg 1
h 1)
Standard deviation
The difference in morphine consumption between the two groups is significant, P = 0.01, Student’s t-test.
a
makes this the technique of choice. . .’ (10). Postopera-
tively, however, only 2 of 46 epidurals were managed
without the addition of fentanyl to the epidural infusate
and the contribution of the epidural opiate to the overall
analgesic effects of the epidural infusion may have been
significant. The incidence of moderate to severe pain was
17%, and the incidence of severe pain was 2%, which is
very similar to our study. The suggested audit standards
for these were 21% and 8%, respectively.
There has been a recent national audit on pediatric
epidural complications, and indications for the place-
ment of an epidural catheter (which include ‘open tho-
racic surgery’) are being refined (12). In addition, recent
evidence from adult thoracic anesthetic literature has
questioned the superiority of thoracic epidural over par-
avertebral regional block (13–15). These studies indi-
cated that, when the two techniques are compared,
analgesia is equivalent; side effects such as hypotension,
urinary retention, nausea, and vomiting are less com-
mon; and perhaps most importantly, indices of pul-
monary function are superior.
A number of other regional techniques have been
described in the setting of postthoracotomy pain relief
in children, including intrathecal, intrapleural, and inter-
costal nerve blocks; wound infiltration catheters (16);
and in adult patients, phrenic nerve block. There are
descriptions of both single-shot and continuous infusion
techniques. The benefits of a regional technique in the
setting of pediatric thoracic surgery have recently been
summarized by Bosenberg (17).
Evidence on the use of paravertebral catheters in chil-
dren is limited. The technique described by Carey et al.
(22 patients aged 0.5–16 years) included a ‘single-shot’
paravertebral block prior to surgical incision together
with oral medication postoperatively (18). Further
reports on a variety of settings for paravertebral block
suggest that the technique is satisfactory (19–22).
Gibson et al. (23) showed in a small study that there
was a significant difference in morphine consumption
between two groups of thoracic surgical patients, one
managed with a retropleural catheter plus rescue mor-
phine and the other with morphine alone, in whom the
morphine dose was 23 lg
kg 1
h 1.
© 2016 John Wiley & Sons Ltd
Pediatric Anesthesia 26 (2016) 722–726
Analgesic management after thoracotomy
Background infusion No background infusion
41 42
24.0 17.0
10.8 13.2
0.29 0.29
0.10 0.10
Hammer suggests a morphine infusion regimen for
older infants and children after thoracic surgery of
10–30 lg
kg 1
h 1 (24). The interquartile range for
morphine consumption for our 83 patients lies within
these limits. We analyzed mean morphine consumption
according to whether the infusion pump was set up
either with or without a background infusion rate.
Patients whose pump was set up without a background
rate received a significantly smaller dose of morphine.
Bupivacaine infusion rate among the two groups was
identical.
Active intrathoracic infection may be considered to be
a contraindication to a percutaneous paravertebral tech-
nique, although not necessarily to the surgical placement
of a paravertebral catheter (25). Patients in our series
were all receiving intravenous antibiotics prior to the
commencement of surgery which may make cutaneous
infective complications following the surgical placement
of a percutaneous paravertebral catheter less likely.
There are several limitations to our study, which was
retrospective and observational. Our intraoperative
management, though similar among all patients, was
not standardized. Postoperative morphine infusions
were not standardized (patients could receive morphine
in one of three ways—continuous infusion, infusion plus
boluses, or boluses alone). This may make it difficult to
define the effectiveness of the paravertebral block purely
by measuring postoperative morphine use. This study
fails to identify clearly which component of the postop-
erative analgesic regimen is responsible for the relief of
pain; it might be the case that the paravertebral infusion
of bupivacaine made a smaller contribution to analgesia
than the infusion of morphine. The pain and nursing
teams regularly assessed patients but there was no for-
mal testing of dermatomes to establish more precisely
the extent to which the paravertebral block was func-
tioning. We collected data for up to 48 h after the surgi-
cal procedure but the mean duration of paravertebral
infusion and morphine infusion was 39.3 and 38.2 h,
respectively. Generally, intravenous morphine and
paravertebral infusions were discontinued shortly after
the chest drain was removed and in our series, the
median duration of drain insertion was 29.5 h (range
725
Analgesic management after thoracotomy T. Murphy et al.

5–456 h). Only 12 patients (14%) had either a morphine
or paravertebral bupivacaine infusion of more than
48 h.
Although we did not observe any significant compli-
cations relating to the paravertebral infusion of bupiva-
caine in our series of patients, we could not conclude
that this technique has proven safety. However, we
maintain that the placement of the paravertebral cathe-
ter by the surgeon, under direct vision, is likely to be
more accurate than placement via a percutaneous tech-
nique.
Conclusion
We believe this to be the largest study reporting the use
of any regional technique after thoracotomy for decorti-
cation in children. We conclude that a combination of a
paravertebral infusion of bupivacaine, parenteral opiate,
and simple oral analgesics appears to be safe and pro-
vides satisfactory pain outcomes. It is the method of
choice in our institution for other pediatric thoracic sur-
gical procedures and some cardiac surgical procedures
performed via a thoracotomy. Although further studies
of the technique are warranted, we argue that there is
evidence to show it is a very reasonable alternative to an
epidural-based regime and may prevent some of the rare
but potentially serious complications associated with the
placement of an epidural catheter.
Ethics approval
Approval for the audit was granted by the hospital.
Funding
The study received no external funding.
Conflicts of interest
The authors report no conflict of interest.

