ORIGINAL ARTICLE

Efficacy of intravenous paracetamol compared to dipyrone

and parecoxib for postoperative pain management after
minor-to-intermediate surgery: a randomised, doubleblind trial
Gerhard Brodner, Wiebke Gogarten, Hugo Van Aken, Klaus Hahnenkamp, Carola Wempe, Hendrik Freise,
Irmgard Cosanne, Markus Huppertz-Thyssen and Bjorn Ellger
Background and objective Paracetamol has a well established
pharmacological profile, but its postoperative efficacy is in
question. This double-blind, placebo-controlled study was
designed to compare the efficacy of intravenous paracetamol
with other intravenous non-opioids as part of a multimodal
concept for perioperative pain therapy.
Methods Patients undergoing minor-to-intermediate surgery
under general anaesthesia were randomly assigned to receive
infusions of paracetamol (1 g every 6 h), dipyrone (1 g every
6 h), parecoxib (40 mg every 12 h) separated by infusions of
physiological saline 0.9%, or placebo (0.9% saline every 6 h),
respectively, for at least 48 h as part of a multimodal pain
concept. Patient-controlled piritramide was administered as
rescue medication. Dependent variables were recorded 1, 6,
18, 30 and 42 h after extubation and 1 week after surgery.
Surgical and associated pain was scored as the primary

Introduction
Non-opioid analgesics are commonly used in a balanced
multimodal analgesia regimen in the perioperative
period. The intention is to provide satisfactory relief
for minor discomfort or, for severe pain, in combination
with an opioid, to improve pain relief and decrease opioid
consumption.1 The ideal non-opioid should be well
tolerated, provide rapid and effective pain relief, should
be convenient and easy to administer and cheap. Intravenous administration is the route of choice when oral or
rectal administration is not possible. At present, nonsteroidal anti-inflammatory drugs (NSAIDs), coxibs,
dipyrone and paracetamol are available for this purpose.
Recent evidence has shown that NSAID2 5 and coxibs6
are associated with renal, gastrointestinal, cardiac and
haematological complications. Furthermore, dipyrone
carries the risk of neutropenia and agranulocytosis.7
Paracetamol is available as an intravenous preparation
From the Department of Anaesthesiology, Intensive Care and Pain Therapy,
Fachklinik Hornheide, Muenster (GB, IC), Department of Anaesthesiology,
Intensive Care and Pain Therapy, Hospital Harlachingen, Munich (WG),
Department of Anaesthesiology and Intensive Care Medicine, University Hospital
of Muenster, Muenster (HVA, KH, CW, HF, BE), Department of Anaesthesiology,
Intensive Care und Pain Therapy, St Marien-Hospital, Dueren (MHT), Germany
Correspondence to Bjorn Ellger, MD, PhD, Department of Anaesthesiology and
Intensive Care Medicine, University Hospital of Muenster, Albert-SchweitzerStrasse 33, D-48149 Muenster, Germany
Tel: +49 251 83 47251; fax: +49 251 88704;
e-mail: ellger@anit.uni-muenster.de
0265-0215 2011 Copyright European Society of Anaesthesiology

outcome on a visual analogue scale. Additionally, time to first dose

and total piritramide dosage, satisfaction, respiratory depression,
nausea, vomiting, sedation, itching and sweating were recorded.
Results A total of 196 patients were recruited. The efficacy of
paracetamol was similar to that of the other non-opioid
analgesics. Surgical pain was reduced with all non-opioids
compared to placebo; there was no effect on associated pain.
Piritramide dosage and incidence of side effects were not
reduced.
Conclusion Intravenous paracetamol has equivalent efficacy to
non-opioids dipyrone and parecoxib that improves postoperative
pain therapy when used as part of a multimodal concept
after minor-to-intermediate surgery.
Eur J Anaesthesiol 2011;28:125132
Published online 4 October 2010
Keywords: analgesia; pain, postoperative; paracetamol

and has a well established pharmacological profile with

only few contraindications.8 It might be the non-opioid
of choice for most in- and outpatient procedures but data
on its efficacy are conflicting. Paracetamol was less
effective than NSAIDs and coxibs after dental surgery
and less effective than dipyrone after lumbar discectomy, but after extremity or breast surgery it was as
effective as ketorolac or dipyrone.9 12 This apparent
difference might be explained, at least in part, by pharmacokinetic interactions with comedication such as
5-HT(3)-receptor antagonists that are frequently given
for prophylaxis of postoperative nausea and vomiting
(PONV) and might reduce its analgesic effect.13
Although a tailored individual pain concept might be
desirable, daily clinical practice requires the implementation of a straightforward standard protocol that
is safe and effective, and here there is room for improvement.14 Most protocols employ non-opioids as the basis
of their multimodal perioperative pain plan, supplemented, if necessary, by opioids and regional anaesthesia.15 It is, however, unclear which non-opioid best meets
the needs of patients undergoing a variety of different
types of surgery.
The aim of this study was to compare the efficacy of
intravenous paracetamol as the basis of a routine
clinical regimen providing pain relief after a variety
of minor and intermediate surgical procedures, with
DOI:10.1097/EJA.0b013e32833fedfa

Copyright European Society of Anaesthesiology. Unauthorized reproduction of this article is prohibited.

126 Brodner et al.

two other commonly used intravenous non-opioids

(dipyrone and parecoxib) and placebo. Furthermore,
we sought to differentiate effects on two different
kinds of pain, surgical and associated pain, and record
relevant side effects.

