Efficacy of amoxicillin/clavulanic acid in preventing infectious and

inflammatory complications following impacted mandibular

third molar extraction
Iciar Arteagoitia, MD, PhD,a Antonia Diez, PhD,b Luis Barbier, MD, PhD,c Gorka
Santamara, BSc,d and Joseba Santamara, MD, PhD,e Bilbao, Spain
UNIVERSITY OF THE BASQUE COUNTRY AND CRUCES HOSPITAL

Objective. To find out whether the frequency of postoperative infectious and inflammatory complications (IC) in subjects treated
with placebo (Pl) is greater than those treated with antibiotic (Ab) after extraction of an impacted mandibular third molar (M3).
Our hypothesis is there are more IC in Pl than in Ab, with a maximum ratio difference of 0.067.
Study design. A double-blind placebo-controlled randomized clinical trial. The sample was derived from the population of
subjects attending Cruces Hospital for evaluation and extraction of 1 M3 under local anesthesia. Patients were treated with
postoperative placebo or amoxicillin/clavulanic acid 500/125 mg 3 times a day during 4 days. The outcome variable was
infectious and inflammatory complications. Sex, age, smoking, molar depth, angulation, need for sectioning, ostectomy, and
operation time were recorded. Analysis was by intention to treat, risk measures, and logistic regression.
Results. In 490 subjects (259 Ab and 231 Pl), the frequency of IC was 1.9% in the Ab and 12.9% in the Pl group (OR 7.6, 95%CI
2.9-19.9; P \ .001). The number needed to treat was 10 (7-16). Unadjusted relative riskwas 0.15 (0.06-0.38) (P \ .001).
Absolute reduction risk was 0.11(0.066-0.155)]. Therefore, the hypothesis cannot be rejected. Multivariate analysis shows
treatment with antibiotic (OR = 8.66 (3.17-23.67); P \ .001) and age (OR = 1.08 (1.00-1.16); P = .029) are the only variables to
be included in the logistic regression model.
Conclusion. Amoxicillin/clavulanic acid is efficacious in reducing the incidence of IC following third molar extraction but
should not be prescribed in all cases.
(Oral Surg Oral Med Oral Pathol Oral Radiol Endod 2005;100:E11-8)

The incidence of infectious and inflammatory complications (IC) following an impacted mandibular third
molar (M3) extraction varies between 0% and 45%,
according to different published studies1-16 Controversy

Financed by the Health Research Fund FIS/GRAN dossier no.

00/0585. The trial patients insurance was taken out by the Basque
Health Department, Basque Health Service/Osakidetza, Osakidetza,
pursuant to the conditions laid down in RD 561/1993. The antibiotic
and placebo were supplied free of charge by Geminis (Novartis
generics). Chlorhexidine was supplied free of charge by LACER.
This clinical trial was presented at IADR/AADR/CADR 82nd
General Session (March 10-13, 2004), Honolulu, Hawaii.
a
Associate Professor, Stomatology Department, University of the
Basque Country.
b
Pharmacologist, private practice.
c
Professor, Stomatology Department, University of the Basque
Country. Maxillofacial Surgeon, Cruces Hospital, Basque Health
Service.
d
Associate Professor, Stomatology Department, University of the
Basque Country.
e
Head, Maxillofacial Surgery Department, Cruces Hospital, Basque
Health Service. Professor and Chairman, Stomatology Department,
University of the Basque Country.
Received for publication Jun 12, 2004; returned for revision Feb 13,
2005; accepted for publication Mar 24, 2005.
1079-2104/$ - see front matter
2005 Mosby, Inc. All rights reserved.
doi:10.1016/j.tripleo.2005.03.025

exists over the use of systemic antibiotics for its prevention. While there is some evidence that antibiotics
can reduce the incidence of postoperative complications,2-4,9,10,14,15,17,18 there is equally convincing evidence to the contrary.1,5-7,11-13,16,19,20 As the Cochrane
Library has published, better evidence is needed
regarding the use of antibiotic prophylaxis in patients
undergoing tooth extraction.21
The specific aim of this study was to assess the
efficacy of amoxicillin/clavulanic acid 500/125 mg in
preventing infectious and inflammatory complications
(IC) in M3 subjects using a double-blind placebocontrolled randomized clinical trial. Our hypothesis
was that the occurrence of IC was greater in patients
treated with placebo (Pl) than in those treated with
antibiotic (Ab), with a maximum ratio difference of
0.067.
Likewise we analyzed variables such as age, sex,
smoking, molar depth and angulation, ostectomy, need
for sectioning and intervention time to identify risk
factors associated with post-operative IC.8,10,22-26
PATIENTS AND METHODS
The randomized clinical trial (RCT) was unicentric,
prospective, placebo-controlled, and double-blinded
of 2 parallel groups. The RCT was approved by the
E11

