A Randomized, Double-Blind, Comparative Study
of the Efficacy of Ketorolac Tromethamine Versus
Meperidine in the Treatment of Severe Migraine
From the Department of
Surgery, Emergency Medicine
Service, West Virginia University
Health Sciences Center,
Received for publication
January 9, 1992. Revision
received April 2, ] 992. Accepted
for publication April 13, ] 992.
Gregory Luke Larkin, MD
John E Prescott, MD, FACEP
Study objective: To evaluate the relative efficacy of ketorolac
tromethamine and meperidine hydrochloride in the emergency depart-
ment treatment of severe migraine.
Design: Prospective, randomized, double-blind trial.
Setting: University hospital ED.
Participants: Patients presenting to the ED with an isolated diagnosis
of common or classic migraine.
Interventions: Subjects were randomized to receive a single intra-
muscular injection of either 30 mg ketorolac or 75 mg meperidine.
Measurements and main results: Of the 31 patients completing
the trial, 15 received ketorolac and 16 received meperidine. The demo-
graphic characteristics of beth groups were comparable. At one hour,
ketorolac was significantly less effective than meperidine in reducing
headache pain (P= .02) and in improving clinical disability (P= .01).
Ketorolac also was less effective at reducing nausea, photophobia,
and the need for rescue medication (P< .05). Sustained headache relief
was experienced by 44% of the patients treated with meperidine at
12- to 24-hour follow-up, compared with 13% of the patients treated
with ketorolac (P= NS). No significant side effects were observed for
Conclusion: IM ketorolac tromethamine is less effective than
meperidine in the ED treatment of severe migraine.
[Larkin GL, Prescott JE: A randomized, double-blind, comparative study
of the efficacy of ketorolac tromethamine versus meperidine in the
treatment of severe migraine. Ann Emerg MedAugust 1992;21:919-924.]
AUGUST 1992 21:8 ANNALS OF EMERGENCY MEDICINE 9 1 9 / 3 7
Larkin & Prescott
Migraine headache is an extremely common disorder affect-
ing more than 8 million Americans and about 15% to 20% of
the general population. ~,2 Treatment of the acute attack has
traditionally included ergot derivatives or narcotics, although
neither is ideal in the emergency setting. Ergots, for example,
can lose their efficacy once the headache is already well
established. 3 Moreover, ergot compounds are contraindicat-
ed in patients with severe neurologic symptoms, peripheral
vascular disease, and coronary artery disease. They are also
contraindicated in pregnancy, which is particularly relevant
because most migraine patients are women of childbearing
Narcotic analgesics are often used instead of ergot prepa-
rations for the emergency department treatment of severe
migraine; however, narcotics have well-known limitations of
their own: sedation, nausea, respiratory depression, urinary
retention, hypotension, and, over time, addiction. In addi-
tion, narcotics may confound the neurologic examination
and thereby obscure other differential diagnoses.
The development of tolerance to narcotics and ergot alka-
loids makes apparent the need for alternative therapy, z To
this end, various nonsteroidal anti-inflammatory analgesic
preparations have been used orally to abort acute migraine
with some limited success. 4-9 However, giving oral medica-
tions to nauseated migraineurs is fraught with obvious prob-
lems. Recently, the availability of a parenteral nonsteroidal
anti-inflammatory analgesic in the form of ketorolac
tromethamine has afforded the vomiting patient with a non-
narcotic, nonsedating, parenteral alternative. Ketorolac
(Toradol ®) is a new, parenteral analgesic indicated for the
short-term management of pain. This drug has demonstrated
efficacy superior or equal to the opiates in the treatment of
renal colic and postoperative, dental, and cancer pain. x°-z5
We could find no previous reports of its efficacy in the treat-
ment of pain from migraine headache.
The purpose of this study was to compare the efficacy of
meperidine, a standard therapy for severe migraine, with
ketorolac in the ED treatment of acute migraine attacks.
Exclusion criteria for patients in the migraine study
Age less than 18 or more than 60 years
Risk of nonsteroida] or opiate intolerance
Known or potential pregnancy* or breast-feeding
First migraine, nonmigraine headache, or mild migraine~ headache
Suspected substance abuse and/or previous treatment with meperidine
Hepatic or renal dysfunction, cardiovascular disease, or other serious illness
History of recent trauma or seizure
Marked hyperpyrexia, hypertension, or orthostatic hypotension
*Female patients of childbearing age with an unreliable menstrual history.
tMild migraine refers to a migraine that does not interfere with the capacity for normal
function (eg, working).