Methods
The protocol was approved on 2 August 2003 by the
Ethics Committee of the Medical Association WestfalenLippe and the Faculty of Medicine, University of
Muenster, Germany (chair: Professor Dr O. Schober,
protocol number 3II-Aken3).
After written informed consent, patients aged 1875 years
undergoing elective minor-to-intermediate surgery
under general anaesthesia in our university centre were
enrolled in this prospective, double-blind, placebo-controlled study. We included consecutive patients undergoing the following types of surgery: plastic surgery
(breast surgery, inguinal or axillary dissections), oral
and maxillofacial surgery (correction of retrognathism
and prognathism), gynaecological (laparoscopy, breast
surgery) and urological (cystoscopy, transurethral prostratectomy) surgery and orthopaedic surgery (hip endoprosthesis for coxarthrosis).
Primary exclusion criteria were any contraindication for
non-opioid analgesics (significant coronary artery stenosis, pulmonary disorders, liver or kidney dysfunction,
haematological disorders); increased intracranial pressure; history of alcohol or drug abuse; hypersensitivity to
any of the study drugs; the American Society of
Anesthesiologists (ASA) class status more than 3; pregnancy or breast feeding; intake of any non-opioid within
24 h before administration of the study drugs and
reduced understanding due to mental disorders or
poor language comprehension. Secondary exclusion
criteria were difficulty with immediate postoperative
extubation, need for postoperative intensive therapy
and insufficient postoperative analgesia: that is, pain
scores more than 40 on a visual analogue scale (VAS)
with endpoints 0 meaning no pain and 100 meaning
worst pain imaginable.
Patients were assigned by random numbers to one of four
groups receiving intravenous paracetamol 1 g every 6 h
(group 1 paracetamol), dipyrone 1 g every 6 h (group 2
dipyrone), parecoxib 40 mg every 12 h (group 3 parecoxib) or placebo (0.9% saline) every 6 h (group 4
placebo) for at least 48 h (Fig. 1). If analgesia was
required for more than 48 h, treatment was continued
according to the protocol and randomisation as long
as necessary.

bottles were labelled with patient number and time of

administration. Infusions were administered by a blinded
attending physician. Group parecoxib received saline
between the two administrations, giving every patient
similar 6-h schedules of intravenous infusions. The
first dose of a non-opioid was started 30 min before the
end of surgery and all drugs were infused over 15 min.
In the postanaesthesia care unit (PACU), a patientcontrolled analgesia (PCA) pump containing piritramide
(2 mg ml1) was connected to the intravenous infusion.
The PCA device was programmed to deliver 2 mg piritramide boluses on demand with a lockout time of 10 min.
A background infusion was not provided due to a possible
increased risk of respiratory depression. PCA was stopped
when the patient made fewer than three demands per
24 h or prior to discharge.
General anaesthesia was induced with propofol
(2 mg kg1), sufentanil (0.25 mg kg1) and rocuronium
(0.6 mg kg1). After tracheal intubation, all were ventilated with 30% oxygen in air, and anaesthesia was maintained with sevoflurane (11.5 Vol%). Anaesthesia was
monitored using standard clinical methods. Further
sufentanil administration, fluids, transfusions and other
procedures were left to the discretion of the attending
anaesthesiologist. Dehydrobenzperidol 0.625 mg was
given for routine PONV prophylaxis. Dimenhydrinate
was added as rescue medication for those vomiting in the
recovery room. When emesis required additional treatment by 5-HT(3)-receptor antagonists, which affect the
action of paracetamol, these patients were excluded from
the pain analysis.13
The PCA devices and the use of VASs were explained to
patients preoperatively. If a patient experienced insufficient pain relief (surgical VAS >40), rescue analgesia was
provided as additional boluses of intravenous piritramide
2 mg until pain relief was satisfactory (VAS <40). If this
was unsuccessful, or if important side effects potentially
attributable to the study drug (anaphylaxis, vasoplegia)
occurred within 30 min, the patient was excluded from
the study and received treatment according to the standard ward protocol: nurse-provided titration of piritramide infusion 0.1 mg kg1, which was repeated every
30 min until the patient reported a VAS score less than 40.

Patients, all care providers and the clinical observer.s who

scored were blinded to the study allocation.

Investigators blinded to the study allocation began

recording the study variables once 60 min had passed
since extubation. The next assessment was after 6 h, on
the evening of the day of surgery, thereafter at 12-h
intervals until the morning of the second postoperative
day (18, 30, 42 h) and 1 week after surgery in patients who
had not been discharged. To control for potential confounders, pain in the same anatomical location as the site
of surgery was assessed prior to the operation at rest and
during movement.

All study drugs were prepared by the hospital pharmacy

in identical glass bottles as infusions of 100 ml. The

Dynamic VAS for pain localised to the site of surgery was

the primary outcome variable. The dose of opioid was of

European Journal of Anaesthesiology 2011, Vol 28 No 2

Copyright European Society of Anaesthesiology. Unauthorized reproduction of this article is prohibited.

