Fundom Clin Pharmacol1996;10387-392 387

8 Elsevier. Paris

Original article

A randomized, double-blind,multicentre controlled trial

of ibuprofen versus acetaminophen and placebo
for symptoms of acute otitis media in children

L Berth', G Pons lo,P d'Athis l, JF Duhame13,C Maudelonde2,

G Lasfargues4, M Guillots, A MarSac2, B Debregeas6, G Olive
Dlpartenunt de Pharmacologie Clinque P C r i d e et PCdiatrique. 2Service d'Oto-Rhino-Loryngobgie,
HBpital Saint-Vincent& Paul, 82 avenue Denfert-Rochereau, 75674 Paris cedex 14; 3Service & PCdiatrie du CHRU & Caen.
avenue Georges ClCmcnceau, 14033 Caen; 4Service & PCdiatrit & I'HBpital Trousseau, 26 rue du Dr Arnold Nener, 75012 Paris:
JServke h PCdiatrie de I'HBpital de Lisiery 4 me Roger Aini. BP 223.14107 Lisieux:
6Laboratoins Ethyphann, 21 rue Saint-Mazthieu.ZI. 78550 Houdan, France
(Received 20 Fcbruary 1995; accepted 5 February 1996)

-
Summary Two hundrul and nineteen children (boys: 56%. girls: 44%) were included in a randomid. double-blind, multiccntrc
(4 ccntrcs) controlled trial designed to assess the efficacy and safety of ibuprofen (IBU) in the (reatmcnt of 1 to 6 year-old children
with otoscopically proven acute otitis media (AOM). either unilateral or bilateral. l h c y randomly rcccived 10 m a g IBU (n = 71).or
acetaminophen (PARA) (n = 7 3 ) or placebo (PLA) (n = 75). orally, tid, for 48 hours. All m i v e d oral ccfaclor (Alfatilo. Lilly.
France) for sevm days. They were evaluated b c f m (DO) and at the end of treatment 0 2 ) . The main criterion of response was the
aspect (landmarks and color) of the tympanic membrane asessai on a semiquantitative scale from 0 to 6. Otha criteria, assessed on
semiquantitative scales. included relief of pain (0 or 1). rectal temperature (0 to 2). and overall evaluation by parents of the
improvement of quality of life on thrce items: appetite (0 to 2). sleep (0 to 2) and playing activity (0 to 2). lhc results at D2 w m as
follows: thcn was no signifmnt difference bctwm trcptmcnt groups as to the main criterion, but only a trcnd for IBU and PARA to
do bmcr than PLA but not for IBU to do bcttcx than PARA. From these data thae is no argument to emphasize the utility of non-
steroidal anti-inflammatory drugs (NSAIDs) in treating the inflammatory signs of the tympanic membrane in otitis. Thcre was a
statistically significant difference W e e n tnxbncnt groups at D2 for pain. IBU being superior to PLA ( P < 0.01): 7%.lWb and 25%
of the children w m still suffaing at D2 in the IBU.PARA and PLA treatment groups. respectively. The difference between PARA
and PLA for pain was not statistically significant. There was no significant difference bctwccn treatment p u p s for thc othcr criteria.
All treatments w e n well and cqually tolerated. Although no significant difference was found between the treatment groups on the
aspect of the tympanic membrane. the efficacy of IBU was evidenced on the relief of pain, the symptom that most disturbs the child.

~ n - s t c r o M pmti-lnllnmmntory
I drug I otitb mcdip I pnin I childhood

INTRODUCTION well tolerated and available by prescription for

Ibuprofen is a non-steroidal anti-inflammatory
drug (NSAID) belonging to the propionic acid
group. Its efficacy as an analgesic (Moore et al,
1985; Laska et al, 1986). antipyretic (Walson et
al. 1989; Wilson et al, 1984; Simila et al, 1976)
and anti-inflammatory (Menon et al, 1974) drug
has been demonstrated in children. Ibuprofen is
children in the USA and the UK. Ibuprofen is
frequently used empirically for pain control in
children with acute otitis media (AOM), but,
until now, its efficacy has not been established
in this indication. The present study was
designed to assess the efficacy and safety of ibu-
profen given orally, compared to acetaminophen
and to placebo, for the treatment of 1 to 6 year-

'Comspondence and reprints.