OOOOE
July 2005

E12 Arteagoitia et al
Table I. Characteristics of the control variables recorded for each of the groups
Variable
Age:
Mean
SD
Gender:
Man
Woman
Smoking, cigs/day:
None
0-10
11-20
[20
Depth type*:
II
III
IV
Angulation:
vertical
mesioangular
distoangular
horizontal
Intervention time, sec**:
mean
SD
Sectioning
No
Yes
Ostectomy***:
None
1/3
2/3
3/3

Ab, n = 259

Pl, n = 231

Total, n = 490

Pl lost, n = 2

Ab lost, n = 2

26.33
2.56

24.33
2.58

0.59a
24.26
5.1

24.01
4.89

24.15
5.01
0.36b

99
160

98
133

197
293

1
1

1
1

1
1
0
0

0
0
2
0

2
0
0

0
1
1

1
0
0
1

0
0
2
0

330.5
13.4

832.5
378.3

1
1

0
2

0
2
0
0

0
2
0
0

0.43c
160
65
31
3

139
62
30
4

295
127
61
7

85
107
67

70
82
79

155
189
146

77
85
21
76

64
82
19
66

141
167
40
142

0.13c

0.93c

0.84a
517.21
276.29

522.03
255.78

519.5
266.56

120
139

96
135

216
274

0.32b

0.03c
14
208
34
3

6
170
50
5

20
378
84
8

Ab, Subjects treated with antibiotic; Pl, subjects treated with placebo.
*Type II: totally covered by soft tissue and without bone coverage; type III: totally covered by soft tissue and partially by bone; type IV: totally covered by bone.
**Time measured using chronometer, from commencement of incision to the cutting of the last suture, by an external observer.
***1/3: Between crown and dental neck; 2/3: between dental neck and 2/3 of roots; 3/3: between 2/3 of roots and apexes.
a,b,c
P, statistical significance of the homogeneity tests between the groups Pl and Ab. a: Logistics regression; b: Fisher exact bilateral test; c: Pearsons bilateral
chi square.

Cruces Hospital Ethics Committee and by the Spanish

Medicines Control Agency under protocol no. 00-0314.
All the patients taking part in the study fully understood
its scope and signed the informed-consent form.
Furthermore, they had a contact telephone number
throughout the trial duration to manage concerns.
The study sample was derived from the population
of subjects attending Cruces Hospital for evaluation
and extraction of an M3 under local anesthesia. The
study was carried out at the Maxillofacial Surgery
Department of Cruces Public Hospital between March
2001 and February 2003. Healthy patients of both sexes
were included. A single M3 was extracted from each
patient.
Patients with any bacterial endocarditis risk factors
were excluded, as were pregnant and breastfeeding
women. Also excluded were patients with acute
infections 10 days prior to the intervention and/or those

who had had to take antibiotics; and those with a

history of allergy or intolerance to the drugs used in this
study.
Ab received amoxicillin/clavulanic acid 500/125 mg
oral 3 times a day for 4 days after the intervention, and Pl
received placebo.
To determine sample size, we considered that the
difference in IC between Pl (10%) and Ab (3.3%) would
be equal to or less than 0.067, which would be equal to a
relative risk (RR) of 0.33. To test this hypothesis with a
type I error of .05 with a power of 80%, we needed a final
sample of 490 subjects. A sample of 540 subjects was
randomized, calculating a 10% expected loss.
The simple randomization was performed using the
C4-SDP program (Glaxo Biometry, Madrid, Spain)
and each of the enrolled patients was assigned the
corresponding blinded random successive treatment
number, which was used as the patients number.

OOOOE
Volume 100, Number 1

Arteagoitia et al E13

Fig 1. Sample distribution diagram.