Intravenous non-opioids for postoperative pain therapy 127

Fig. 1

Enrollment
Assessed for eligibility (n = 202)

Excluded (n = 6)
Not meeting inclusion criteria (n = 0)
Declined to participate (n = 6)
Other reasons (n = 0)

Randomised (n = 196)

Allocated to intervention (n = 196)

Received allocated intervention (n = 196)
Did not receive allocated intervention (n = 0)

Allocation

Group paracetamol (n = 49)

Group dipyrone (n = 49)

Group parecoxib (n = 49)

Group placebo (n = 49)

Lost to follow-up 42 h (n = 0)

Lost to follow-up 42 h (n = 0)

Lost to follow-up 42 h (n = 0)

Lost to follow-up 42 h (n = 0)

Discontinued after at least 2 doses

Discontinued after at least 2 doses

Discontinued after at least 2 doses

Discontinued after at least 2 doses

of study medication (n = 4):

Pain (n = 1)
Re-operation (n = 1)
Protocol violation (n = 2)

of study medication (n = 8):

Pain (n = 2)
Emesis (n = 2)
Protocol violation (n = 1)

of study medication (n = 5):

Resp. depression (n = 1)
Emesis (n = 2)
Protocol violation (n = 1)

of study medication (n = 4):

Headache (n = 1)
Protocol violation (n = 1)

Withdrew consent (n = 3)

Withdrew consent (n = 1)

Withdrew consent (n = 2)

Follow-up 7 days
Eligible for evaluation n = 40

Eligible for evaluation n = 41

Eligible for evaluation n = 41

Eligible for evaluation n = 39

Withdrew consent n = 1
Protocol violation n = 2
Lost because of discharge n = 6

Withdrew consent n = 3
Protocol violation n = 1
Lost because of discharge n = 4

Withdrew consent n = 1
Protocol violation n = 1
Lost because of discharge n = 6

Withdrew consent n = 2
Protocol violation n = 1
Lost because of discharge n = 7

Analysis
Flow chart of the study design. Administration of the study drug was started 30 min prior to the end of surgery and was then repeated at 6-h intervals
for at least 48 h. Scores were then recorded at 7 days after surgery.

minor relevance because it did not permit assumptions

regarding pain control or comfort, particularly because
less opioid does not necessarily mean that there
are less opioid-related side effects.16 Considering the
various biological, psychological and sociological
aspects of pain, the two following types of pain were
differentiated: surgical pain (pain localised to the site
of surgery) and associated pain (acute pain not localised to the site of surgery, such as headache, or
musculoskeletal). There were several secondary outcome variables: the time to first piritramide bolus and
piritramide consumption as quantified by the number
of boluses demanded and administered; satisfaction

rated as 1, excellent; 2, good; 3, moderate; 4, insufficient; and 5, poor. The following side effects of nonopioids and opioids were recorded: respiratory depression (1 normal respiratory rate; 2 respiratory rate of
812 breaths min1; 3 respiratory rate of <8 breatheaths min1); nausea and vomiting (1 no; 2 yes);
sedation (1 awake, 2 tired, easy to wake up,
3 asleep, can easily be woken by light glabellar
tap, 4 coma, deeply sedated, only sluggish response
to stimuli); itching (1 no, 2 sometimes, a little,
3 bearable, 4 intolerable) and sweating (1 no,
2 for 5 min after infusion, 3 for 30 min after infusion,
4 diffused, often).
European Journal of Anaesthesiology

2011, Vol 28 No 2

Copyright European Society of Anaesthesiology. Unauthorized reproduction of this article is prohibited.

128 Brodner et al.

were assessed by x2 test or Fishers exact test, if matched

cells were rare (expected frequencies <5). Ordinal variables were reported as median [interquartile range
(IQR)], and metric variables were described as mean
(SD). KruskallWallis test, analysis of variance
(ANOVA) and repeated-measures ANOVA were used
to compare groups.

Patient characteristics, medical history, physical status,

medication, duration of surgery and anaesthesia, blood
loss, fluid balance, transfusions and length of hospital stay
were recorded in a standardised protocol.
A significant difference among groups in the primary
outcome was defined according to two findings from
previous studies.17,18 First, VAS scores revealed large
variations with SDs of 2040 on a 100-point VAS
scale. Second, VAS scores were found to correlate with
a 4-point numeric rating scale. This indicates that
patients cluster the VAS scores into several distinct
categories representing different degrees of pain.
Hence, for our power calculation, assuming a SD of
30, we considered a difference of 15 VAS points, that
is, a half SD, as a significant and relevant difference
between two such clusters. According to the criteria
defined by Cohen,19 this corresponds to a medium
effect size. A sample size of 196 patients would allow
us to detect an effect of this magnitude of f equal to 0.3
in VAS pain scale (significance level: a 0.05, statistical power: 1b 0.95).

Results
Two hundred and two patients were asked to participate in this study. Six did not consent, leaving 196 for
randomisation. In 21 patients, the medication was
stopped after at least two doses due to events given
in Fig. 1. Those who withdrew consent or were subject
to protocol violations did not receive further study
medication.
A total of 22.1% of the patients underwent plastic surgery,
26.7% oral and maxillofacial surgery, 25.6% gynaecological or urological surgery and 25.6% orthopaedic surgery
for hip endoprosthesis.
Personal characteristics and selected background variables of the four groups did not differ significantly
among groups (Table 1) with three exceptions: patients
of group 3 parecoxib had a significantly shorter duration
of anaesthesia and needed significantly less intraoperative sufentanil compared to group 4 placebo, and there

