British Journal of Anaesthesia 101 (4): 535–41 (2008)

doi:10.1093/bja/aen215 Advance Access publication July 23, 2008

PAIN
Pregabalin and dexamethasone for postoperative pain control:
a randomized controlled study in hip arthroplasty
O. Mathiesen1*, L. S. Jacobsen2, H. E. Holm2, S. Randall2, L. Adamiec-Malmstroem2,
B. K. Graungaard2, P. E. Holst2, K. L. Hilsted1 and J. B. Dahl1
1
Department of Anaesthesia, Copenhagen University Hospital, Ndr. Ringvej, DK-2600 Glostrup, Denmark.
2
Department of Anaesthesia, Hoersholm Hospital, Hoersholm, Denmark
*Corresponding author. E-mail: olemat@dadlnet.dk
Background. Optimal pain treatment with minimal side-effects is essential for early mobility
and recovery in patients undergoing total hip arthroplasty. We investigated the analgesic effect
of pregabalin and dexamethasone in this surgical procedure.
Methods. One hundred and twenty patients were randomly allocated to either Group A
( placebo), Group B ( pregabalin 300 mg), or Group C ( pregabalin 300 mgþdexamethasone
8 mg). The medication and acetaminophen 1 g were given before operation. Spinal anaesthesia

Downloaded from bja.oxfordjournals.org by guest on May 10, 2011

was performed. Postoperative pain treatment was with acetaminophen 1 g three times daily
and patient-controlled i.v morphine, 2.5 mg bolus. Nausea was treated with ondansetron.
Morphine consumption, pain intensity at rest and during mobilization, nausea and vomiting,
sedation, dizziness, and consumption of ondansetron were recorded 2, 4, and 24 h after oper-
ation. P,0.05 was considered statistically significant.
Results. Twenty-four hour morphine consumption was significantly reduced in Groups B
[mean (SD) 24 (14) mg] and C [25 (19) mg] compared with Group A [47 (28) mg]. Vomiting
was reduced in Group C compared with Group B (P¼0.03). Sedation was significantly
increased in Group B compared with the other groups.
Conclusions. Pregabalin resulted in a 50% reduction in 24 h postoperative morphine require-
ments. This was not associated with a reduced incidence of nausea or vomiting. Pregabalin
resulted in increased levels of sedation. Combining pregabalin and dexamethasone provided no
additional effects on pain or opioid requirements.
Br J Anaesth 2008; 101: 535–41
Keywords: analgesia, postoperative; pain, postoperative
Accepted for publication: June 16, 2008

Total hip arthroplasty is associated with postoperative pain
of moderate intensity in the early postoperative period.1
Patients are often elderly and may have significant comor-
bid conditions; therefore, opioid-related side-effects such
as nausea and sedation are especially undesirable in this
population. Consequently, optimal pain treatment with
minimal side-effects is essential to allow early mobility,
optimal functional recovery, and to reduce postoperative
morbidity and mortality.2
It has been suggested that postoperative pain treatment
should be specific in relation to the surgical procedure2
and that a combination of different, preferably non-opioid,
analgesics should be administered in order to provide addi-
tive or synergistic effects together with reduced, opioid-
related side-effects.2 – 4
Both gabapentinoids5 – 11 and glucocorticoids12 – 16 have
demonstrated opioid-sparing effects in a number of clinical
studies of postoperative pain, but procedure-specific data
in relation to total hip arthroplasty of these medications, or
their combination, are not available.2 Consequently, for
our study, we have chosen a dose of pregabalin that
demonstrated efficacy in the first published clinical study
of pregabalin in acute pain,9 and a dose of dexamethasone
that demonstrated opioid-sparing and pain-relieving effects
in a study of laparoscopic cholecystectomy.15 The hypo-
thesis of our study was that ‘protective’5 administration of
pregabalin would reduce opioid requirements, pain, and
side-effects, and that the addition of dexamethasone would
further improve analgesia and reduce side-effects after hip
surgery. We expected that this combination of drugs

