Hong Kong Journal of Emergency Medicine

A randomised control trial comparing the efficacy of tramadol and

paracetamol against ketorolac and paracetamol in the management of
musculoskeletal pain in the emergency department

HKH Lee , SM Ting , FL Lau

Background: This study aimed to compare the efficacy, acceptance and side effects of intramuscular tramadol
and ketorolac in combination with oral paracetamol in the emergency setting. Materials and methods: This
was a randomised, double blind controlled trial. Patients aged 18 years or above with moderate to severe
musculoskeletal pain were recruited. Patients with known allergy, currently on psychiatric medication, with
alcohol or opioid dependence, during pregnancy and with major systemic illness were excluded. Tramadol
100 mg or ketorolac 30 mg was given intramuscularly together with paracetamol 500 mg per oral. They
were then bed rested in the observation ward for one hour. Visual analogue scale (VAS), satisfaction score,
vital signs and side effects were recorded. Results: Altogether, 78 (M: 43, F: 35) subjects were recruited from
June to September 2005, with equal number in each arm. The mean age was 39.9 for the tramadol group
and 43.9 for the ketorolac group. Most of them suffered from back pain (66.7%). There was a significant
difference in VAS improvement between the two groups (0.88; P=0.01). However, there were no significant
differences in patients' satisfaction score and admission rate. The incidence of side effects was similar between
the two groups (tramadol 19, ketorolac 17), mainly dry mouth, and none were major. The tramadol group
had more nausea. Conclusion: The analgesic effect of the tramadol and paracetamol combination is as effective
as the ketorolac and paracetamol combination. Tramadol is well tolerated and relatively safe. It is also cheaper
than ketorolac. Hence, we recommend tramadol and paracetamol combination for acute moderate to severe
musculoskeletal pain in the emergency setting. (Hong Kong j.emerg.med. 2008;15:5-11)

18

500 mg 100 mg 30 mg
2005 6 9
78 43 35 39.9
43.9 66.7% 0.88; P = 0.01
19

Correspondence to:
Lee Ka Hing, Herman, MBChB, MRCSEd, FHKAM(Emergency Medicine)
United Christian Hospital, Accident & Emergency Department,
130 Hip Wo Street, Kwun Tong, Kowloon, Hong Kong
Email: kahing74@yahoo.com

Ting Soo Moi, MRCP(London), FRCSEd, FHKAM(Emergency Medicine)

Lau Fei Lung, FRCP(Edin), FRCSEd, FHKAM(Emergency Medicine)
6 Hong Kong j. emerg. med. „ Vol. 15(1) „ Jan 2008
17
Keywords: Analgesia, ketorolac tromethamine, opioid analgesics
Background
Acute musculoskeletal painful conditions are common
problems presenting to the emergency department. In
moderate and severe pain, early analgesic treatment is
effective in reducing suffering, promoting early
discharge and early return to work.1-3
Ketorolac has been the major injectable analgesic used
in acute pain in our department for some time. In our
experience, it was effective. However, being a non-
steroidal anti-inflammatory drug (NSAID), we worry
about its important potential risks of hypersensitivity
reactions and renal, hepatic and gastrointestinal side
effects.
Tramadol was recently introduced to our department
as an injectable analgesic. It achieves analgesia by two
mechanisms: an opioid effect and an enhancement of
serotoninergic and adrenergic pathways.
This study aimed to compare the efficacy, acceptance
and side effects of intramuscular tramadol and
ketorolac in combination with oral paracetamol in the
emergency setting. This may affect our prescribing
preference and avoid the major adverse reactions of
the NSAIDs.
Materials and methods
This was a prospective randomised double-blinded
control study (Figure 1). Subjects were randomised by
block randomisation, with envelops drawn randomly
from preset equal number in the two groups. Patients
were blinded to the treatment they received. They were
asked to complete the questionnaire before they
received treatment. A clinically independent specialist
in the observation room would perform the post
injection assessment.
Patients of both sexes from 18 to 65 years of age, with
moderate to severe acute musculoskeletal pain
(irrespective of trauma history) and also acute arthritis
(including gout) presenting within 72 hours of onset
were recruited. Patients functionally affected by
other medical or surgical conditions, currently on
psychiatric medication, with alcohol or opioid
d e p e n d e n c e , d u r i n g p r e g n a n c y, w i t h k n ow n
hypersensitivity to paracetamol, NSAID or opioid,
with contraindications to NSAIDs or contraindications
to opioid/tramadol, patients with neurological deficit,
bone fracture or active malignancy were excluded.
Subjects included were randomised to receive
intramuscular tramadol 100 mg with oral paracetamol
500 mg or intramuscular ketorolac 30 mg with oral
paracetamol 500 mg accordingly.
The sample size (n=34 in each group) was calculated with
effect size 3.38.4 The calculated standardised effect size
was 0.8, with alpha 0.05 and power 0.90, we can look
up in a table and find out the required sample size.
The primary end-point of the study was pain control. It
was assessed by a self-completed questionnaire with a
visual analogue scale (VAS) before injection and 60
minutes after injection of either drug plus oral
paracetamol. The VAS we used was a 10-unit horizontal
straight line with each unit equal to 1.5 cm, whereas,
0 represented no pain and 10 represented maximal pain.
The secondary end-points were complications and the
satisfaction score measured from 0 to 100 points. The
post-treatment VAS, satisfaction score and side effects
Lee et al./Tramadol versus ketorolac 7