Analysis was supported by Statistical Package for the

Social Sciences (SPSS 12.0; SPSS Inc., Chicago, USA;
2005) and was performed using the intention-to-treat
principle. Nominal variables were described as relative
and absolute frequencies; differences among groups
Table 1

Patient characteristics, physical condition, intraoperative characteristics and postoperative recovery

Paracetamol (n 49)

Dipyrone (n 49)

Parecoxib (n 49)

Placebo (n 49)

Variable

Mean

SD

Mean

SD

Mean

SD

Mean

SD

Age (years)
BMI (kg m2)
Number of cigarettes
Preoperative pain (surgical
area; VAS score)
S intraoperative Sufentanil (mg)M
Intraoperative blood loss (ml)
Duration of surgery (min)
Duration of anaesthesia (min)M
Duration of stay in the
postoperative recovery
room (min)
First mobilisation (h)
Hospital stay (days)

50.5
25.4
2.67
9.2

17.5
4.4
7.6
17.1

45.5
25.3
3.2
10.8

17.9
5.5
9.8
17.2

49.4
27.7
3.3
13.3

14.6
6.1
7.4
16.6

42.8
25.3
4.9
6.0

16.8
4.0
8.7
13.2

37.6
229.6
96.4
137.6
100.9

14.2
250.0
54.2
64.0
57.3

43.1
306.7
113.2
154.7
111.2

20.4
258.4
67.4
67.8
59.0

34.4
224.1
82.9
125.3
103.67

12.2
247.6
50.0
63.8
46.4

44.5
177.6
110.2
164.6
115.9

18.6 M(Parecoxib/placebo)
168.6
59.0
75.4 M(Parecoxib/placebo)
193.3

15.1
11.0

11.8
3.6

16.5
11.1

13.0
6.5

13.8
10.9

12.8
5.5

16.6
10.3

Sex (male/female)
ASA physical status
(I/II/III/IV)
Diabetes mellitus (diet/oral
antidiabetics/insulin)
Mild chronic obstructive
pulmonary disease (yes)
Coronary artery disease
(no risk/low risk/high risk)
Arterial hypertension (yes)
Renal insufficiency
(compensated retention)
Liver disease (no/transamin.
<100/transamin. 100)

rel. frequency (%)

rel. frequency (%)

rel. frequency (%)

rel. frequency (%)

26.5/73.5
32.7/51.0/16.3/0

44.9/55.1
28.6/59.2/12.2/0

26.5/73.5
18.3/65.3/16.3/0

49.0/51.0
42.9/49.0/13.3/0

95.9/2.0/2.0

100 //0/0

95.9/4.1/0

98.0/2.0 7 0

8.2

6.1

8.2

83.7/16.3/0

83.7/16.3/0

83.7/16.3/0

87.8/14.3/0

34.7
0

28.6
2.0

24.5
0

28.1
4.1

98.0/2.0/0

100/0/0

95.9/4.1/0

95.9/4.1/0

8.4
3.9

Values are presented as means with SD or relative frequency [rel. frequency (%)]. ASA, American Society of Anaesthesiologists. MP < 0.05.

European Journal of Anaesthesiology 2011, Vol 28 No 2

Copyright European Society of Anaesthesiology. Unauthorized reproduction of this article is prohibited.

were more women in group 1 paracetamol and group 3

parecoxib than in group 2 dipyrone and group
4 placebo.

(10.48.6)
(9.010.1)
(12.14.7)
(13.13.0)
(13.23.1)
MMM

P < 0.01.

Prior to surgery, there were no significant differences in

VAS among groups regarding pain localised to the site of
surgery during rest (range of means: 6.013.3) or movement (range of means: 11.023.8).

1 h, 1 h after extubation; 6 h, 6 h after extubation; 18 h, 18 h after extubation; 30 h, 30 h after extubation; 42 h, 42 h after extubation. Mean, mean score. MP < 0.10,

1.1
3.8
4.0
4.5
4.0
(9.89.3)
(14.05.3)
(8.88.2)
(7.68.7)
(7.29.1)
(3.915.3)
(5.413.8)
(3.313.6)
(4.311.9)
(3.912.6)
24.5
26.3
18.9
14.8
12.3
(14.8)
(15.8)
(12.7)
(12.3)
(11.7)
18.5
17.7
13.5
11.5
9.0

(17.7)
(18.4)
(15.2)
(13.6)
(12.9)

25.4
25.7
22.7
19.8
17.3

(16.6)
(16.5)
(15.4)
(16.2)
(16.9)

5.7
4.2
5.1
3.8
4.3

0.2
4.3
0.3
0.5
1.0

(10.78.4)
(13.45.8)
(12.44.4)
(12.53.5)
(12.14.0)

5.9
8.6M
5.5
3.3
3.4

(15.53.6)
(18.11.0)
(13.92.9)
(11.44.9)
(11.75.0)

MM

P < 0 0.05,

0.9
0.5
3.7
5.0
5.0
6.9
8.0
9.2MM
8.3MM
8.4MM

(16.42.7)
(17.51.5)
( 17.5 to 0.9)
(16.3 to 0.2)
(16.6 to 0.1)