# The Board of Management and Trustees of the British Journal of Anaesthesia 2008. All rights reserved. For Permissions, please e-mail: journals.permissions@oxfordjournals.org
 Mathiesen et al.
administered before surgery would demonstrate additive
and prolonged postoperative analgesia.
The aim of this prospective, double-blind, randomized,
placebo-controlled study was therefore to investigate the
effect of pregabalin, and of a combination of pregabalin
and dexamethasone, on morphine consumption ( primary
end point), pain scores, and side-effects (secondary end
points) in patients undergoing hip joint alloplastic surgery.
Methods
The study was carried out at Hoersholm Hospital, the
Capital Region of Denmark. It was conducted in compli-
ance with guidelines for Good Clinical Practice (GCP) and
was monitored by Copenhagen University Hospital GCP
Unit. The design and the description adhere to the
Consolidated Standards of Reporting Clinical Trials state-
ment (CONSORT).17 Approval was obtained from the
Danish Medicines Agency, the local Regional Ethics
Committee, and The Danish Data Protection Agency.
The study was registered at www.ClinicalTrials.gov:
NCT00235261.
Patients undergoing primary alloplastic hip joint replace-
ment surgery, aged 55– 75 yr, BMI between 18 and 35,
ASA I– III, and having a planned spinal anaesthesia for
the procedure, were eligible for the study. Patients were
not admitted to the study if any of the following criteria
were present: (i) inability to cooperate, (ii) allergy to any
drugs in the study, (iii) treatment with antacids or anti-
depressants, (iv) a history of diabetes or epilepsy, (v)
known impaired kidney function, (vi) alcohol, drug abuse,
or both, (vii) a daily intake of analgesics, except for non-
steroidal anti-inflammatory drugs, COX-2 inhibitors, or
acetaminophen, and (viii) treatment with systemic gluco-
corticoids within 4 weeks before surgery.
When referred to the Department of Orthopaedics,
patients received written information regarding the trial
and were later enrolled in the study during the preoperative
anaesthetic evaluation. Signed and dated informed consent
was obtained from all patients.
All patients received acetaminophen and spinal anaes-
thesia. Oral acetaminophen 1 g was given as premedica-
tion 1 h before anaesthesia. Spinal anaesthesia was
performed at the L3/4 (alternatively L2/3 or L4/5) using
3 ml plain bupivacaine 5 mg ml21. Hypotension was treated
with isotonic sodium chloride, Voluven (hydroxyethyl
starch, HES, 130/0.4), i.v. ephedrine 5 – 10 mg, or both, at
the discretion of the anaesthetist. The same five orthopae-
dic specialists performed all surgical procedures using the
lateral approach for the arthroplasty.
Postoperative pain treatment consisted of oral acetamino-
phen 1 g every 8 h, initiated 4 h after operation, and patient-
controlled i.v. (PCA) morphine (CADD-Legacy PCA model
6300, Astra Tech, Taastrup, Denmark) 2.5 mg bolus, 10
min lockout time. Nausea of at least moderate level was
536
treated with i.v. ondansetron, 4 mg starting dose and 1 mg
supplemental doses. No other analgesic, anti-emetic, or
sedative drugs were used during the 24 h study period.
The study was randomized, double-blind, and placebo-
controlled. Study medication was prepared by the hospital
pharmacy into identical capsules of either 300 mg prega-
balin or placebo. A separate package containing either
dexamethasone or isotonic sodium chloride was opened
and prepared into a neutral syringe by a nurse, who was
not a part of the study or any handling of the patient. All
medications were given to the patient by one of the inves-
tigators. Study medication was marked with the name of
the project, the investigator’s name, and consecutive
numbers according to a computer-generated block ran-
domization schedule prepared by the hospital pharmacy.
Each block contained nine numbers and the patients were
assigned consecutively to their group according to their
number. The investigators did all assessments. No person
was aware of group assignment until all patients had been
included and assessments were completed.
Patients were randomly allocated to one of the following
three treatment groups: Group A: placeboþplacebo;
Downloaded from bja.oxfordjournals.org by guest on May 10, 2011
Group B: pregabalin (Lyricaw, Pfizer)þplacebo; and Group
C: pregabalinþdexamethasone (Fortecortinw, Merck). One
hour before anaesthesia and according to their groups,
patients received pregabalin 300 mg or placebo orally.
Likewise, before the induction of anaesthesia, i.v. dexa-
methasone 8 mg or placebo was given.
The primary outcome measure was patient-controlled
morphine consumption from 0 to 4 and 0 to 24 h after
operation. Secondary outcome measures were postopera-
tive pain score at rest and during mobilization, and side-
effects: nausea, vomiting, sedation and dizziness.
Before operation, all patients were instructed in the use of
the visual analogue scale (VAS) (0 mm, no pain; 100 mm,
worst pain imaginable) and patient-controlled analgesia.
Pain scores (before operation) at rest and during mobilization
(from sitting to standing position) were assessed.
Patients were visited by the investigators 2, 4, and 24 h
after operation. At each visit, outcomes were measured in
the following order: morphine consumption, VAS pain
score at rest and during mobilization, nausea, number of
vomits, sedation, and dizziness.
Total PCA morphine consumption from 0 to 2, 0 to 4,
and 0 to 24 h after operation were recorded. VAS pain
scores were assessed by the patients at 2, 4, and 24 h after
operation, both supine and during mobilization from the
supine to the sitting position. Levels of nausea, sedation,
and dizziness, at the time of visit by the investigator, were
assessed by the patient using a four-point verbal scale
(none, mild, moderate, and severe). Number of vomits in
the postoperative periods 0 –2, 2 – 4, and 4 – 24 h was
assessed by the patients. Finally, total 24 h consumption
of ondansetron was recorded.
On the basis of morphine usage in a retrospective
sample of 30 patients in our department, who had total hip
 Pain relief after hip surgery