388 L Bertin et al
old children suffering from AOM and treated by
cefaclor.
PATIENTS AND METHODS
Patients
Two hundred and nineteen children (122 males) with
otoscopically proven AOM, either unilateral or bilateral,
were included (30 November 1988 to 3 March 1990) in
this multicentre (4 centres) trial. They were aged from 1
to 6.75 years (mean age: 2.98 f 1.33 years). All children
were seen as outpatients. Informed consent was obtained
from parents. This study was approved by the hospital
Ethics Committee.
Patients with cardiac, hepatic or renal disorders, gas-
troduodenal disease, known hypersensitivity to NSAIDs
and penicillins or receiving either antibiotic, analgesic,
diuretic or anti-inflammatory drugs within one week
prior to the study were not eligible. The diagnosis of
AOM was based on the aspect of the tympanic mem-
brane (TM)using a semiquantitative scale ("tympanic
score") from 0 to 6 as follows: A- alterations of the
ossicular landmarkr (TM bulging in its posterior quad-
rant or purulent discharge = 3; total disappearance of the
landmarks of the short process of the malleus, with no
posterior bulging = 2; short process of the malleus still
visible but blurred = 1; normal landmarks = 0) + B-
modifications of the TM color (red TM with a posterior
orange zone indicating a suppuration, or acute bullous
myringitis = 3; red TM with disappearance of the light
reflex = 2; generalized dilatation of the TM blood ves-
sels = 1; normal color, ie, grey-white TM with light
reflex = 0). This score (A+B) had to be 1 3 for at least
one ear for patients to be eligible. Patients with AOM
requiring myringotomy, or with chronic otitis were not
eligible.
Treatments
Patients were randomized (a computer generated list was
given to each center) into three parallel treatment
groups: 7 1 children were given ibuprofen (microgra-
nules in packets of 100 or 200 mg); 73 and 75 children
received identically looking acetaminophen and placebo
microgranules. respectively. All treatments were sup-
plied by Ethypharm. Drugs were given orally at a dos-
age of 10 m g k g tid (just before breakfast, lunch and
dinner) for 48 hours. All children received cefaclor
(Alfatile) for 7 days (15-30 mgkg bid). External cooling
techniques such as sponging, undressing the child, low-
ering the environmental temperature, etc were allowed
in case of fever > 39 "C. In case of fever 139 "C. chil-
dren could be given acetaminophen (30-60 mg/kg/d)
additionaly to the studied treatment, whichever treat-
ment group they belonged to. The total dose of acetami-
nophen had to be noted in the case report form. No local
treatment was allowed.
Response criteria
EYF=Y
The criteria were assessed in the clinic just before (DO)
and after 48 hours (D2) of treatment. The main criterion
was the aspect (ossicular landmarks and color) of the
tympanic membrane assessed using the above described
semi-quantitative scale from 0 to 6.
Other criteria, assessed on semi-quantitative scales.
included: a) evaluation of otalgia (or pulling of the ear in
the young infant) (0 = no pain; 1 = pain in the ear);
b) rectal temperature (0 = 5 37 "C. 1 = 37.1 to 38.4 "C,
and 2 = 2 38.5 "C); c) overall evaluation by the parents
of the quality of life using three items (appetite, sleep
and playing activity), each item being evaluated using a
three-point scale of increasing severity from 0 to 2:
0 (normal), 1 (moderate impairment), or 2 (severe
impairment).
Tolerance
Tolerance was assessed after 48 hours of treatment on
the basis of clinical criteria: nausea, vomiting, abdomi-
nal pain and cutaneous rash. The presence or absence of
these symptoms at onset of treatment was recorded.
Spontaneous reports of other side effects were also
recorded.
Statistical methods
The sample size was determined for the main criterion
with a = 0.05 and = 0.05 assuming that the frequency
of symptom recovery (tympanic score = 0 to 2) would
be 3540% in the placebo group and 50-555 in the ibu-
profen group (20% difference) within 48 hours after the
onset of treatment. As the choosen statistical strategy
was to compare the three treatment groups, and not to
compare the score between Do and D,, the treatment
groups were compared twice: first at DO then at D2. The
quantitative data were compared by use of an ANOVA,
and, in case of significant overall difference, the treat-
ments were then compared to each other with Scheffe's
test, a modified version of Student's method to compare
two means. The semi-quantitative data were processed
by the x z test: we first looked for an overall difference
between the three treatment groups; in case of statistical
significance (P c 0.05). the treatments were then com-
pared to each other with a calculated significance thresh-
old P = 0.01.
The main criterion was the aspect of the TM. If the
otitis was bilateral, we used for the statistical analysis
the following rule: 1) at DO, if the right and left scores
differed, we chose the highest score (3 to 6), and, 2) at
 Ibuprofen for acute otitis media in children 389

Table I. Clinical data (mean, standard deviation, range).