(16.43.3)
(16.14.9)
(19.8 to 0.9)
(20.0 to 3.4)
(19.7 to 1.0)
6.5
5.6
10.3MM
11.7MMM
10.4MM
(17.91.7)
(22.9 to 2.0)
(22.5 to 3.8)
(20.7 to 4.2)
(20.5 to 1.6)
8.1
12.4MM
13.1MMM
12.4MMM
11.1MM
(11.48.3)
(17.33.7)
(12.26.6)
(9.27.7)
(10.38.9)
1.6
6.8
2.8
0.7
0.7
(12.27.6)
(18.82.3)
(19.9 to 1.0)
(18.9 to 2.4)
(17.90.6)
2.3
8.3
10.4MM
10.6MMM
8.6M
(5.714.1)
(13.38.0)
(9.69.4)
(7.39.5)
(7.611.1)
4.2
2.7
0.1
1.1
1.7
(4.115.7)
(6.414.7)
(6.712.1)
(6.610.2)
(7.011.9)
5.8
4.1
2.7
1.8
2.4
(19.2)
(22.3)
(22.3)
(17.2)
(20.2)
33.8
35.6
32.1
28.9
24.6
(16.7)
(16.9)
(15.8)
(15.3)
(14.7)
27.2
30.0
21.8
17.2
14.2
(15.3)
(16.9)
(12.0)
(12.1)
(13.3)
25.7
23.2
19.0
16.4
13.5

Dipyrone
vs. parecoxib
Paracetamol
vs. placebo
Parecoxib
(n 49)
Mean (SD)
Dipyrone
(n 49)
Mean (SD)
Paracetamol
(n 49)
Mean (SD)

Surgical pain
1h
31.5 (17.5)
6h
27.3 (16.3)
18 h
21.7 (13.2)
30 h
18.3 (11.7)
42 h 16.0 (14.3)
Associated pain
1h
24.3 (17.5)
6h
21.9 (15.8)
18 h
18.6 (14.9)
30 h 15.3 (13.0)
42 h 13.3 (13.1)

Table 2

Visual analogue scale scores

Placebo
(n 49)
Mean (SD)

Paracetamol
vs. dipyrone

Paracetamol
vs. parecoxib

Mean difference (95% confidence interval)

Dipyrone
vs. placebo

Parecoxib
vs. placebo

Intravenous non-opioids for postoperative pain therapy 129

Four patients were excluded following administration of

5-HT(3)-receptor antagonists for emesis. Three patients
were excluded from analysis because pain relief was
inadequate with non-opioid and additional piritramide
2 mg in the recovery room (Fig. 1). The number of
piritramide administrations in the recovery room did
not differ among groups: group 1 paracetamol 7.6
(8.3), group 2 dipyrone 5.9 (7.3), group 3 parecoxib 7.8
(9.2) and group 4 placebo 6.4 [(7.3); P 0.6]. Postoperative surgical pain did not differ significantly among the
groups receiving a non-opioid analgesic. From the morning after surgery, these groups reported less surgical pain
than those receiving placebo. Additionally, the administration of dipyrone decreased associated pain compared
to placebo (Table 2). There was no significant difference
in time to first piritramide bolus and the total number
administered or demanded (Table 3). One week after
surgery, surgical pain did not differ among the four
groups except for significantly lower scores in group 1
paracetamol compared to group 3 parecoxib. Scores for
associated pain did not differ among the four groups
(Table 4).
All patients were satisfied with their pain treatment,
there was no difference among groups and satisfaction
improved with time after surgery [mean score in groups
(range): 1 h, 1.9 (1.82.0); 6 h, 1.6 (1.41.8); 18 h, 1.4 (1.3
1.5); 30 h, 1.3 (1.21.4); 42 h, 1.4 (1.31.4), P for group
effect 0.18, P for time effect 0.001 and P for group
interaction 0.17).
The incidence of side effects did not differ among groups
(Table 5). Nausea and vomiting were frequent; the range
of relative frequencies for nausea was between 30.6 and
42.9% and the range of relative frequencies for vomiting
was between 18.4 and 24.5%. Respiratory depression with
1012 breaths per minute was observed in the evening of
the day of surgery in three patients (two in group 3
parecoxib and one in group 4 placebo). Significant sedation was not observed.

Discussion
In this prospective, double-blind, placebo-controlled
trial, we showed that the analgesic efficacy of paracetamol
was no different from that of dipyrone and parecoxib
when used as the basis of a multimodal approach to treat
postoperative pain following a variety of minor or intermediate surgical procedures. All three non-opioid analgesics under study provided good patient satisfaction and
superior pain relief compared to placebo. There were no
European Journal of Anaesthesiology

2011, Vol 28 No 2

Copyright European Society of Anaesthesiology. Unauthorized reproduction of this article is prohibited.

130 Brodner et al.

Table 3

Piritramide dosage

Time to first piritramide bolus (min)

Total number of piritramide boluses administered
Total number of piritramide boluses demanded

Paracetamol (n 49)
Mean (SD)

Dipyrone (n 49)
Mean (SD)

Parecoxib (n 49)
Mean (SD)

Placebo (n 49)
Mean (SD)

61.5 (39.1)
31.0 (27.1)
46.9 (64.1)

104.4 (179.8)
29.7 (27.1)
35.6 (34.2)

89.4 (188.6)
39.1 (32.9)
49.7 (43.9)

100.2 (163.9)
40.0 (35.1)
55.1 (59.5)

0.53
0.24
0.35

Mean, mean score; P, probability.