alloplastic surgery, the anticipated morphine requirement

Assessed for eligibility
was 30 (15) mg in 24 h. We considered a 40% reduction (n =514)
(12 mg) to be clinically relevant. With a type I error (a) of
5% and a power (12b) of 80%, sample size calculations
showed that 31 patients in each group were required. Excluded
(n =388)
Data are presented as median (IQR) or mean (SD) as Lack of time (n =115); did not want to participate (n =39);
appropriate. P,0.05 was considered statistically signifi- diabetes (n = 32); medication (antidepressants, antacids,
analgesics, antiepileptics, glucocorticoids) (n =109);
cant. To assess for normality, the Kolmogorov – Smirnov BMI (n =15); inability to cooperate (n =14);
(K– S) test was performed on the data set. Assumption of cave drugs used in study (n =4); planned general
normality was rejected for most data, and consequently anaesthesia (n =52); alcohol abuse (n =3); impaired
kidney function (n =3); participant in another study (n =1);
data were compared using the non-parametric Kruskal– already participated in this study (n =1)
Wallis (K– W) test for independent samples. Significant
Kruskal – Wallis values were Bonferroni corrected with a
factor 3 for the three different time points investigated. If Randomized
(n =126)
the Bonferroni corrected Kruskal – Wallis test value was
significant, groupwise comparisons were performed at
each time point using Mann – Whitney rank sum test for
unpaired data. Significant Mann – Whitney values were Group C
Group A Group B
(n =42)
again Bonferroni corrected with a factor 3 for the number (n =42) (n =42)
(pregabalin
(placebo) (pregabalin)
of groups. + dexamethasone)

Categorical data were analysed with x2 test. For seda-

Excluded
tion and dizziness, the arithmetic ‘mean’ scores for each Insufficient spinal
Excluded