MATERIALS AND METHODS
This study was approved by the West Virginia University
Institutional Review Board for the protection of human sub-
jects in biomedical research. Patients who were diagnosed by
the attending emergency physician as having an acute attack
of classic migraine (with aura) or common migraine (without
aura) as defined by the International Headache Society 16
were considered for entry into the study. Eligible patients
were enrolled if they met the exclusion criteria shown
(Figure 1). Written informed consent was obtained from all
eligible patients, and the principles of the Declaration of
Helsinki and patient confidentiality were maintained
throughout the study.
Clinical evaluation included a thorough history, physical
examination, and laboratory tests as deemed appropriate by
Comparison of demographic characteristics among ketorolac and meperidine
12 (75) 1.0
31.5 -+ 4.4
33.8 _+ 5.0
Migraine type (%)
Classic (with aura)
4 (25) .70
Initial headache severity/
clinical disability (%)
Grade 3--most severe
(incapacitated or bedridden)
Grade 2--marked (interferes
with work, job, ete)
Grade 1--mild (slightly impaired
but able to work)
11 (73) 14 (87)
4 (27) 2 (13) .39
Most common associated symptoms
at presentation (%)
Numbness or tingling
Family history of migraine (%) 10 (67) 8 (50)
Medication taken before
ED visit (%)
analgesic (including aspirin)
Prophylactic medications t
*Values are _+ 95% confidence intervals based on the standard deviation and the standard error of
qncludes calcium channel blockers, ~-blockers, and antidepressants.
38/920 ANNALS 0FEMERGENOYMEDICINE 21:8 AUGUSTt992
Larkin & Prescott
the attending emergency physician. Physicians completed a
headache history form on all patients detailing the onset,
duration, intensity, location, and type of migraine. Preci-
pitants, exacerbating and ameliorating factors, associated
complaints, family history, prodromal aura, self-medication,
and result of self-medication (both in the past and as
pertains to the present headache) also were recorded.
After giving informed consent, patients were assigned to one
of the two treatment groups by a computer-generated ran-
dom number. The subjects were randomized in a 1:1 fashion
to receive a single IM dose of either 75 mg meperidine
(Demerol ®) or 30 mg ketorolac (Toradol®). Both the investi-
gator and patient were blinded as to the drug used. A nurse
who was not involved in patient care except to give the injec-
tion was the only person who knew which medication the
subject received. This nurse was considered to be a disinter-
ested party and, by protocol, did not communicate informa-
tion regarding the case to anyone for the duration of the
Patients filled out questionnaires every 15 minutes to mea-
sure pain relief and any side effects experienced. Pain relief
was recorded on a verbal analog scale from "no relief" to
"complete pain relief." The scale included intermediate
labels of "some relief but less than half better" and "marked
relief/more than half better." Although less sensitive than a
visual analog scale, the simple descriptive scale was easier
for our photophobic patients to use. Blood pressure, pulse,
and headache-associated nausea, vomiting, photophobia,
phonophobia, and numbness were assessed on the same
schedule as pain relief.
The severity of the headache and the corresponding degree
of clinical disability were graded on an analog scale both
immediately before and one hour after treatment as follows:
Pain relief at 60 minutes
i ¸ 7~
(more than half better) (less than half better)
grade 0, none (able to work and function normally); grade l,
mild (working ability mildly impaired); grade 2, marked
(working ability markedly impaired); grade 3, severe (total
incapacitation, bedridden). Only patients with grade 2 or 3
disability were eligible for enrollment into the study.
If little or no relief was afforded at the end of the 60-
minute trial, rescue medication (eg, analgesics and/or anti-
nauseants) were offered to the patient. Any use of rescue
medication was documented by the clinician, as were any
adverse reactions or side effects. Treating physicians also
recorded their impression of treatment efficacy after 60 min-
utes as poor, fair, good, or excellent. Patients received fol-
low-up telephone calls the next day to establish the frequency
of migraine relapse and side effects and to assess overall
Clinical disability: Baseline and one hour after treatment
Grade 2 Grade 1 Grade 0
AUGUST1992 21:8 ANNALS OF EMERGENCY MEDICINE 921/39
Larkin & Prescott
Statistical analysis was carried out on the pain relief scores
and the improvement in disability using the Wilcoxon rank
sum or Mann-Whitney U test. All other comparisons were
made using the Fisher's exact test. All P values reported are
two-sided exact values calculated using Stat Xact ® software
(Cytel Software Corp, Cambridge, Massachusetts).
Thirty-three patients were enrolled in the trial; however, two
were subsequently removed from the study because they
failed to meet the exclusion criterion of first migraine attack.