Diagnosis of postoperative infection and inflammatory complication was performed by the main
researcher, according to previously published clinical
criteria:7,9,10,24,27,28
Oral temperature [37.88 after 24 h for no other
justifiable cause
Intraoral abscess diagnosed via fluctuation pus
drainage (yes/no)
Dry socket defined as absence of clot with necrotic
remains present in the alveolus accompanied by
severe mandibular pain (yes/no)
Severe pain persisting or increasing 48 h after
surgery accompanied by intraoral inflammation
(moderate or severe) and/or intraoral erythema
(moderate or severe)
Severe pain after 7th day accompanied by intraoral
inflammation (moderate or severe) and/or intraoral erythema (moderate or severe) for no other
justifiable reason which improves with antibiotic
treatment
Inflammation, erythema and pain were measured
qualitatively (none, slight, moderate, or severe). In
addition, pain was measured quantitatively, using the
visual analog scale (VAS) of 100 mm.
Likewise we analyzed whether variables such as age,
sex, smoking, molar depth and angulation, ostectomy,
odontosection, and intervention time, could be considered risk factors associated with postoperative infectious and inflammatory complications (Table I).
All extractions were performed by maxillofacial
surgeons, under locoregional anesthetic of the inferior

alveolar and buccal nerves with Ultracain (articaine

40 mg/mL 1 epinephrine 0.005 mg/mL). A vestibular
envelope incision, mucoperiostical flap, ostectomy with
irrigation of physiologic serum and/or sectioning
(depending on tooth position) were performed. Tooth
removal, elimination of the pericoronal follicle, curettage of the alveolus, cleaning, and flap suture followed.
Three sutures were placed.
In all cases after the removal, irrigation of the
alveolus was performed using chlorhexidine digluconate at 0.12% (Lacer, Barcelona, Spain) for 1 minute.
Normal postoperative treatment was prescribed for
both groups with diflunisal capsules 500 mg (Dolobid,
Madrid, Spain) every 12 hours during 2 days and subsequently as needed. If pain was moderate to severe,
a capsule of metamizol 575 mg (Nolotil, Barcelona,
Spain) every 8 hours was added. In addition, chlorhexidine mouthwashes at 0.12% were prescribed 3 times a
day for 7 days.
Ab patients were treated with amoxicillin/clavulanic
acid 500/125 mg (Geminis, Barcelona, Spain) tablets 3
times a day for 4 days, and Pl patients were given
identical tablets with placebo, dosed as for Ab.
After exodontia, each patient was given a numbered envelope containing 12 antibiotic or placebo
tablets, identical in shape, color, texture, size, and
packaging. Neither the patient nor the surgeon knew
its contents.
The follow-up period was 8 weeks. All the cases were
examined on the 7th postoperative day. Patients were
also examined whenever necessary at any time during
the follow-up period.

OOOOE
July 2005

E14 Arteagoitia et al
Table II. Description of patients diagnosed with IIC
Variables defining IIC
Patient
no.
61
62
66
75
77
87
99
109
129
131
151
161
183
185
200
211
222
224
225
239
251
290
330
339
342
353
367
392
395
413
415
436
442
458
424

Groupa
1
2
2
2
1
2
2
2
2
2
2
1
2
1
2
2
2
2
2
2
2
2
2
2
2
1
2
2
2
2
2
2
2
2
2

Infection
typeb
early
early
early
early
late
early
early
late
late
late
late
late
early
late
early
early
late
early
early
early
early
early
late
early
early
early
early
early
early
early
late
late
early
early
early

Fever

Intraoral
Abscess

Alveolitis

Inflam.c

Erythemad

Dehis.e

Pain 48 hf

Pain 6 daysg

Ageh

Inclusion
typei

yes

yes

yes

yes

yes

yes
yes
yes

yes

yes

yes

yes

yes

yes
yes
yes

yes
yes

yes

yes

yes
yes

yes

yes
yes

yes

yes

mod
mod
mod

mod

mod
mod
mod
mod
mod
mod
mod

mod

mod
mod
mod
sev

sev
sev

sev

sev
sev
sev

mod
mod
mod

mod
mod
mod
mod

mod

mod

sev
sev

sev

sev
mod
sev

yes

yes

yes

yes

yes
yes

yes
yes

yes
yes

yes

yes

yes
yes

yes

yes

yes

yes

yes

yes

yes
yes

yes
yes

yes
yes
yes

yes
yes
yes
yes
yes

yes

yes

yes
yes

yes
yes
yes

yes

26.5
25.3
27.5
19.2
18
24.4
42.2
22.4
24.1
28.4
28.4
19.7
22.2
32.9
18.2
21.9
22.2
24.8
23.2
25.2
22.9
40.1
29.1
20.4
29.1
29.6
21.9
21.5
33.1
36.6
24.3
31.4
25.8
25.4
20.7