Downloaded from bja.oxfordjournals.org by guest on May 10, 2011

Local anaesthetic in
patient was calculated and compared by attributing anaesthesia (n =2);
wound (n =1); wanted
Excluded
allergic reaction (n =1); None
numerical values to the scores from each patient: none, 0; medicine from another
to drop out because
slight, 1; moderate, 2; and severe, 3. of nausea (n =1)
study (n =1)
Calculations were performed using SPSS 13.0 for
Windows (SPSS, Chicago, IL, USA). Analysed Analysed Analysed
(n =38) (n =40) (n =42)

Fig 1 Flow diagram of patient distribution.

Results
nausea and one patient having severe nausea in Group A
From November 28, 2005 to June 18, 2007, we considered
(Table 2). The incidence of vomiting was generally low
514 consecutive patients aged 55– 75 yr for inclusion in
with 32 (Group A), 30 (Group B), and 40 (Group C)
the study. One hundred and twenty-six patients were
patients not experiencing vomiting at all. Total 24 h
included and randomly assigned to their treatment group.
number of vomits and number of patients vomiting were
However, six patients were later excluded from the study,
not significantly different between Group A and Group B.
resulting in data from 120 patients in the final analyses
In Group C, total 24 h number of vomits was significantly
(Fig. 1).
lower than in Group B (P¼0.03), but not different from
There were no significant differences between the
that of Group A. Likewise, total 24 h number of patients
groups for patient characteristics and peroperative data,
vomiting was significantly lower in Group C vs Group B
except for the weight (Group B vs C; P¼0.01) (Table 1).
(P¼0.03), but not different from that of Group A
The median numbers of days staying in hospital were 4, 5,
(Table 2). For the consumption of ondansetron, there were
and 4 for Groups A, B, and C, respectively (NS).
no significant differences between the groups (Table 2).
For the first 2 and 4 h after operation, there were no sig-
The mean 24 h sedation scores were significantly
nificant differences between the groups in morphine con-
higher in Group B compared with Group A (P,0.003) and
sumption (Fig. 2). The total 24 h morphine consumption
Group C (P¼0.02). Sedation was most prominent at 2 and
was significantly reduced in Groups B [24 (14) mg] and
4 h, whereas, at 24 h, there were no significant differences
C [25 (19) mg] compared with Group A [47 (28) mg]
between the groups (Table 3). For mean 24 h dizziness
(P,0.003). Between Group B and Group C, there was no
scores, there were no significant differences between treat-
significant difference (P¼0.90) (Fig. 2).
ment groups (P¼0.24).
For VAS pain scores at rest, or during movement, there
were no significant differences between the groups at any
time points (Figs 3 and 4). For nausea scores, there were
no significant differences between treatment groups at any Discussion
time points. Nausea was of mild nature in most patients, Preoperative administration of pregabalin 300 mg resulted
with two patients in each group experiencing moderate in a nearly 50% reduction in 24 h postoperative morphine

537
 Mathiesen et al.

Table 1 Patient characteristics. Data are median (range) or mean (SD)

Group A (placebo) Group B (pregabalin) Group C

(pregabalin1 dexamethasone)

Number of patients (n) 38 40 42

Age 66 (63 – 71) 67 (62– 71) 68 (64– 71)
Height (cm) 171 (9) 170 (9) 170 (8)
Weight (kg) 78 (14) 73 (13) 81 (13)
Gender (male/female) 18/20 14/26 22/20
Preoperative VAS pain at rest (mm) 2 (0 –16) 5 (0– 21) 7 (2– 19)
Preoperative VAS pain at mobilization (mm) 21 (8 –30) 15 (5– 27) 21 (14– 31)
Duration of surgery (min) 60 (52 – 67) 52 (46– 60) 60 (50– 71)
Bleeding (ml) 425 (300 – 600) 375 (263 –500) 400 (300 –600)
Isotonic sodium chloride (ml) 825 (600 – 1000) 1000 (800 –1000) 1000 (800 –1000)
Voluven (ml) 250 (0 –500) 275 (0– 500) 425 (0– 500)
Type of alloplastic
No cement 36 39 40
Total cement 0 0 1
Hybrid 2 1 1