relevant differences in their effects on different kinds

of pain.
The efficacy of non-opioids in the perioperative setting is
a matter of intense debate. When opioids are required to
supplement non-opioids, the opioid-sparing effect is
relatively modest and might be of minor clinical importance because the reduction in total dose might not
necessarily lead to a reduction in opioid-related side
effects.16 Despite the debate, paracetamol, dipyrone
and the selective cyclooxygenase-2 inhibitors are widely
used in the perioperative period.8,20 There has been
renewed interest in paracetamol since it became available
as a parenteral preparation with minor adverse effects.8 It
does not cause gastric irritation or interfere with coagulation21 and has no relevant cardiac, renal and pulmonary
side effects. When used in therapeutic doses, hepatotoxicity appears an insignificant risk.8 Hence, it has
become a mainstay of many multimodal postoperative
analgesia regimen,22 but the analgesic efficacy is still
debated,912 especially as its pharmacokinetic is affected
by other drugs used in the perioperative setting.13
Although they may have a more pronounced risk profile,
other intravenous non-opioids like dipyrone or parecoxib
are frequently preferred. Indeed, dipyrone is not licensed
in many countries due to potential risk of serious blood
dyscrasias.7 Parecoxib and consequently the entire group
of selective cyclooxygenase-2 inhibitors are under review
because of the risk of cardiac complications,16,23 and
impaired wound healing and osteogenesis.6
In our study, the efficacy of all tested non-opioids was
comparable. The sample size was not big enough to
detect differences of a subtle nature and we should regard
these as clinically irrelevant. Our power analysis was
based on the assumption that detection of a difference
of 15 points or more on a 100-point VAS scale was
clinically relevant and should form the basis of our trial.
A non-opioid analgesic effect of this magnitude might be
expected from published data.17,18,24,25 Our approach is
perhaps more conservative than that of a similar study, in
which a difference of 1 point on a 5-point rating scale was
regarded as significant.26 It is unlikely that important
clinical differences were hidden by insufficient statistical
power.
The opioid-sparing effect represented by the total
number of piritramide doses administered did not differ
significantly among the four groups. Patients in the

paracetamol and dipyrone group used fewer boluses

compared to placebo, and it is of interest that despite
taking slightly less opioid, the quality of pain relief as
indicated by the satisfaction scores was better.
Although this failed to achieve statistical significance
in this study, it concurs with the findings of a metaanalysis that demonstrates a minor opioid-sparing effect
of paracetamol and coxibs over 24 h after surgery.16 It is
possible to speculate that higher satisfaction scores
might be related to fewer opioid side effects in these
groups, but sound data are lacking, and conclusions on
side effects based on our results must be treated with
caution.
Episodes of nausea and vomiting were lower in the
paracetamol group compared to placebo, but this did
not achieve statistical significance. Again our results
compare with data16 that report the incidence of nausea
at 4351% and postoperative vomiting at 1822% in
patients receiving NSAID, coxibs or paracetamol. This
meta-analysis found a slight reduction in the incidence of
nausea and vomiting when a non-opioid is added to an
opioid-based postoperative pain protocol.
Significant sedation was not observed in our study; however, one patient (2%) receiving parecoxib was excluded
from analysis due to respiratory depression 6 h after
surgery, and mild respiratory depression of 812 breaths
min1 during sleep was observed in two more patients
(one in group parecoxib and one in group placebo). This
also corresponds to earlier findings in which a 33.9%
incidence of respiratory depression was observed in
patients treated with paracetamol or NSAID. Our results
together with published evidence cast doubt on a major
opioid-sparing effect of non-opioid analgesics and do not
support the notion that their use decreases the incidence
of opioid side effects.
The non-opioids decreased the mean VAS of 30 in
patients receiving placebo to a mean VAS of approximately 20 in patients receiving a non-opioid. The clinical
relevance and the true benefit of this reduction remain
unclear because the use of a PCA device is subject to
individualised criteria that include personal pain
thresholds. Moreover, despite a higher VAS score in
the placebo group, satisfaction did not differ among
groups. This raises the question of whether a reduction
of 1020 VAS points within the range of bearable pain
actually makes a real difference for the patient. These

European Journal of Anaesthesiology 2011, Vol 28 No 2

Copyright European Society of Anaesthesiology. Unauthorized reproduction of this article is prohibited.

5.9 (1.313.1)
6.8 (1.214.7)
0.8 (6.68.2)
0.0 (8.28.2)
5.1 (12.42.2)
6.8 (14.91.3)
1.3 (8.56.0)
0.9 (7.29.0)
7.2 M (14.3 to 0.1)
5.8 (13.82.2)

data underline the need for studies about the validity and
metric characteristics of pain ratings. However, as the
quality of analgesia is the most relevant criterion in
studies on pain therapy, we are left with VAS scores as
the only clinically applicable outcome variable.
Interestingly, some patients managed their pain without any opioid; the non-opioid and even placebo
reduced pain to a certain extent. It remains speculative
just how many of these patients might be satisfied
without any regular analgesic, leading to requests for
rescue opioid. Any potential benefits that might arise
from non-opioid protocols would be lost if these individuals were excluded.
Although our data on opioid consumption did not reach
statistical significance, our results are largely in line with
previous findings showing that non-opioids as part of a
multimodal approach might not only reduce the amount
of administered opioids but also tend to improve the
quality of perioperative pain therapy.16 It is tempting to
speculate that in daily routine care, providers, and perhaps patients as well, fear high doses of opioids and refuse
to comply with their administration. Non-opioids might
supplement modest doses of opioid to obtain pain
relief and thus make the patient and care provider feel
more comfortable.