Age Weight Gender Otitis

(Years) (kX) Male Female Unilateral Bilateral
(96) (96) (96) (96)
Ibuprofen (n = 71)
3.00 (1.36) 14.3 (3.3) 56.30 43.70 21.10 78.90
(1 .O - 6.58) (9.60 - 22)
Acetaminophen (n = 73)
3.07 (1.37) 14.3 (3.6) 63.00 37.00 16.40 83.60
(1.5 - 5.92) (9.95 - 25)
Placebo (n = 75)
2.87 (1.27) 14.4 (3.6) 48.00 52.00 14.70 85.30
(1.5 - 6.75) (1 0.0 - 26)

Table IL Tympanic score as main criterion (mean. standard deviation, range).

Alterations of the ossicular Modifications of the Tympanic score (A + B )

landmarks (A) TM (*) colour (€3)
kfr Right kfr Right Left Right Main criterion
DayO Day2 DayO Day2 DayO D a y 2 DayO Day2 DayO Day2 DayO Day2 DayO D a y 2

Ibuprofen (n = 71)
1.55 0.56 1.8 0.79 1.45 0.54 1.73 0.87 3 3.54 1.66 4.23 1.93
(0.89) (0.73) (0.73) (0.7) (0.82) (0.65) (0.74) (0.67) (1.66) (1.39) (1.3) (0.8) (1.31)
(0-3) (0-3) (0-3) (0 - 3) (0- 3) (0- 2) (0- 3) (0- 3) (0 - 6) (0- 6) (0 - 6) (3 - 6) (0 - 6)
Acetaminophen (n = 73)
1.73 0.84 1.75 0.78 1.66 0.79 1.67 0.75 3.38 3.42 1.53 4.3 2.07
(0.93) (0.65) (0.85) (0.61) (0.84) (0.69) (0.75) (0.62) (1.67) (1.5) (1.12) (0.92) (1.13)
(0-3) (0-2) (0-3) (0- 2) (0- 3) (0- 2) (0- 3) (0 - 2) (0- 6) (0- 6) (0- 4) (3 - 6) (0- 4)
Placebo (n = 75)
1.75 1 1.65 0.93 1.68 0.99 1.63 0.88 3.43 3.28 1.81 4.25 2.4
(0.81) (0.68) (0.78) (0.7) (0.74) (0.69) (0.69) (0.59) ( 1.47) (1.41) (1.19) (0.82) (1.05)
(0-3) (0-3) (0-3) (0- 3) (0- 3) (0- 3) (0- 3) (0- 2) (0- 6) (0- 6) (0 - 5 ) (3 - 6) (0- 6)

D2. if the right and left scores were different at DO. we
chose the score of the ear with the highest score at DO;if
not, we used the highest score at D2 (0 to 6). In this
way, a child was taken only once into account in the
analysis and each ear was compared to itself. A comple-
mentary analysis was performed on the results obtained
with the remaining ear. in case of bilateral otitis.
The 95% confidence interval of the difference between
treatment groups was calculated for each criterion.
RESULTS
The data obtained in each of the three treatment
groups are presented in tables I to 111.
Comparability of treatment groups
At the time of inclusion (DO), there was no signifi-
cant difference between the three treatment groups
as regards sex distribution, age, weight and evalua-
tion criteria. There was a majority of bilateral
otitis, and no significant difference between treat-
ment groups in this respect (79, 84 and 85% of
bilateral AOM in the ibuprofen, acetaminophen
and placebo groups, respectively).
Efficacy
After 48 hours of treatment (D2). there was no sig-
nificant difference between the 3 treatment groups
as regards the main criterion (fig 1): the observed
scores 0 to 2 (symptom recovery) were more fre-
quent than expected for both ibuprofen (76% vs
50-55 %) and placebo (65% vs 3540%). the corre-
sponding difference (1 1%) being smaller than the
targeted difference (20%); the power of the test
(a= 0.05) for a theoretical difference of 11% was
1 - p = 0.30. It should be noted, however, that the
mean tympanic score was equal to 1.9. 2.1 and 2.4
in the ibuprofen, acetaminophen and placebo
group, respectively, and that the TM was normal
(score equal to 0) in 12 (16.9%). 8 (1 1%) and 2
390 L Bertin et al

Table III. Scores for secondary criteria of efficacy (mean, standard deviation, range).