requirements in patients undergoing total hip arthroplasty
under spinal anaesthesia. This opioid-sparing effect was
not associated with a reduced incidence of nausea or
vomiting. Pregabalin 300 mg resulted in increased levels
of sedation. The combination of pregabalin 300 mg and
Glucocorticoids have both anti-inflammatory and
anti-emetic properties and have demonstrated prolonged
postoperative analgesic effects in several procedures.12 13
Dexamethasone has been shown to reduce postoperative
analgesic requirements in both dental (8 and 16 mg),14
laparoscopic (8 mg),15 and breast (8 and 16 mg) surgery.13 16

Downloaded from bja.oxfordjournals.org by guest on May 10, 2011

i.v. dexamethasone 8 mg reduced sedation and vomiting
compared with pregabalin 300 mg alone, but provided no
additional effects on pain or opioid requirements.
Gabapentin has demonstrated substantial opioid-sparing
and pain-relieving abilities in a large number of clinical
studies of postoperative pain.5 – 8 In addition, opioid-related
side-effects, such as nausea and vomiting, may be
reduced.5 – 8 In the first trial investigating the postoperative
analgesic effect of pregabalin, a dose of 300 mg demon-
strated significant pain-relieving properties for patients in
dental surgery.9 In patients undergoing spinal fusion
surgery,10 pregabalin 150 mg 2 reduced morphine con-
sumption compared with placebo, and in laparoscopic hys-
terectomy,11 pregabalin 300 mg 2 reduced 24 h
The main results of our study could not confirm our
hypothesis that pregabalin and dexamethasone would
demonstrate prolonged and additive analgesic effects. We
observed reduced morphine requirements in the groups
receiving pregabalin, but no additional opioid-sparing or
clinically significant prolonged pain-relieving effects of
adding dexamethasone to pregabalin. We only have data on
accumulated morphine requirements at 2, 4, and 24 h after
operation, and our results cannot document if the opioid-
Group placebo
60 Group pregabalin
Group pregabalin + dexamethasone
VAS pain score (mm) at rest (mean and SD)

postoperative oxycodone consumption. However, pregaba-

lin has not been evaluated in hip replacement surgery. 50

80
Group placebo
Cumulative morphine (mg) consumption

40
Group pregabalin
Group pregabalin
60 + dexamethasone
30
(mean and SD)

*
40 *
20

20
10

0 0
0–2 h 0–4 h 0–24 h 2h 4h 24 h
Time periods Hours after operation

Fig 2 Consumption of morphine in 0– 24 h after operation in placebo, Fig 3 Pain score (VAS) at rest at 2, 4, and 24 h after operation in
pregabalin, and pregabalinþdexamethasone groups. *P,0.003 compared placebo, pregabalin, and pregabalinþdexamethasone groups. There were
with placebo. no differences between the groups.