References
1 McIntosh K. Community-acquired pneumo-
nia in children. N Engl J Med 2002; 346:
429–437.
2 Balfour-Lynn IM, Abrahamson E, Cohen G
et al. BTS guidelines for the management of
pleural infection in children. Thorax 2005; 60
(Suppl 1): i1–i21.
3 Clark JE, Hammal D, Spencer D et al. Chil-
dren with pneumonia: how do they present
and how are they managed? Arch Dis Child
2007; 92: 394–398.
4 Spencer D. Empyema thoracis: not time to
put down the knife. Arch Dis Child 2003; 88:
842–843.
5 Spencer DA, Iqbal SM, Hasan A et al.
Empyema thoracis is still increasing in UK
children. BMJ 2006; 332: 1333.
6 Kerek O, Hilliard T, Henderson J. What is
the best treatment for empyema? Arch Dis
Child 2008; 93: 173–174.
7 Alexiou C, Goyal A, Firmin RK et al. Is
open thoracotomy still a good treatment
option for the management of empyema in
children? Ann Thorac Surg 2003; 76: 1854–
1858.
8 The Association of Paediatric Anaesthetists
of Great Britain and Ireland. Good practice
in postoperative and procedural pain man-
agement, 2nd edition. Pediatr Anesth 2012;
22 (Suppl 1): 1–79.
9 Wong DL, Baker CM. Pain in children:
comparison of assessment scales. Pediatr
Nurs 1988; 14: 9–17.
10 Kotze A, Hinton W, Crabbe DC et al. Audit
of epidural analgesia in children undergoing
thoracotomy for decortication of empyema.
Br J Anaesth 2007; 98: 662–666.
11 Lonnqvist PA, MacKenzie J, Soni AK et al.
Paravertebral blockade. Failure rate and
complications. Anaesthesia 1995; 50: 813–
815.
12 Moriarty A. Pediatric epidural analgesia
(PEA). Pediatr Anesth 2012; 22: 51–55.
13 Daly DJ, Myles PS. Update on the role of
paravertebral blocks for thoracic surgery: are
they worth it? Curr Opin Anaesthesiol 2009;
22: 38–43.
14 Richardson J, Sabanathan S, Jones J et al. A
prospective, randomized comparison of pre-
operative and continuous balanced epidural
or paravertebral bupivacaine on post-thora-
cotomy pain, pulmonary function and stress
responses. Br J Anaesth 1999; 83: 387–392.
15 Davies RG, Myles PS, Graham JM. A com-
parison of the analgesic efficacy and side-
effects of paravertebral vs epidural blockade
for thoracotomy–a systematic review and
meta-analysis of randomized trials. Br J
Anaesth 2006; 96: 418–426.
16 Anell-Olofsson M, Lonnqvist P-A, Bitkover
C et al. Plasma concentrations of levobupi-
vacaine associated with two different inter-
mittent wound infusion regimens following
surgical ductus ligation in preterm infants.
Pediatr Anesth 2015; 25: 711–718.
17 Bosenberg A. Benefits of regional anesthesia
in children. Pediatr Anesth 2012; 22: 10–18.
18 Carey JA, Hamilton JR, Spencer DA et al.
Empyema thoracis: a role for open thoraco-
tomy and decortication. Arch Dis Child 1998;
79: 510–513.
19 Lonnqvist PA. Continuous paravertebral
block in children. Initial experience. Anaes-
thesia 1992; 47: 607–609.
20 Eng J, Sabanathan S. Continuous paraverte-
bral block for postthoracotomy analgesia in
children. J Pediatr Surg 1992; 27: 556–557.
21 Lonnqvist PA, Olsson GL. Paravertebral vs
epidural block in children. Effects on postop-
erative morphine requirement after renal sur-
gery. Acta Anaesthesiol Scand 1994; 38: 346–
349.
22 Karmakar MK, Booker PD, Franks R et al.
Continuous extrapleural paravertebral infu-
sion of bupivacaine for post-thoracotomy
analgesia in young infants. Br J Anaesth
1996; 76: 811–815.
23 Gibson MP, Vetter T, Crow JP. Use of con-
tinuous retro pleural bupivacaine in postop-
erative pain management for pediatric
thoracotomy. J Pediatr Surg 1999; 34: 199–
201.
24 Hammer GB. Pediatric thoracic anesthesia.
Anes Clinics N Am 2002; 20: 153–180.
25 Murphy T, McCheyne A. Feedback. Contin
Educ Anaesth Crit Care Pain 2011; 11: 70.

726 © 2016 John Wiley & Sons Ltd

Pediatric Anesthesia 26 (2016) 722–726