otalgia Rectal temperature Appetite Sleep Playing activity

Day 0 Day 2 Day 0 Day 2 DayO Day2 DayO Day2 Day 0 Day 2

Ibuprofen (n = 71)
0.86 0.07 1.27 0.35 1.21 0.59 1.27 0.39 0.94 0.3 1
(0.35) (0.26) (0.5 1 ) (0.51) (0.61) (0.6) (0.61) (0.55) (0.56) (0.47)
( 0 - 1) ( 0 - 1) (0 - 2) (0 - 2) (0-2) (0-2) (0-2) (0-2) (0-2) (0- 1)
Acetaminophen (n = 73)
0.78 0.1 1.22 0.38 1.21 0.52 1.19 0.37 0.97 0.25
(0.61) (0.52) (0.62) (0.58) (0.7) (0.51) (0.62) (0.46)
(0 - 2) (0 - 2) (0-2) (0-2) (0-2) (0-2) (0-62) (0-2)
1.27 0.52 1.25 0.61 1.39 0.44 0.88 0.32
(0.5) (0.53) (0.55) (0.57) (0.61) (0.58) (0.59) (0.47)
(0 - 2) (0 - 2) (0-2) (0-2) (0-2) (0-2) (0-2) (0-1)

IBUPROFEN PLACEBO ACXTAMINOPHEN

100 100

8u

60 60

40 40

m m
0 0
w m w M

Fig 1. Tympanic score at DO andD2: 0 = 0.1= m. 2 = m, 3 = m, 4 = m. 5 = B,6 =.. The difference betwcentnatment p u p s
was not statistically significant.