538
 Pain relief after hip surgery

VAS pain score (mm) during mobilization (mean and SD) In spite of the relatively large reduction in morphine
60 Group placebo consumption, we did not observe any reduction of nausea
Group pregabalin
Group pregabalin + dexamethasone or vomiting with pregabalin. The incidence of nausea and
vomiting in this population was very low, and our study is
50 most likely underpowered to detect any difference. The
well-known anti-emetic effect of dexamethasone18 was
40
seen as a significant reduction in vomiting in patients
receiving the combination of pregabalin and dexametha-
sone compared with pregabalin alone in our study.
30 Sedation and dizziness are well-known side-effects of
gabapentinoids,7 and we observed increased sedation in
patients receiving pregabalin alone in the early postopera-
20 tive period. The clinical relevance is arguable as most
patients were only slightly sedated at this stage and all
patients were able to follow the postoperative routine of
10
care and mobilization. When adding dexamethasone to
pregabalin (Group C), sedation diminished and we
0 observed no significant differences in mean sedation score
2h 4h 24 h between Group A and Group C. In a previous study,
Hours after operation dexamethasone 8 mg significantly reduced postoperative
Fig 4 Pain score (VAS) on movement at 2, 4, and 24 h after operation in fatigue scores after laparoscopic cholecystectomy.15
placebo, pregabalin, and pregabalinþdexamethasone groups. There were Consequently, dexamethasone may have important ben-

Downloaded from bja.oxfordjournals.org by guest on May 10, 2011

no differences between the groups. eficial effects other than analgesia in multimodal
regimens.
Table 2 Postoperative nausea, vomiting, and consumption of ondansetron The combination of pregabalin and dexamethasone had
Group A Group B Group C P-value no effect on pain or opioid requirements compared with
n538 n540 n542 pregabalin alone. One reason for this finding could be that
(placebo) (pregabalin) (pregabalin1
dexamethasone)
the dose of dexamethasone was too low. We used 8 mg of
dexamethasone, but a somewhat higher dose of 0.2– 0.4
Nausea mg kg21 dexamethasone i.v. is recommended by some
2h
authors.19 Another reason could be the low pain score at
None 37 39 40 0.82
Slight 1 1 2 most time points in all groups (,30 mm on the VAS),
Moderate 0 0 0 since adequate sensitivity in trials of analgesics for acute
Severe 0 0 0
pain may only be achieved when patients are experiencing
4h
None 37 35 42 0.08 at least moderate pain.20 21
Slight 0 5 0 In a recent procedure-specific systematic review, recom-
Moderate 1 0 0
mendations were provided for optimal pain treatment after
Severe 0 0 0
24 h total hip arthroplasty.2 These recommendations included
None 32 35 36 0.90 acetaminophen and in addition, a number of medications,
Slight 3 3 4
including gabapentinoids, were suggested to have potential
Moderate 2 2 2
Severe 1 0 0 usefulness, but procedure-specific data were not yet avail-
Total no. of 18 27 3 0.34 (A vs B) able.2 Our study presents procedure-specific data for prega-
vomits (0– 24 h) 0.09 (A vs C)
balin in hip arthroplasty. However, in spite of the opioid-
0.03 (B vs C)
No. of patients 6 10 2 0.32 (A vs B) sparing effect demonstrated, general recommendations for
vomiting (0 –24 h) 0.10 (A vs C) the use of pregabalin in hip alloplastic surgery cannot be
0.03 (B vs C)
made based on this single study. The optimal dosing
Ondansetron (mg) 2.9 (4.3) 2.1 (2.9) 1.0 (1.9) 0.11
(mean and SD) regimen for pregabalin is unsolved, and the potential clini-
Number of 15 15 10 0.26 cal benefit from the opioid-sparing was not demonstrated.
patients requiring
The mechanisms of action of N-methyl-D-aspartic acid
ondansetron
(NMDA) antagonists and gabapentinoids differ from that
of ‘traditional’ anti-nociceptive drugs. Whereas the latter
sparing effect of pregabalin was prolonged beyond the reduce the afferent input from both intact and traumatized
expected clinical duration of action, or if it took place during tissue, NMDA antagonists and gabapentinoids have no
the first few postoperative hours. Future studies with more effect on nociception per se, but reduce the hyperexcitabil-
frequent assessments are needed to answer this question. ity of dorsal horn neurones induced by tissue damage.