Thus, 31 patients were included in the analysis: 15 in the
kctorolac group and 16 in the meperidine group. Comparison
of demographic characteristics showed that both treatment
groups were similar with respect to age, sex, type of migraine,
clinical disability, family history, associated symptoms,
duration of headache, and previous treatment (Table). No
patients had re-entered the study.
After one hour, patients treated with ketorolac reported a
smaller reduction in pain than patients treated with meperi-
dine (Figure 2). Two-sided analysis of the raw pain relief
score data using the Wilcoxon rank sum test shows the dif-
ference in pain relief to be significant (P = .02).
Assessment of clinical disability after 60 minutes revealed
that there was significantly less improvement in the ketoro-
lac group (P = .01). No patients treated with ketorolac indi-
cated that they could return to work unimpaired, but 25%
(four of 16) of the meperidine group reported the complete
absence of disability at 60 minutes (Figure 3). Fifty-three
percent of the ketorolac group reported no improvement in
disability grade during the study compared with 12.5% of
the meperidine group (P = .001).
Migraine patients with photophobia versus time
No. Photot:)hobic Patients
0 15 30
Ketorolac was also significantly less effective at reducing
the most common symptom accompanying the headache,
photophobia (P = .03) (Figure 4). Ketorolac did not reduce
nausea as well as meperidine, but the difference was not sta-
tistically significant (P =. 10). The numbers of patients with
other symptoms were too small for meaningful analysis.
Eleven of 15 patients treated with ketorolac required
rescue medication after one hour, compared with six of 16
(37.5%; mid P = .05) who received meperidine. Of the
patients who did not require rescue medication, five had
rebound at 12- to 24-hour follow-up: two in the ketorolac
group and three in the meperidine group (P = NS). Sustained
headache relief at 12 to 24 hours was reported by seven (44%)
of those who received meperidine and by two (13%) of those
who received ketorolac as monotherapy (P = NS).
Side effects of treatment were few. Increased nausea was
reported by one patient in each group, and drowsiness was
reported as a side effect in five patients: three from the
ketorolac group and two from the meperidine group. Nonethe-
less, no patients left the trial before the end of the 60-minute
observation period. There were no reported adverse reactions
to either drug.
The response to therapy was consistent between both com-
mon and classic migraine however, the sample size used was
too small to allow stratification of these data.
Although derangements in serotonin, endorphins, and
prostaglandins have all been suggested to play a role in the
pathogenesis of acute migraine headache, 17 no one mecha-
nism adequately explains the response (or lack thereof) to
different therapies. This confusion has led to a bewildering
array of suboptimal treatment options. 18 Antinauseants,
narcotics, and ergot compounds, for example, are standard
therapy, yet they carry with them well-known risks and
undesirable side effects, is The results of this study demon-
strate that 30 mg ketorolac is well tolerated in migraine
patients and causes very few side effects; however, it was not
as efficacious as 75 mg meperidine in relieving migraine
headache pain and disability.
Although our study used a small sample size and verbal
analog scales, the differences in pain relief and clinical
improvement between groups were consistent and statistically
significant. We found the verbal pain relief scale easier for
our patients to use than the more sensitive visual analog
scale. This was particularly true of patients experiencing
scintillating scotomata or other visual disturbances.
Measuring pain relief instead of pain ensured that all
patients started at the same baseline and had the same
degree of potential response. The use of a "pain relief"
rather than a "pain" scale receives some support from the
work of Huskisson, 19 who found it superior to subtracting
pain scores after treatment from initial pain scores.
40/922 ANNALS OF EMERGENCY MEDICINE 21:8 AUGUST1992
Larkin & Prescott
Previous studies using various nonsteroidal anti-inflamma-
tory analgesics in the treatment of acute migraine reported
good efficacy; 4-8 nevertheless, these data do not necessarily
apply to our study with ketorolac. First, none of the prior
studies were conducted in controlled ED or hospital-based
settings. In addition, the patients were neither monitored for
compliance nor controlled for ingestion of nonstudy drugs.
Because the subjects in these previous studies were told to
take the study medication at home at the first sign of
headache, many nonmigraine headaches may have been
included in the analyses.
Our study, however, addressed only 12 migraines as defmed
by stringent International Headache Society criteria. 16 More-
over, the headaches treated in our study were uniformly
severe and disabling; the pain was not only refractory to
treatments at home, but it caused the patients to seek relief
in an ED. Last, photophobia, a reportedly strong indicator
of attack severity, 2° was present in 100% of the study popu-
Our results show ketorolac to have only marginal efficacy
when compared with meperidine in the treatment of acute
migraine headache. A placebo control was not used because
it was deemed inhumane to delay or withhold treatment from
patients in acute pain. Surprisingly, our data are in marked
contrast to other trials, which found ketorolac, at 30 mg or
less, to deliver analgesia superior or equivalent to 100 mg of
meperidine.Z°,n,13,15 For example, Brown et a115 stated that
100 mg of meperidine and 5 mg of ketorolac were indistin-
guishable in the treatment of pain after major oral surgery.