VAS = visual analogue scale; SE = side effect

Figure 1. Study flow chart.

were assessed and recorded by a clinically independent
Senior Medical Officer doing the observation room
duty. The blood pressure, pulse rate, oxygen saturation,
and rescue analgesics required after the first hour were
recorded.
All the statistical calculation employed the online
engine SISA (Simple Interactive Statistical Analysis).
Results

The difference in the VAS improvement between the
two groups and vital sign parameters were analysed by
the unpaired t-test. A p-value less than 0.05 was
considered as statistically significant. Side effects were
analysed with Fisher's exact test and Chi-square test.
Seventy-eight patients were recruited between 21 June
2005 and 21 September 2005, with equal number of
subjects in each group. The two groups were
comparable in sex, age and diagnosis. The subjects'
characteristics are shown in the Table 1. The majority

Table 1. Patient characteristics and presenting problems

Tramadol Ketorolac P-value
Number of subjects 39 39 1
Sex (M/F) 22/17 21/18 0.888
Mean age (range) 39.9 (18-57) 43.9 (18-65) 0.471
Back pain 28 (71.8%) 24 (61.5%) 0.249
Ankle & foot pain 4 (10.3%) 5 (12.8%) 0.807
Gout 1 (2.6%) 2 (5.1%) 0.380
Others 6 (15.4%) 8 (20.5%) 0.554
Initial visual analogue scale 7.39 6.72 0.353
8 Hong Kong j. emerg. med. „ Vol. 15(1) „ Jan 2008

(52) suffered from back pain (66.7%), with 28 in the
tramadol group and 24 in the ketorolac group, followed
by ankle and foot pain (11.5%). There was a
statistically significant difference in improvement of
VAS score between the two groups, with the tramadol
group being better. The mean pre-treatment VAS for
the tramadol group was 7.39 with a reduction of 3.12
(42.2%) after treatment, while that for the ketorolac
group was 6.72 with an improvement of 2.24 (33.3%)
after treatment. There was no significant difference for
the initial VAS between the two groups (Table 1). The
difference in VAS improvement for the two groups was
0.88 (P=0.01) (Table 2). There were no significant
differences in satisfaction score and admission rate.
Nineteen patients (48.7%) in the tramadol group and
17 patients (43.6%) in the ketorolac group experienced
side effects. The number of patients in the two groups
who experienced and reported side effects was not
statistically significant (Chi-square test, P=0.206). Dry
mouth was the commonest, with 13 patients (33.3%) in
the tramadol group and 12 patients (30.8%) in the
ketorolac group. No major life-threatening effect
occurred. However, significantly more patients
experienced nausea in the tramadol group than in the
ketorolac group (6 versus 1, P<0.05) (Table 3).
In both the tramadol and ketorolac groups, there was
a slight drop in the mean systolic and diastolic blood

Table 2. Analysis result of the difference in visual analogue scale (VAS) improvement, satisfaction score between the tramadol and
ketorolac groups and the change in physiological parameters (blood pressure, heart rate and capillary oxygen saturation) before and
60 minutes after the injection of the respective drug
Paired parameter (P1 vs P2) Absolute value (P1/P2) Mean difference (P1-P2) (95% CI) P-value
VAS improvement (tramadol vs ketorolac group) 3.12/2.24 0.88 (0.013 - 1.75) 0.01
Satisfaction score (tramadol vs ketorolac group) 66.6/58.7 7.9 (-2.94 - 18.7) 0.164
Tramadol group: systolic BP (baseline vs 60 min) 132/123 9 (1.24 - 16.8) 0.330
Tramadol group: diastolic BP (baseline vs 60 min) 83/77 6 (0.884 - 12.1) 0.337
Tramadol group: pulse rate (baseline vs 60 min) 80/68 12 (7.59 - 16.4) 0.040
Tramadol group: O 2 saturation (baseline vs 60 min) 99/98 1 (0.293 - 1.71) 0.0006
Ketorolac group: systolic BP (baseline vs 60 min) 139/125 14 (5.42 - 22.6) 0.459
Ketorolac group: diastolic BP (baseline vs 60 min) 82/78 4 (-1.87 - 9.47) 0.337
Ketorolac group: pulse rate (baseline vs 60 min) 83/67 16 (10.4 - 21.6) 0.007
Ketorolac group: O2 saturation (baseline vs 60 min) 98/98 0 (-0.577 - 5.77) 0.358