P < 0 0.05.
Mean, mean score.

7.1 (9.4)
5.6 (10.7)
5.0 (7.9)
6.5 (11.2)

12.2 (14.2)
12.4 (15.4)

6.3 (12.6)
5.6 (12.7)

2.1 (9.45.3)
0.9 (7.39.2)

Parecoxib vs.
placebo
Dipyrone vs.
placebo
Dipyrone vs.
parecoxib
Paracetamol vs.
placebo
Paracetamol vs.
parecoxib
Paracetamol vs.
dipyrone
Placebo
(n 39)
Mean (SD)
Parecoxib
(n 41)
Mean (SD)
Dipyrone
(n 43)
Mean (SD)
Paracetamol
(n 40)
Mean (SD)

Surgical pain
Associated pain

Table 4

Visual analogue scale scores 1 week after surgery

Mean difference (95% confidence interval)

Intravenous non-opioids for postoperative pain therapy 131

NSAIDs other than those tested in our trial might be

equally effective after different kinds of surgery as a
few studies that include ketorolac and diclofenac
show.11,27,28 However, in the perioperative period, the
parenteral application of drugs has advantages and intravenous NSAIDs are not available in Germany. Published
studies tend to favour one or other drug for specific types
of surgery.912,20 Although it might be tempting to use
the data to provide tailor-made analgesic protocols for
individuals, in the daily routine of a hospital where
every kind of surgery is performed, and pain therapy is
devolved to various members of the care team with
different degrees of training and responsibility, this individualised approach might add to the existing difficulties
of postoperative pain management.14 Implementation of
a safe, effective and straightforward pain protocol that
meets most patients needs is a reasonable goal. As our
study has shown that paracetamol was at least as effective
as the other non-opioids for various forms of surgery and
showed few side effects, it can be recommended as the
analgesic of first choice following minor-to-intermediate
surgery.
In summary, analgesic efficacy of intravenous paracetamol as the basis of a multimodal therapeutic concept
is not different from the efficacy of dipyrone and parecoxib following different kinds of minor-to-intermediate surgical procedures. The analgesic efficacy of these
non-opioids was superior to placebo; however, the PCA
piritramide dosage and the incidence of opioid-related
side effects were not reduced. Paracetamol did not cause
European Journal of Anaesthesiology

2011, Vol 28 No 2

Copyright European Society of Anaesthesiology. Unauthorized reproduction of this article is prohibited.

132 Brodner et al.

Table 5

Side effects

Relative frequency
Episode of:
Nausea
Vomiting
Itching
Respiratory depression (812 breaths min1)
Number of episodes with side effects (none/1/2 /3)
Sweating
Repeated measures ANOVA

Paracetamol (n 49)
rel. frequency (%)

Dipyrone (n 49)
rel. frequency (%)

Parecoxib (n 49)
rel. frequency (%)

Placebo (n 49)
rel. frequency (%)

30.6
18.4
0.0
0.0
59.2/22.4/18.4/0.0
8.2

34.7
18.4
2.0
0.0
59.2/22.4/14.3/4.1
10.2

40.8
18.4
0.0
4.1
59.2/16.3/20.4/4.1
8.2

42.9
24.5
2.0
2.0
55.1/18.4/26.6/0.0
4.1

Mean (SD)

Mean (SD)

Mean (SD)

Mean (SD)

2.0
1.6
1.1
1.0
1.0

2.1
1.7
1.2
1.1
1.1

SedationM
1h
6h
18 h
30 h
42 h

(0.5)
(0.6)
(0.2)
(0.0)
(0.3)

(0.6)
(0.7)
(0.4)
(0.4)
(03)

2.0
1.7
1.2
1.1
1.0

(0.5)
(0.6)
(0.5)
(0.4)
(0.1)

2.1
1.8
1.2
1.2
1.1

(0.7)
(0.7)
(0.5)
(0.4)
(0.3)

1 h, 1 h after extubation; 6 h, 6 h after extubation; 18 h, 18 h after extubation; 30 h, 30 h after extubation; 42 h, 42 h after extubation. Mean, mean score. No significant
differences between groups in relative frequencies (x2 test, Fishers exact test). MRepeated measures ANOVA: between subjects effect, not significant; within subjects
effect for time, P < 0.05; and interaction (time M groups), not significant.

more side effects as compared to placebo or the other

tested drugs.

13

14

Acknowledgement
The work was supported by Bristol-Myers Squibb, Munich,
Germany, and Pfizer Pharma GmbH, Karlsruhe, Germany.