Another comparative trial concluded that 10 mg of ketorolac
was as efficacious as 100 mg of meperidine for the treatment
of renal colic. 10
Although the migraine pain model studied is different than
the aforementioned pain models, it is noteworthy that all the
previous studies used higher doses of meperidine and lower
doses of ketorolac than our study. It is unlikely, therefore,
that doses of ketorolac of more than 30 mg would have been
of significant benefit. Numerous studies have already
demonstrated equivalent analgesic efficacy between 30 and
90 mg of ketorolac at one hour after treatment. 11,12,21-23
Similarly, there are no published data to suggest that 60-mg
doses are superior to 30 mg at one hour after dosing.
One possible reason for our study results showing ketoro-
lac to be less effective than meperidine is that prostaglandins
are not as important in migraine as serotonin (5-hydroxy-
tryptamine) TM or other mediators. 24 The relative success
of drugs that affect serotonin (eg, dihydroergotamine 25 and
sumatriptan 26) and dopamine (eg, metoclopramide 27 and
phenothiazines 28) suggests that ~-receptor agonists (eg, nar-
cotics) are often unnecessary. Regrettably, many non-narcotic
modalities provide inconsistent relief and have undesirable
side effects of their own. Metoclopramide and the pheno-
thiazines, for example, can induce acute dystonic reactions.
In addition to extrapyramidal reactions, phenothiazines can
also cause neuroleptic malignant syndrome, which, although
rare, is fatal in 20% of cases. 29 Both sumatriptan and dihydro-
ergotamine have demonstrated efficacy,25, 26 but both can
cause flushing, tingling, chest pain, and lightheadedness.
Moreover, the vasoconstrictive effects of ergotamine can lead
to claudication, angina, and in rare instances, myocardial
Comparing ketorolac and other new therapies to standard
narcotic analgesia is difficult because migraine trials are
particularly susceptible to large placebo effects. 31 In addi-
tion, the euphoric effects from narcotics may augment their
analgesic effects. 31 We did not include suspected substance
abusers or persons known to have been treated previously
with meperidine in our study. Nevertheless, further studies
comparing ketorolac with less-euphoric agents (eg, a narcot-
ic agonist-antagonist) may be warranted.
It is possible that the peak onset of action of meperidine is
faster than ketorolac. This may explain, in part, why
meperidine is still widely used as a standard antimigraine
therapy in many EDs.
There are several limitations to the use of single-dose intra-
muscular analgesics as used in our study. For example, there
is large interindividual variation in responsiveness to nar-
cotics that demands dosage flexibility. Moreover, the optimal
way to titrate narcotic analgesia is through the IV route.
Ketorolac, however, is not currently approved for IV use in
the United States.
Although ketorolac tromethamine has been demonstrated to be
superior to meperidine in several different pain models, it is not
as efficacious as meperidine in the ED treatment of acute severe
migraine. Meperidine was significantly better than ketorolac in
several outcome measures, including improvement in migraine-
associated pain, photophobia, and disability. Ketorolac,
although well tolerated, was associated with more refractory
migraine symptoms and a greater need for rescue medication,
The authors thank Raymond Jackson, MD, FACEP, for his thoughtful editing of the
1. Waters WE: Headache, Clinical Epidemiolagy, vol 2. Littleton, Massachusetts, John
Wright-PS6, 1986, p 37-39.
2. US Bureau ofthe Census: Statistical Abstract of the US, 1989, ed 109. Washington, BC,
US National Center for Health Statistics, 1989, p 118.
3. Scheife RT, Hills JR: Migraine headaches: Signs, symptoms, biochemistry, and
current therapy. Am J Hasp Pharm 1980;37:365.
4. Nestvold K, Kloster R, Partinen M, et al: Treatment of acute migraine attack:
Naproxen and placebo compared. Cephalalgia 1985;5:115-119.
5. Havana-Kanniainen H: Treatment of acute migraine attack: Ibuprofen and placebo
compared. Headache 1989;29:507-509.
6. Pradalier A, Rancurel G, Dordain 6, et al: Acute migraine attack therapy: Comparison
of naproxen sodium and an ergotramine tartrate compound. Cephalalgia 1985;5:107-113.