Table 3. Side effects reported by the two study groups at 60 minutes post drug administration
Symptom Tramadol Ketorolac P-value
Dry mouth 13 (33.3%) 12 (30.8%) 0.862*
Headache 0 2 (5.1%) 0.247**
Nausea 6 (15.4%) 1 (2.6%) 0.048**
Dyspnoea 1 (2.6%) 1 (2.6%) 0.385**
Dizziness 4 (10.3%) 3 (7.7%) 0.285**
Abdominal discomfort 1 (2.6%) 0 0.500**
Sweating 1 (2.6%) 1 (2.6%) 0.385**
Drowsiness 2 (5.1%) 3 (7.7%) 0.321**
Others Sleepy 1; tired 1 (5.1%) Sweet taste 1; dyspepsia 1 (5.1%) 0.385**
* Chi-square test; **Fisher's exact test
Lee et al./Tramadol versus ketorolac
pressure, which was not clinically or statistically
significant. The mean pulse rate dropped in both
groups, which were statistically but not clinically
significant for both groups. There was a statistically
significant drop in mean oxygen saturation (from 99%
to 98%) in the tramadol group, but was not clinical
significant.
Six patients required rescue drugs. Two in the tramadol
group and three in the ketorolac group required further
analgesics (P=0.321) and one in the tramadol group
required medication for dizziness (P=0.50). Two
patients in the tramadol group and one patient in the
ketorolac group were admitted to the orthopaedic
department due to severe persistent low back pain
(P=0.38).
Discussion
Acute musculoskeletal pain presents commonly to
emergency departments. The choice of injectable
analgesics has been narrow, namely non-steroidal anti-
inflammatory drug and sometimes, narcotics in selected
patients and conditions.
Ketorolac tromethamine is a member of non-steroidal
anti-inflammatory drugs, and has a rapid onset of
action of 30 minutes for the intramuscular route, with
peak effect at 1 to 2 hours. Like other NSAIDs, it has
the potential risks of hypersensitivity reactions,
nephotoxicity, hepatotoxicity and gastrointestinal
bleeding/ulceration after repeated use. NSAID given
intramuscularly may also be very painful and associated
with severe local complications. A study by Chung
showed that the use of injectable NSAID in local
accident and emergency department practice was
excessive.5 He suggested stringent justifications for its
intramuscular administration.
The recent introduction of tramadol in our department
pr ov i d e d a p r o m i s i n g a l t e r n a t i ve . Tr a m a d o l
hydrochloride is a synthetic analogue of codeine that
has low affinity for the mu-opiate receptors. It has been
claimed to be safer with few opioid side effects: notably,
respiratory depression, constipation and addiction
9
potential. It has a rapid onset of action by intramuscular
injection, with maximum effect within 15 to 30
minutes and duration of action lasting 3 to 6 hours.
Siu and Chung found that tramadol was a safe drug in
the emergency setting with few and insignificant side
effects.4 Its efficacy as compared with other analgesics
required further studies.
Two studies compared the analgesic efficacy of
ketorolac and tramadol. Both studies were conducted
by anaesthetists in patients undergoing abdominal
surgery. Ollé Fortuny et al showed that during the first
12 hours following surgery, a 100 mg dose of tramadol
provided more effective pain relief than 30 mg of
ketorolac, each administrated intravenously every
6 hours. However, intravenous tramadol was associated
with a higher incidence of postoperative vomiting
(38%).6 In Putland and McCluskey’s study, patients
who received tramadol had less pain in the recovery
room and at discharge from the day-surgery unit and
required less rescue analgesia with morphine than
patients who received ketorolac, but dry mouth was
significantly more common after the administration
of tramadol than ketorolac (60% versus 27%). 7
Therefore, we conducted the study to assess tramadol's
effectiveness and tolerability as an analgesic in
comparison with the currently widely used NSAID in
acute musculoskeletal pain in the emergency setting.
Rosenthal showed that combination therapy had
promising result. 8 Tramadol/paracetamol add-on
therapy effectively managed painful osteoarthritis flare
in the elderly subset and was generally well tolerated.
We chose combinations of paracetamol with the study
drugs to maximise the analgesic effects and minimise
side effects.
In our study, we found that the intramuscular tramadol
and oral paracetamol combination had statistically
significantly better analgesic effect than the ketorolac
and paracetamol combination at 60 minutes after
injection. However, we could not show any significant
clinical difference as measured by patients' satisfaction,
reduction in admission rate, and the use of rescue
drugs. This could be due to the short study period