(2.7%) children in the ibuprofen, acetaminophen
and placebo group, respectively.
The three groups were significantly different (P
< 0.05) as regards otalgia (fig 2): 5 (7%). 7 (9.6%)
and 19 (25.3%) children were still suffering at D2
in the ibuprofen, acetaminophen and placebo
groups, respectively. The difference between ibu-
profen and placebo was statistically significant (P
< 0.01): the 95% confidence interval of the differ-
ence between ibuprofen and placebo was 6.8 -
29.8%. There was no significant difference
between ibuprofen and acetaminophen, nor
between acetaminophen and placebo; the 95%
confidence interval of the difference between ace-
taminophen and placebo was -3.8 - 27.7%.
There was no significant difference between
treatment groups at D2 as regards: rectal tempera-
ture, the assessment of quality of life (appetite,
sleep and playing activity), and the number of chil-
dren requiring the administration of acetamino-
phen despite the treatment studied (7, 6 and 11
children in the ibuprofen, acetaminophen and pla-
cebo groups, respectively).
Tolerance
Eleven children experienced mild unexpected
events: 3 (4%) in the placebo group (nausea,
vomiting and abdominal pain), 3 (4.1%) in the
acetaminophen group (nausea and abdominal
pain) and 5 (7%) in the ibuprofen group (nausea,
abdominal pain and a cutaneous rash in one
child). No other unexpected event was reported.
Treatment was never interrupted because of unex-
pected events.
DISCUSSION
NSAIDs are empirically given to children with
acute otitis media, as their anti-inflammatory prop
 Ibuprofen for acute otitis media in children
IBUPROFEN
100
so
68
40
20
Do DZ
a
Ng 2. Evaluation of otalgia at DO and D2: 0 = 0.
(P < 0.01, ~2 test).
1= ..
w DZ
The difference between treatment groups was statiscally significant at D2
erties are supposed to have beneficial effects in
this disease, but their efficacy has not yet been evi-
denced. It was then of interest to assess whether
ibuprofen, a common antiinflammatory drug,
could be useful in children with AOM, and to
compare its effects with those of a placebo, but
also of acetaminophen, a popular and safe drug
which has analgesic and antipyretic properties but
no anti-inflammatory activity.
In the present study, the main criterion of
response was a “tympanic score” taking into
account the alterations of the ossicular landmarks
and color of the tympanic membrane, so that the
potential benefits of the anti-inflammatory effect
of ibuprofen could be evidenced. At D2, the differ-
ence between ibuprofen and placebo for the tym-
panic score did not reach statistical significance. It
should be noted that all children received cefaclor
and that with appropriate antimicrobial therapy,
most children with acute bacterial AOM are signif-
icantly improved within 48 to 72 hours (Bluestone,
1989).
Despite the lack of repeated measures of pain
relief, difficult to obtain in outpatients, that pre-
cludes the recording of time to onset of reponse, a
significant difference between treatment groups
was noted for otalgia at D2. The difference
between ibuprofen and placebo was significant for
this criterion, but the superiority of ibuprofen
compared to acetaminophen has not been demon-
strated in this respect. However, this effect of ibu-
profen on the relief of pain is of importance, as
this symptom is the one that most disturbs the
children with AOM. It is worth noticing that more
than 80% of all the children had otalgia at the
time of inclusion. The discrepancy that we
39 1
PLACEBO ACETAMINOPHEN
observed between a significant analgesic effect
and the absence of demonstrated antiinflammatory
effect may be explained at least in part by the dif-
ferences in dose requirement for these two effects
since the antiinflammatory effect is usually
obtained with 40 mg/kg/24h. The single dose-
effect relationship studies showed that a 10 mgkg
dose was sufficient to obtain an antipyretic effect
up to 8 hours after dosing allowing a recomman-
dation of 30 mg/kg/day and suggesting a discrep-
ancy between the thresholds to observe antipyretic
and anti-inflammatory effects (Walson et al.
1989).
No significant difference between treatment
groups was evidenced on rectal temperature. This
was not, however, the main goal of the present
study. This criterion was assessed using a semi-
quantitative scale, which is not the best way to
demonstrate an antipyretic effect. Furthermore,
no repeated measures of temperature between DO
and D2 were done and very few children had a
temperature 2 38.5 “C at D2 (1 in each group).
There was no significant difference between ace-
taminophen and placebo as regards pain control
or rectal temperature. Acetaminophen was used
according to the dosage recommended in France
at the time of the study. This dose regimen was
not based on available scientific data. Available
data in the literature are in favor of a 10-15
mgkg single dose repeated every 4 to 6 hours
(Pons et al, 1990). Fifty percent more acetamino-
phen could have been safely given, and the poten-
tial efficacy of this drug may have been underes-
timated.
In both groups all antipyretic on going treat-
ments were stopped at the time of inclusion. Any
392 L Bertin et al
additional dosage of paracetamol was not planned
before at least three hours after inclusion in the
trial. In case of fever, children in the paracetamol
group could have received paracetamol twice
(30 mg/kg/day, as the study treatment, + 30-
60 mg/kg/day, as a rescue treatment). Nahata et al,
1994 showed that in 2 to 8 year-old children
receiving 12-14 mg/kg every 4 hours (72-
84 mg/kg/24h) for 1 to 3 days, plasma concentra-
tions were never above 40 mg/L and no sign of
hepatic toxicity was evidenced. We therefore con-
sidered that the patients in the paracetamol group
were not at risk of overdosage.
We conclude that ibuprofen combined with
antibiotic therapy could have an interest as a
short-term treatment of pain related to acute
otitis media in children. Further studies have to
be performed in order to evaluate the benefitfrisk
ratio.
ACKNOWLEDGEMENTS
We gratefully acknowledge the assistance of Dr
C Moran (Saint-Vincent de Paul). We gratefully thank
V Andriwx for her secretarial assistance.
REFERENCES
Bluestone CD. Modern management of otitis media. Pediotr
Clin N o ~ Am h 1989;61371-87
Laska EM. Sunshine A. Marrero I. Olson N. Siege1 C. McCor-
mick N. The comlation between blood levels of ibuprofen
and clinical analgesic response. Clin Phormacol Ther
1986;4O: 1-7
Menon ND. Whelton JJ, Ansell BM and Goldberg A N . A
comparative study of ibuprofen and aspirin in Still's disease
(juvenile chronic polyarthritis). Abstracts of the XIIIth
international Congress of Rhcumatology. Kyoto, Japan.
Cum Med Res Opin 1974;(suppl): 17-8
Moore PA. Acs G. Hargreaves JA. Postextraction pain relief in
children: a clinical trial of liquid analgesics. Inr J Clin
Pharmacol Thcr Toxic01 1985;23:573-7
Nahata MC. Powell DA, Dumll DE. Miller MA. Acetamino-
phen accumulation in pediatric patients after repeated thera-
peutic doses. Eur J Clin Pharmncol 1994;27:57-9
Pons G. Badoual J, Olive G. Optimal dose of acetaminophen in
children (English abstract). Arch Fr Pedicrtr 1990;47:539-42
Simill S. Kouvalainen K. Keinben S.Oral antipyretic therapy.
S CJ R ~~ ~ I U P1976;5:81-3
U~O~
Walson PD, Galletta G. Braden NJ. Alexander L. Ibuprofen,
acetaminophen, and placebo treatment of febrile children.
Clin Pharmacol Thcr 1989;46:9-17
Wilson G. Guerra AJMS. Santos NT.Comparative study of the
antipyretic effect of ibuprofen (oral suspension) and ace-
taminophen (suppositories) in paediatrics. J Int Med Res
1984;I 2 : 2 5 W