21:8 ANNALS OF EMERGENCY MEDICINE 923/41
KETOROLAC Download full-text
Larkin & Prescott
7. Hakkarainen H, Vapaatalo H, Gothoni G, et al: Tolfenamic acid is as effective as ergot-
amine during migraine attacks. Lancet 1979;2:326-328.
8. Sargent Jd, Baumel B, Peters K, et al: Aborting a migraine attack: Naproxen sodium
vs ergotamine plus caffeine. Headache 1988;28:263-266.
9. Andersson PG, Hinge HH, Johansen 0, et al: Double-blind study of naproxen vs
placebo in the treatment of acute migraine attacks. Cephalalgia 1989;9:29-32.
10. Oosterlinck W, Philip NH, Charig C, et al: A double-blind single dose comparison of
intramuscular ketorolac tromethamine and pethidine in the treatment of renal colic.
J Clin Pharmaco11990;30:336-341.
11. Fricke J, Angelocci D: The analgesic efficacy of IM ketorolac and meperidine for the
control of postoperative dental pain (abstract). Clin Pharmacol Ther 1987;41:181.
12. Yee JP, Koshiver JE, AIIbon C, et al: Comparison of intramuscular ketorelac
tromethamine and morphine sulfate for analgesia of pain after major surgery.
13. Yee J, Bradley R, Stanski D, et el: A comparison of the analgesic efficacy of
intramuscular ketorolac tromethamine and meperidine in postoperative pain (abstract).
Clin Pharmaco! Ther 1986;39:237.
14. Carlson RW, Morrison RA, Shear HB, et el: A multiinstitutional evaluation of the
analgesic efficacy and safety of ketorolac tromethamine, acetaminophen plus codeine,
and placebo in cancer pain. Pharmacotherapy 1990;10:211-216.
15. Brown CB, Moodie JE, Evans SA, et al: Efficacy of intramuscular ketorolac
tremetharnine and meperidine in pain following major oral surgery (abstract). Clin
Pharmacol Thor 1988;.43:161.
16. Headache Classification Committee of the International Headache Society:
Classification and diagnostic criteria for headache disorders, cranial neuralgias and
facial pain. Cephalalgia 1988;8(suppl 7):19-28.
17. Lance JW: A concept of migraine and the search for the ideal headache drug.
18. Peroutka S J: The pharmacology of current anti-migraine drugs. Headache
19. Huskisson EC: Measurement of pain. Lancet1974;2:1127-1131.
20. Dexter SL, Graham AN, Johnston ED, et el: Double-blind controlled study of paramax
in the acute treatment of common and classical migraine. BrJ Clin Pract1985;39:388-
21. Stanski DR., Cherry C, Bradley R, et el: Efficacy and safety of single doses of
intramuscular ketorolac tromethamine compared with meperidine for postoperative
pain. Pharmacotherapy 1990;10:40S-44S.
22. Brown CR, Moodie JE, Dickie 6, et ah Analgesic efficacy and safety of single-dose
oral and intramuscular keterolac trometharnine for postoperative pain.
23. 0'Hara BA, Fragen R J, Kinder M, ot al: Ketorolac tromethamine as compared with
morphine for treatment of postoperative pain. Clin Pharmacol Ther1987;41:556-561.
24. Della Bella D, Carenzi F, Casacci B, et al: Endorphins in the pathogenesis of
headache. Adv Neuro11982;33:75-79.
25. Cailaham M, Raskin N: A controlled study of dibydroergotamine in the treatment of
acute migraine headache. Headache 1986;26:168-171.
28. Cady RK, Wendt JK, Kirchner JR, et el: Treatment of acute migraine with
subcutaneous sumatriptan. JAMA 1991;265:2831-2835.
27. Tek DS, McClellan DS, Olshaker JS, et al: A prospective, double-blind study of
meteclepramide hydrochloride for the control of migraine in the emergency department.
Ann Emerg Med 1990;19:1083+1087.
28. Stiell IG, Dufour D6, Moher D, et el: Methotrimeprazine versus meperidine and
dimenhydrinate in the treatment of severe migraine: A randomized, controlled trial. Ann
29. Varia IM, Taska R: The neuroleptic malignant syndrome. Res Staff Physician
30. Magee R: Saint Anthony's fire revisited: Vascular problems associated with migraine
reed ication. Med J Aust 1991 ;154:145-149.
31. Tfelt-Hansen P, Olesen J: Methodological aspects of drug trials in migraine.
Address for reprints: Gregory Luke Larkin, MD, Department of Emergency Medicine,
William Beaumont Hospital, 3601 West 13 Mile Road, Royal Oak, Michigan 48073.
42/924 ANNALS OF EMERGENCY MEDICINE 21:8