and rapid follow up assessment after one single
10
treatment as compared to the other studies, which were
done in hospitalised patients. Also, the synergistic effect
of paracetamol could have blunted the significance.
Moreover, the sample size was calculated from the effect
size of the change in VAS, which might not be large
enough with the power to detect any significant
difference in the incidence of side effects and
satisfaction score.
Our study, however, showed that intramuscular
tramadol was safe to use. The commonest side effects
of tramadol are minor: namely nausea, dizziness,
drowsiness, sweating, vomiting and dry mouth. The
reported incidence was about 1.6 to 6.1%.4 Our study
showed the incidence of side effects of the tramadol
and paracetamol combination (48.7%) was higher than
the ketorolac and paracetamol combination (43.6%),
but it was not statistically significant. These minor
effects were unexpectedly high compared to the
reported incidence. However, only one patient needed
treatment, which might reflect individual subjective
variation and difference in reporting and recording
symptoms. The tramadol and paracetamol combination
had more nausea than the ketorolac and paracetamol
combination. There was no seizure or psychotic effects
like mania, auditory hallucination and serotonin
syndrome in our patients.
The cost per treatment of tramadol (100 mg per
ampoule) was HK$2.37 and the cost of ketorolac
(30 mg per ampoule) was HK$7.50 per treatment.
Hence, the cost of tramadol was HK$5.13 or 68.4%
less than ketorolac.
The limitation of our study was the small sample size.
However, the sample size was calculated beforehand
with a usual accepted significance, and the difference
in analgesia was shown to be statistically significant.
Randomising patients by drawing from equal preset
numbers of envelopes could cause randomisation error,
but the characteristics of the patients in both groups
were still comparable.
Despite the limitations, our study showed that
intramuscular tramadol and oral paracetamol was as
effective as NSAID in the treatment of moderate to
Hong Kong j. emerg. med. „ Vol. 15(1) „ Jan 2008
severe acute musculoskeletal and arthritic pain. In
addition, it may have less severe side effects and is
cheaper. Further studies are needed to demonstrate its
effectiveness in specific disease entities like arthritic
pain and to assess its effect after repeated dosages and
after a longer period.
Conclusion
Our study showed intramuscular tramadol and oral
paracetamol combination therapy was effective and safe
to use in acute pain management. Though many
patients experienced side effects, all of them were
minor. Patients in the tramadol group experienced
significantly more nausea, but no major adverse event
occurred. An intramuscular tramadol and oral
paracetamol combination is appropriate as first line
analgesic in acute musculoskeletal pain control in the
emergency department. However, the effect of long
term repeated use of intramuscular tramadol has to be
monitored. The adverse events need to be reported and
documented.
Acknowledgement
The authors would like to thank all the staff of the
Accident and Emergency Depar tment, United
Christian Hospital who contributed to the study.
References
1. Natalie S, Lyn M. Management of acute musculoskeletal
pain. APLAR J Rheumatol 2005;8:5-15.
2. Jones JB. Pathophysiology of acute pain: implications
for clinical management. Emerg Med (Fremantle) 2001;
13(3):288-92.
3. Schug S, Macintyre P. The scientific evidence in acute
pain management. Acute Pain 2005;7:161-5.
4. Siu AYC, Chung CH. A pilot study on the use of
tramadol hydrochloride for pain control in an
emergency department. Hong Kong J Emerg Med 2004;
11(1):16-20.
5. Chung CH. The use of injectable nonsteroidal anti-
inflammatory drugs in local accident and emergency
practice. Hong Kong J Emerg Med 2002;9(2):65-71.
6. Ollé Fortuny G, Opisso Julia L, Oferil Riera F, Sánchez
Lee et al./Tramadol versus ketorolac
Pallarés M, Calatayud Montesa R, Cabré Roca I.
Ketorolac versus tramadol: comparative study of analgesic
efficacy in the postoperative pain in abdominal hysterectomy.
Rev Esp Anestesiol Reanim 2000;47(4):162-7.
7. Putland AJ, McCluskey A. The analgesic efficacy of
tramadol versus ketorolac in day-case laparoscopic
11
sterilisation. Anaesthesia 1999;54(4):382-5.
8. Rosenthal NR, Silverfield JC, Wu SC, Jordan D, Kamin
M; CAPSS-105 Study Group. Tramadol/acetaminophen
combination tablets for the treatment of pain associated
with osteoarthritis flare in an elderly patient population.
J Am Geriatr Soc 2004;52(3):374-80.