III
IV
III
III
IV
III
III
II
IV
III
IV
IV
III
IV
IV
III
IV
III
III
III
II
III
III
IV
IV
II
IV
IV
IV
IV
II
IV
II
II
III

IIC, Infective inflammatory complication.

a
1: Antibiotic; 2: placebo.
b
Early: IIC during the first week; late: IIC after the first week.
c
Intraoral inflammation (mod, moderate; sev, severe).
d
Intraoral erythema (mod, moderate; sev, severe).
e
Suture dehiscence.
f
Severe pain persisting or increasing after 48 hours.
g
Severe pain persisting or increasing after day 6.
h
Age in years.
i
II: subgingival; III: osteomucosa; IV: endo-osseous.

On diagnosis of an IC, the blinding was opened

and rescue medication prescribed: in Ab metronidazole
500 mg (Flagyl, Madrid, Spain) 3 times a day for
7 days, and in Pl amoxicillin/clavulanic acid 500/125mg
(Geminis) 1 metronidazole 500 mg (Flagyl) 3 times a
day for 7 days.
The number of tablets (placebo/antibiotic) taken over
the trial was determined. On the 7th postoperative day
visit patients returned any unused medication, and a

count of unused tablets was performed by the investigator.

Only those patients who did not attend the follow-up
appointment were withdrawn from the study.
Table I shows the control variables registered in all
patients. The third molars were divided into 3 types
according to depth: types II (crown not covered by
bone), III (partially covered by bone), and IV (totally
covered by bone). Adverse effects were registered.

OOOOE
Volume 100, Number 1

Arteagoitia et al E15

Table III. Risk measurements

Table IV. Multivariate analysis

Risk
measurements

Complication,
n = 490

Complication,*
n = 49014

Infection/
inflammation

OR

95% confidence
interval

Risk in Ab
Risk in Pl
ARR (95%CI)
NNT (95%CI)

0.019
0.129
0.11 (0.066-0.155)
10 (7-16)

0.027
0.137
0.11 (0.06-0.16)
10 (7-17)

Group
Age
Sex 2
Smokes 2
Smokes 3
Depth 2
Depth 3
Angulation 2
Angulation 3
Angulation 4
Time
Sectioning
Ostectomy 2
Ostectomy 3
Ostectomy 4

8.66
1.08
0.936
1.07
1.34
2.32
2.6
1.03
1.44
2.69
0.98
1.04
0.60
0.46
1.61

0.000
0.029
0.9
0.87
0.59
0.11
0.1
0.96
0.63
0.1
0.38
0.94
0.66
0.54
0.76

3.17-23.67
1.00-1.16
0.33-2.62
0.45-2.59
0.47-3.99
0.84-6.60
0.83-8.11
0.33-3.26
0.33-6.28
0.83-8.75
0.89-1.02
0.39-2.73
0.60-5.95
0.04-5.5
0.076-34.10

ARR, absolute reduction risk; NNT, number needed to treat.

*Risk measurements included the 4 patients lost in the analysis, as if they
had been infected/inflamed for the sensitivity study.

The analysis was by intention to treat.