15
16

References
1

3
4

8
9

10

11

12

Brodner G, Mertes N, Buerkle H, et al. Acute pain management: analysis,

implications and consequences after prospective experience with 6349
surgical patients. Eur J Anaesthesiol 2000; 17:566575.
Irwin MG, Roulson CJ, Jones RD, et al. Peri-operative administration of
rectal diclofenac sodium. The effect on renal function in patients
undergoing minor orthopaedic surgery. Eur J Anaesthesiol 1995; 12:403
406.
Sharma S. Upper gastrointestinal bleeding after hip and knee arthroplasty.
Orthopedics 2006; 29:255257.
Krishna S, Hughes LF, Lin SY. Postoperative hemorrhage with nonsteroidal
anti-inflammatory drug use after tonsillectomy: a meta-analysis. Arch
Otolaryngol Head Neck Surg 2003; 129:10861089.
Joshi GP, Gertler R, Fricker R. Cardiovascular thromboembolic adverse
effects associated with cyclooxygenase-2 selective inhibitors and
nonselective antiinflammatory drugs. Anesth Analg 2007; 105:1793
1804; table.
OConnor JP, Capo JT, Tan V, et al. A comparison of the effects of ibuprofen
and rofecoxib on rabbit fibula osteotomy healing. Acta Orthop 2009;
80:597605.
Hedenmalm K, Spigset O. Agranulocytosis and other blood dyscrasias
associated with dipyrone (metamizole). Eur J Clin Pharmacol 2002;
58:265274.
Kehlet H, Werner MU. Role of paracetamol in the acute pain management.
Drugs 2003; 63 (Spec No 2):1522.
Hyllested M, Jones S, Pedersen JL, Kehlet H. Comparative effect of
paracetamol, NSAIDs or their combination in postoperative pain
management: a qualitative review. Br J Anaesth 2002; 88:199
214.
Skoglund LA. Good evidence for effectiveness of analgesics most
commonly prescribed by UK dentists. What is the relative efficacy of singledose oral analgesics after third molar extraction? Evid Based Dent 2005;
6:66.
Kampe S, Warm M, Landwehr S, et al. Clinical equivalence of IV
paracetamol compared to IV dipyrone for postoperative analgesia after
surgery for breast cancer. Curr Med Res Opin 2006; 22:1949
1954.
Grundmann U, Wornle C, Biedler A, et al. The efficacy of the nonopioid
analgesics parecoxib, paracetamol and metamizol for postoperative pain
relief after lumbar microdiscectomy. Anesth Analg 2006; 103:217222;
table.

17

18

19
20
21

22
23

24

25

26

27

28

Girard P, Pansart Y, Coppe MC, et al. Modulation of paracetamol and

nefopam antinociception by serotonin 5-HT(3) receptor antagonists in
mice. Pharmacology 2009; 83:243246.
Benhamou D, Berti M, Brodner G, et al. Postoperative Analgesic THerapy
Observational Survey (PATHOS): a practice pattern study in 7 central/
southern European countries. Pain 2008; 136:134141.
White PF, Kehlet H. Improving postoperative pain management: what are
the unresolved issues? Anesthesiology 2010; 112:220225.
Elia N, Lysakowski C, Tramer MR. Does multimodal analgesia with
acetaminophen, nonsteroidal antiinflammatory drugs, or selective
cyclooxygenase-2 inhibitors and patient-controlled analgesia morphine
offer advantages over morphine alone? Meta-analyses of randomized trials.
Anesthesiology 2005; 103:12961304.
Brodner G, Pogatzki E, Wempe H, Van Aken H. Patient-controlled
postoperative epidural analgesia. Prospective study of 1799 patients.
Anaesthesist 1997; 46 (Suppl 3):S165S171.
Brodner G, Van Aken H, Hertle L, et al. Multimodal perioperative
management combining thoracic epidural analgesia, forced mobilization,
and oral nutrition reduces hormonal and metabolic stress and improves
convalescence after major urologic surgery. Anesth Analg 2001;
92:15941600.
Cohen J. Statistical power analysis for the behavioral science. Hilsdale:
Erlenbaum; 1988.
Schug SA, Manopas A. Update on the role of nonopioids for postoperative
pain treatment. Best Pract Res Clin Anaesthesiol 2007; 21:1530.
Munsterhjelm E, Munsterhjelm NM, Niemi TT, et al. Dose-dependent
inhibition of platelet function by acetaminophen in healthy volunteers.
Anesthesiology 2005; 103:712717.
Myles PS, Power I. Clinical update: postoperative analgesia. Lancet 2007;
369:810812.
Scott PA, Kingsley GH, Smith CM, et al. Nonsteroidal anti-inflammatory
drugs and myocardial infarctions: comparative systematic review of
evidence from observational studies and randomised controlled trials. Ann
Rheum Dis 2007; 66:12961304.
Sener M, Yilmazer C, Yilmaz I, et al. Efficacy of lornoxicam for acute
postoperative pain relief after septoplasty: a comparison with diclofenac,
ketoprofen, and dipyrone. J Clin Anesth 2008; 20:103108.
Rugyte D, Kokki H. Intravenous ketoprofen as an adjunct to patientcontrolled analgesia morphine in adolescents with thoracic surgery: a
placebo controlled double-blinded study. Eur J Pain 2007; 11:694699.
Sinatra RS, Jahr JS, Reynolds LW, et al. Efficacy and safety of single and
repeated administration of 1 gram intravenous acetaminophen injection
(paracetamol) for pain management after major orthopedic surgery.
Anesthesiology 2005; 102:822831.
Zhou TJ, Tang J, White PF. Propacetamol versus ketorolac for treatment of
acute postoperative pain after total hip or knee replacement. Anesth Analg
2001; 92:15691575.
Hynes D, McCarroll M, Hiesse-Provost O. Analgesic efficacy of parenteral
paracetamol (propacetamol) and diclofenac in postoperative orthopaedic
pain. Acta Anaesthesiol Scand 2006; 50:374381.

European Journal of Anaesthesiology 2011, Vol 28 No 2

Copyright European Society of Anaesthesiology. Unauthorized reproduction of this article is prohibited.