539
 Mathiesen et al.

Table 3 Postoperative sedation

Group A n538 Group B n540 Group C n542 P-value

(placebo) (pregabalin) (pregabalin1dexamethasone)

Sedation
2h
None 33 15 20 ,0.003 (A vs B)
Slight 4 14 14 ,0.003 (A vs C)
Moderate 0 6 6 0.26 (B vs C)
Severe 1 5 2
4h
None 31 11 26 ,0.003 (A vs B)
Slight 4 19 15 0.10 (A vs C)
Moderate 2 7 1 ,0.003 (B vs C)
Severe 1 3 0
24 h
None 26 24 34 0.49 (A vs B)
Slight 9 13 8 0.15 (A vs C)
Moderate 2 2 0 0.08 (B vs C)
Severe 1 1 0
Mean sedation (2– 24 h) 0.30 0.90 0.45 ,0.003 (A vs B)
(none, 0; slight, 1; moderate, 2; 0.09 (A vs C)
severe, 3) 0.02 (B vs C)

This effect seems equally effective irrespective of the Funding

timing of administration, that being if the drugs are admi-

Downloaded from bja.oxfordjournals.org by guest on May 10, 2011

nistered before or after the injury.22
The term ‘anti-hyperalgesic’ has been suggested for
NMDA antagonists, gabapentinoids, and other medications
with similar mechanisms of action. Accordingly, the term
‘protective analgesia’ has been suggested for preoperative
administration of anti-hyperalgesics, since these medi-
cations block the sensitization process in dorsal horn neur-
ones.5 ‘Protective’ was suggested in order to differentiate
this concept from ‘pre-emptive’ analgesia. Although both
concepts share the same objective, reduction of
injury-induced central sensitization, the mechanisms of
action are different. Whereas ‘pre-emptive’ analgesia aims
to hinder the afferent, nociceptive input from ever reaching
the dorsal horn neurones, anti-hyperalgesics are thought to
attenuate the sensitization process itself. In this study, we
have investigated the ‘protective’ analgesic effect of the
anti-hyperalgesic pregabalin alone and in combination
with dexamethasone for patients in hip alloplastic surgery.
In conclusion, we have demonstrated that a single, pre-
operative dose of pregabalin 300 mg resulted in a nearly
50% reduction in 24 h postoperative morphine requirements
in patients undergoing hip surgery. This opioid-sparing was
not, however, associated with a reduced incidence of
nausea or vomiting. In contrast, pregabalin resulted in
increased levels of sedation. The combination of pregabalin
and dexamethasone 8 mg reduced vomiting and sedation
compared with pregabalin alone, but no additional effects
on pain or opioid requirements could be demonstrated.
Acknowledgements
We thank Copenhagen University Hospital GCP Unit, Gentofte Hospital,
Denmark, for monitoring the study and Astra Tech, Taastrup, Denmark,
for supplying CADD-Legacy PCA devices.
J.B.D. has received an unrestricted research grant from
Pfizer, Denmark.
References
1 Nikolajsen L, Brandsborg B, Lucht U, Jensen TS, Kehlet H.
Chronic pain following total hip arthroplasty: a nationwide ques-
tionnaire study. Acta Anaesthesiol Scand 2006; 50: 495 – 500
2 Fischer HB, Simanski CJ. A procedure-specific systematic review
and consensus recommendations for analgesia after total hip
replacement. Anaesthesia 2005; 60: 1189 – 202
3 Kehlet H, Dahl JB. The value of ‘multimodal’ or ‘balanced analge-
sia’ in postoperative pain treatment. Anesth Analg 1993; 77:
1048– 56
4 Kehlet H, Dahl JB. Anaesthesia, surgery, and challenges in post-
operative recovery. Lancet 2003; 362: 1921 –8
5 Dahl JB, Mathiesen O, Moiniche S. ‘Protective premedication’: an
option with gabapentin and related drugs? A review of gabapentin
and pregabalin in the treatment of post-operative pain. Acta
Anaesthesiol Scand 2004; 48: 1130 – 6
6 Mathiesen O, Moiniche S, Dahl JB. Gabapentin and postoperative
pain: a qualitative and quantitative systematic review, with focus
on procedure. BMC Anesthesiol 2007; 7: 6
7 Tiippana EM, Hamunen K, Kontinen VK, Kalso E. Do surgical
patients benefit from perioperative gabapentin/pregabalin? A sys-
tematic review of efficacy and safety. Anesth Analg 2007; 104:
1545– 56
8 Ho KY, Gan TJ, Habib AS. Gabapentin and postoperative pain—a
systematic review of randomized controlled trials. Pain 2006;
126: 91 – 101
9 Hill CM, Balkenohl M, Thomas DW, Walker R, Mathe H,
Murray G. Pregabalin in patients with postoperative dental pain.
Eur J Pain 2001; 5: 119 – 24
10 Reuben SS, Buvanendran A, Kroin JS, Raghunathan K. The analge-
sic efficacy of celecoxib, pregabalin, and their combination for
spinal fusion surgery. Anesth Analg 2006; 103: 1271 – 7
11 Jokela R, Ahonen J, Tallgren M, Haanpaa M, Korttila K. A ran-
domized controlled trial of perioperative administration of