The statistical study was performed using statistics
software SPSS 10 for Windows (SPSS, Chicago, Ill).
The homogeneity of the groups was studied using
Fishers exact test, x2, and logistics regression.
The magnitude of the effect of amoxicillin/clavulanic
acid 500/125 mg against placebo was evaluated using
risk measurements: odds ratio (OR), absolute risk
reduction (ARR), and number needed to treat (NNT).
NNT is the number of patients who need to be treated to
prevent 1 adverse outcome, in this case an IC.
To determine the existence or not of confounding
and/or interaction variables, a multivariate study was
carried out using logistics regression analysis.
Three random intermediate analyses were carried out
for ethical reasons, analyzing the dependent variables.
RESULTS
A cohort of 494 patients between 18 and 60 years old,
199 men (40.2%) and 295 women (59.8), mean age
24.15 SD 5.01, were studied.
To achieve the predetermined sample size (490), 494
patients were enrolled (233 Pl, 261 Ab) because 4
patients (2 Pl and 2 Ab) failed to attend the follow-up
appointment (Fig 1).
Both groups (Pl and Ab) were well balanced with
respect to control variables: age, sex, smoking, depth,
third molar position, intervention time, and need for
sectioning. The ostectomy was greater in Pl (Table I).
The number of patients with IC was 35 (Table II): 30
in Pl (12.9%), 22 of which in the first week, 5 in Ab
(1.9%), 2 of which in the first week. Postoperative
infectious and inflammatory complications are between
2.9 and 19.9 times more frequent if antibiotics are not
taken (OR 7.6; CI 2.9-19.9; P \.001)
The relative risk value of 0.15 (0.06-0.38) (P \.001)
includes 0.33, and the ARR of 0.11 (0.066-0.155)
includes 0.067, whereby our hypothesis cannot be
rejected (Table III).
The multivariate analysis via logistics regression
shows an adjusted estimation of the effect with OR of
8.66, meaning that with a 95% confidence IC frequency

Estimation with 483 patients because the statistical study excluded 7 patients,
who smoked over 20 cigarettes a day and none of whom were infected/
inflamed.
OR, Odds ratio; Sex 2, women; Smokes 2, #10 cigarettes/day; Smokes 3, 11-20
cigarettes/day; Depth 2, type III (partially covered by bone); Depth 3, type IV
(totally covered by bone); Angulation 2, mesioangular; Angulation 3,
distoangular; Angulation 4, horizontal; Sectioning, yes; Ostectomy 2, 1/3
(between the crown and dental neck); Ostectomy 3, 2/3 (between the dental
neck and 2/3 of the roots); Ostectomy 4, 3/3 (between 2/3 of the roots and
apexes).

is between 3.17 and 23.67 times greater in Pl solely

because they did not take the antibiotic (Table IV).
The logistics regression with the control variables:
age, sex, smoking, depth, angulation, intervention time,
odontosection, and degree of ostectomy show there is no
interaction. Only age behaves as a confounding variable
(OR 1.08, CI 1.00-1.16; P = 0.029) (Table IV).
The possibility of IC was
Px 1=11e

3:7410:074 AGE
2:075 ANTIBIOTIC1 :

The angulation and depth variables show a greater

tendency to be a significant variable in relation with IC.
This is why we performed a stratified analysis of the risk
measurements of these 2 variables. Table V shows that
for third molars in vertical position and third molars type
II, antibiotic treatment was inefficacious.
There were no serious side effects. Sixteen mild side
effects were recorded (14 in Ab and 2 in Pl): 2 vomiting,
2 gastric pain, 1 mycosis and 11 diarrhea. The number
needed to harm was 28 (15-147).
DISCUSSION
The main purpose of our study was to obtain scientific
evidence on the efficacy of amoxicillin/clavulanic acid
500/125 to prevent IC after M3 extraction.
The study shows postoperative treatment with amoxicillin/clavulanic acid to prevent complication, is efficacious from a statistical point of view, although the
hypothesis put forward in the study cannot be rejected.

OOOOE
July 2005

E16 Arteagoitia et al
Table V. Stratified analysis of the risk measurements of the depth and angulation variables
Depth
Type II
Type III
Type IV
Angulation
Vertical
Mesioangular
Distoangular
Horizontal

P bilateral

NNT (95% CI)

ARR (95% CI)

OR (95% CI)

0.091
0.000
0.053

17 (8-infinite)
7 (5-13)**
9 (5-102)

0.059 (
0.001-0.12)
0.149 (0.077-0.221)
0.107 (0.01-0.204)

0.155 (0.015-1.58)
0.049 (0.005-0.504)
0.261 CI 0.074-0.92

0.092
0.017
0.033
0.001

15
12
9
5

0.065
0.086
0.11
0.186

0.155
0.112
0.268
0.099

(8-infinite)
(7-54)
(3-infinite)
(4-12)**

(
0.001-0.13)
(0.019-0.153)
(
0.073-0.293)
(0.086-0.286)

(0.015-1.581)
(0.014-0.924)
(0.024-2.986)
(0.198-0.498)

**The number of patients needed to treat (NNT) with antibiotic to prevent infectious and inflammatory complications shows antibiotic efficacy has clinical
relevance in third molars type III and those in horizontal position.

We carried out a pragmatic clinical trial, analyzed

according to intention to treat, analyzing the behavior of
a systemic antibiotic, under the usual working conditions of our public hospital service. Surgery was
performed by experienced surgeons. Irrigation of the
alveolus with chlorhexidine 0.12% was included and
likewise as a postoperative mouthwash.29-31
The sample is representative of the population on
which this surgery is typically performed. The patients
were recruited consecutively from the population
attending this department. Exclusion criteria were based
solely on ethics.
The characteristics of the variables studied in those
patients who did not sign informed consent, and as such
were not included in the study, showed no significative
differences.
The preliminary calculation of the sample size was
done based on a hypothesis. The complications with
antibiotic forecast was estimated on a retrospective
study and the expected results with placebo on consulted
bibliography.1,2,6,7,11
To limit classification bias, the infectious and inflammatory complication diagnosis was carried out including clinical signs and symptoms previously defined by
other researchers.
Nevertheless, we did not determine either the sensitivity or the specificity of the infection/inflammatory
diagnostic test.
Not satisfied with diagnosing infectious and inflammatory complications in the first week, as occurs in the
majority of preliminary studies, we have included in the
study as IC all those cases notified and diagnosed up to
the 8th postoperative week, based on our clinical
experience.
In week 1, we obtained 9.5% of complications in Pl as
against 0.8% in Ab. When trying to compare our results
with previous clinical trials, we noted that they estimated infection percentages with placebo between
0% and 45.5%.1,2,4,5,6,7,11-16 This wide range of IC,
regardless of the antibiotic, reflects the methodologic

differences among studies. The differences may also be

due to the characteristics of the samples in relation to
control variables such as patient age, degree of third
molar inclusion, surgical procedural differences, and
surgeons experience. Also contributing to the variability in IC frequency may be how the variable response is
defined and measured (eg, clinical complication criterion, patient follow-up guidelines).
It is important to point out that in our trial all patients
were treated with intraoperational irrigation of 0.12%
chlorhexidine. Although we cannot analyze its efficacy,
the results are probably affected.
The rate of early postoperative complications in our
Pl is similar to those of Curran et al1 (8.7%), Bystedt
et al2 (10%), Happonen et al6 (11.1%), Ritzau et al7
(5.2%), Monaco et al11 (5.3%), and Bulut et al13 (6.6%).
When comparing the IC rate in Ab with other studies
that used amoxicillin/clavulanic acid, we noted very
different results. Delilbasi et al15 obtained rates of 8.9%
in the group treated with amoxicillin/clavulanic acid
and 20.9% in the group treated with chlorhexidine.
Their percentages were higher than ours, although we
coincide in the antibiotic efficacy but with a different
guideline (amoxicillin/clavulanic acid 500/125 twice
a day for 5 days). In contrast, Poeschl et al,16 who used
amoxicillin/clavulanic acid 1 g for 5 days, found no difference between the groups with or without antibiotic.
In our study the total of IC cases during an 8-week
follow-up period was 12.9% in Pl and 1.9% in Ab. Of the
5 IC cases in Ab, 3 (60%) became infected/inflamed
after week 1, and of the 30 in Pl, 8 (26.6%) did. These
data suggest patient follow-up should include longer
periods.
There are no prospective studies with an 8-week
follow-up. Piecuch et al,10 in a retrospective study on
a sample of 3,443 low third molars, verified that
complications after week 1 is proportionally higher in
the group treated with antibiotics.
The prescription of antibiotics in the exodontia of low
third molars should be based on criteria of efficacy,