540
 Pain relief after hip surgery
pregabalin for pain after laparoscopic hysterectomy. Pain 2008;
134: 106 – 12
12 Kehlet H. Glucocorticoids for peri-operative analgesia: how far
are we from general recommendations? Acta Anaesthesiol Scand
2007; 51: 1133– 5
13 Hval K, Thagaard KS, Schlichting E, Raeder J. The prolonged
postoperative analgesic effect when dexamethasone is added to a
nonsteroidal antiinflammatory drug (rofecoxib) before breast
surgery. Anesth Analg 2007; 105: 481 – 6
14 Numazaki M, Fujii Y. Reduction of postoperative emetic
episodes and analgesic requirements with dexamethasone in
patients scheduled for dental surgery. J Clin Anesth 2005; 17: 182– 6
15 Bisgaard T, Klarskov B, Kehlet H, Rosenberg J. Preoperative dexa-
methasone improves surgical outcome after laparoscopic chole-
cystectomy: a randomized double-blind placebo-controlled trial.
Ann Surg 2003; 238: 651 – 60
16 Fujii Y, Nakayama M. Reduction of postoperative nausea and
vomiting and analgesic requirement with dexamethasone in
women undergoing general anesthesia for mastectomy. Breast J
2007; 13: 564 –7
541
17 Moher D, Schulz KF, Altman DG. The CONSORT statement:
revised recommendations for improving the quality of reports of
parallel-group randomised trials. Lancet 2001; 357: 1191– 4
18 Henzi I, Walder B, Tramer MR. Dexamethasone for the preven-
tion of postoperative nausea and vomiting: a quantitative systema-
tic review. Anesth Analg 2000; 90: 186 – 94
19 Romundstad L, Stubhaug A. Glucocorticoids for acute and per-
sistent postoperative neuropathic pain: what is the evidence?
Anesthesiology 2007; 107: 371 – 3
20 Bjune K, Stubhaug A, Dodgson MS, Breivik H. Additive analgesic
effect of codeine and paracetamol can be detected in strong,
but not moderate, pain after Caesarean section. Baseline pain-
intensity is a determinant of assay-sensitivity in a postoperative
analgesic trial. Acta Anaesthesiol Scand 1996; 40: 399 – 407
21 Collins SL, Moore RA, McQuay HJ. The visual analogue pain
scale: what is moderate pain in millimetres? Pain 1997; 72: 95 – 7
22 Woolf CJ, Thompson SW. The induction and maintenance of
central sensitization is dependent on N-methyl-D-aspartic acid
receptor activation; implications for the treatment of post-injury
pain hypersensitivity states. Pain 1991; 44: 293– 9
Downloaded from bja.oxfordjournals.org by guest on May 10, 2011