OOOOE
Volume 100, Number 1

safety, convenience, and cost. As to efficacy, our results

demonstrate the antibiotic is efficacious, because IC
frequency is between 3 and 24 times greater if not
prescribed.
The antibiotic as administered is safe. For an adverse
effect it is necessary to treat between 15 and 147 patients
with the antibiotic. In any event, the side effects would
be mild. We must take into account the antibiotic
resistences.
As to convenience, we should take into account that
the frequency of postoperative complication without
antibiotics was 12.9%, which in all cases was resolved
using the rescue antibiotic. Only 1 patient, who was in
the Ab group, of the 490 cases had to be admitted to
hospital.
Therefore, we question the convenience of treating
all patients with antibiotics to prevent IC. Taking as a
base analysis of the risk measurements and the multivariate study, we estimate tht systematic prescription of
preventive antibiotic is not indicated. Age, molar depth,
and its angulation must be borne in mind. Whereas the
possibility of postoperative complications is 10% without antibiotic at 20 years, it exceeds 30% at 40 years, the
risk of IC increasing by 1.08 per year of age.
In the submucosa low third molars and in those in
vertical position, the number of patients to be treated to
avoid 1 being infected is between 8 and infinity. Our
clinical interpretation is that the antibiotic is probably
inefficacious in these 2 situations.
The antibiotic is more efficacious in third molars
partially covered by bone and those in a horizontal
position. As the patients age increases, the possibility
of IC increases. We should consider the possibility of
prescribing an antibiotic as the patient ages. We also
believe a cost/efficiency study is necessary to establish
the most appropriate therapeutical decision taking.
The authors thank the patients and the staff of the Oral and
Maxillofacial Surgery Department, Cruces Hospital. We are
also grateful to Dr JI Pijoan for helping us to analyze the results.
REFERENCES
1. Curran JB, Kennett S, Young AR. An assessment of the use of
prophylactic antibiotics in third molar surgery. Int J Oral Surg
1974;3:1-6.
2. Bystedt H, Von Know L, Nord CE. Effect of tinidazole on
postoperative complications after surgical removal of impacted
mandibular third molars. Scand J Infect Dis Suppl 1981;26:135-9.
3. Kaziro GSN. Metronidazole (Flagyl) and Arnica montana in the
prevention of post-surgical complications, a comparative placebo
controlled clinical trial. Br J Oral Maxillofac Surg 1984;22:42-9.
4. Mitchell DA. A controlled clinical trial of prophylactic tinidazole for chemoprophylaxis in third molar surgery. Br Dent J
1986;160:284-6.
5. Barclay JK. Metronidazole and dry socket: prophylactic use in
mandibular third molar removal complicated by nonacute
pericoronaritis. N Z Dent J 1987;83:71-5.

Arteagoitia et al E17
6. Happonen RP, Backstrom AC, Ylipaavalniemi P. Prophylactic
use of phenoxymethylpenicillin and tinidazole in mandibular
third molar surgery, a comparative placebo, controlled clinical
trial. Br J Oral Maxillofac Surg 1990;28:12-5.
7. Ritzau M, Hillerup S, Branebjerg Ersbol B. Does metronidazole
prevent alveolitis sicca dolorosa? A double-blind, placebocontrolled clinical study. Int J Oral Maxillofac Surg 1992;21:
299-302.
8. Chiapasco M, De Cicco L, Marrone G. Side effects and
complications associated with third molar surgery. Oral Surg
Oral Med Oral Pathol Oral Radiol Endod 1993;76:412-20.
9. Lloyd J, Earl PD. Metronidazole: two or three times daily,
a comparative controlled clinical trial of the efficacy of two
different dosing schedules of metronidazole for chemoprophylaxis following third molar surgery. Br J Oral Maxillofac Surg
1994;32:165-7.
10. Piecuch F, Arzadon J, Lieblich S. Prophylactic antibiotics for
third molar surgery: a supportive opinion. J Oral Maxillofac Surg
1995;53:53-60.
11. Monaco G, Staffolani C, Gatto MR, Checchi L. Antibiotic
therapy in impacted third molar surgery. Eur J Oral Sci 1999;
107:437-41.
12. Sekhar CH, Narayanan V, Baig MF. Role of antimicrobials in
third molar surgery: prospective, double blind, randomized,
placebo-controlled clinical study. Br J Oral Maxillofac Surg
2001;39:134-7.
13. Bulut E, Bulut S, Etikan I, Koseoglu O. The value of routine
antibiotic prophylaxis in mandibular third molar surgery: acutephase protein levels as indicators of infection. J Oral Sci 2001;
43:117-22.
14. Yoshii T, Hamamoto Y, Muraoka S, Furudoi S, Komori T.
Differences in postoperative morbidity rates, including infection
and dry socket, and differences in the healing process after
mandibular third molar surgery in patients receiving 1-day or 3day prophylaxis with lenampicillin. J Infect Chemother 2002;8:
87-93.
15. Delilbasi C, Saracoglu U, Keskin A. Effects of 0.2% chlorhexidine gluconate and amoxicillin plus clavulanic acid on the
prevention of alveolar osteitis following mandibular third molar
extractions. Oral Surg Oral Med Oral Pathol Oral Radiol Endod
2002;94:301-4.
16. Poeschl PW, Eckel D, Poeschl E. Postoperative prophylactic
antibiotic treatment in third molar surgery a necessity? J Oral
Maxillofac Surg 2004;62:3-8.
17. Krekmanov L. Alveolitis after operative removal of third molars
in the mandible. Int J Oral Surg 1981;10:173-9.
18. Lyall JB. Third molar surgery. The effect of primary closure
wound dressing and metronidazole on postoperative recovery.
J R Army Med Corps 1991;137:100-3.
19. Peterson LJ. Antibiotic prophylaxis against wound infections in
oral and maxillofacial surgery. J Oral Maxillofac Surg 1990;48:
617-20.
20. Capuzzi P, Montebugnoli L, Vaccaro MA. Extraction of
impacted third molars. A longitudinal prospective study on
factors that affect postoperative recovery. Oral Surg Oral Med
Oral Pathol Oral Radiol Endod 1994;77:341-3.
21. van der Sanden WJM, Mettes TG, Verdonschot EH, vant Hof
MA, Nienhuijs M, Plasschaert AJM. Interventions for treating
trouble-free impacted wisdom teeth in adults (Protocol).
Cochrane Oral Health Group. Cochrane Database of Systematic
Reviews. Available online: http://www.cochrane.org/cochrane/
revabstr/ORALAbstractIndex.htm
22. de Boer MP, Raghoebar GM, Stegenga B, Schoen PJ, Boering G.
Complications after mandibular third molar extraction. Quintessence Int 1995;26:779-84.
23. Santamara J, Arteagoitia I. Impacted inferior third molar:
predictive model of surgical difficulty. J Dent Res 1997;76:
1128.
24. Muhonen A, Venta I, Ylipaavalniemi P. Factors predisposing to
postoperative complications related to wisdom tooth surgery
among university students. J Am Coll Health 1997;46:39-42.

OOOOE
July 2005

E18 Arteagoitia et al
25. Macgregor AJ, Addy A. Value of penicillin in the prevention
of pain, swelling and trismus following the removal of
ectopic mandibular third molars. Int J Oral Surg 1980;9:
166-72.
26. Worrall SF. Antibiotic prescribing in third molar surgery. Br J
Oral Maxillofac Surg 1998;36:74-5.
27. Al-Kahateeb T, El-Marsafi A, Butler N. The relationship
between the indications for the surgical removal of impacted
tird molars and the incidence of alveolar osteitis. J Oral
Maxillofac Surg 1991;49:141-5.
28. Marx RE. Chronic osteomyelitis of the jaws. Oral Maxillofac
Surg Clin N Am 1991;3:367-81.
29. Hermesch CB, Hilton TJ, Biesbrock AR, Baker RA, Cain-Hamlin
J, McClanahan SF, et al. Perioperative use of 0.12% chlorhexidine
gluconate for the prevention of alveolar ostetis. Efficacy and risk
factor analysis. Oral Surg Oral Med Oral Pathol Oral Radiol Endod
1998;85:381-7.

30. Larsen PE. The effect of a chlorhexidine rinse on the incidence

of alveolar osteitis following the surgical removal of impacted
mandibular third molars. J Oral Maxillofac Surg 1991;49:932-7.
31. Bonine FL. Effect of chlorhexidine rinse on the incidence of dry
socket in impacted mandibular third molar extraction sites. Oral
Surg Oral Med Oral Pathol Oral Radiol Endod 1995;79:154-7.

Reprint requests:
Itziar Arteagoitia
Departamento de Estomatologa
Universidad del Pas Vasco
Barrio Sarriena s/n Leioa 48940
Spain
arteagoitia@infomed.es