Journal of

Oral Rehabilitation

Journal of Oral Rehabilitation 2014 41; 941--956

Review

Prophylactic antibiotic regimen and dental implant failure:

a meta-analysis
B. R. CHRCANOVIC*, T. ALBREKTSSON* & A. WENNERBERG*
dontics, Faculty of Odontology, Malmo University, Malmo,

*Department of Prostho-

Sweden and Department of Biomaterials, Goteborg University, Goteborg,

Sweden

SUMMARY The aim of this meta-analysis was to

investigate whether there are any positive effects
of prophylactic antibiotic regimen on implant
failure rates and post-operative infection when
performing dental implant treatment in healthy
individuals. An electronic search without time or
language restrictions was undertaken in March
2014. Eligibility criteria included clinical human
studies, either randomised or not. The search
strategy resulted in 14 publications. The I2 statistic
was used to express the percentage of the total
variation across studies due to heterogeneity. The
inverse variance method was used with a fixed- or
random-effects
model,
depending
on
the
heterogeneity. The estimates of relative effect were
expressed in risk ratio (RR) with 95% confidence
interval. Six studies were judged to be at high risk
of bias, whereas one study was considered at
moderate risk, and six studies were considered at
low risk of bias. The test for overall effect showed
that the difference between the procedures (use

Introduction
The risk of infection in any dental implant surgical
procedure depends on several patient factors, management procedures during implant placement and the
care of the surgical team in maintaining the basic
principles of surgery and asepsis (1). Although many
surgeons believe that antibiotic coverage for implant
placement is not necessary, they continue to use
them to protect against claims of malpractice (2).
However, the benefits of antibiotic prophylaxis in
2014 John Wiley & Sons Ltd

versus non-use of antibiotics) significantly affected

the implant failure rates (P = 0
0002), with a RR of
0
55 (95% CI 0
410
75). The number needed to
treat (NNT) to prevent one patient having an
implant failure was 50 (95% CI 33100). There were
no apparent significant effects of prophylactic
antibiotics on the occurrence of post-operative
infections in healthy patients receiving implants
(P = 0
520). A sensitivity analysis did not reveal
difference when studies judged as having high risk
of bias were not considered. The results have to be
interpreted with caution due to the presence of
several confounding factors in the included
studies.
KEYWORDS: dental implants, antibiotic prophylaxis,
implant failure rate, post-operative infection,
meta-analysis
Accepted for publication 15 June 2014

healthy patients undergoing routine surgical

procedures, such as the placement of endosseous
dental implants, are still a controversial issue (3),
even for the placement of implants into infected sites
(4). Moreover, there is a general trend of providing
inappropriate antibiotic prescriptions in dental
practice, usually in excess, in both therapeutic and
prophylactic use (5).
Among the possible disadvantages of routine
antibiotic prophylaxis is the risk of drug-related complications (such as hypersensitivity and anaphylaxis),
doi: 10.1111/joor.12211

942

B . R . C H R C A N O V I C et al.
which may be greater than the risk of post-operative
infection occurring with that particular operation, and
the possibility that its use may lead to lax surgical
techniques and actually increase the complication rate
(3). In addition, in view of the growing concerns that
healthcare professionals have regarding the potential
for an increase in new antibiotic-resistant bacteria,
antibiotic prophylaxis guidelines are being reviewed
and modified by healthcare professionals (2).
The indiscriminate use of antibiotics may produce
unwanted adverse effects in patients, cause
unnecessary economic waste and add to the increasing emergence of resistant bacterial strains. The recent
increase in resistant bacteria observed in the healthcare environments represents a major concern that
requires all healthcare professionals to closely examine each procedure in which prophylactic antibiotics
are frequently used to determine their true effectiveness (6). Even though oral surgeons often prescribe
antibiotics routinely following implant surgery, the
usefulness of pre- and post-operative antibiotics in
reducing the failure rates of endosseous implants
remained unclear and unsubstantiated for several
years (7). Most of the conclusions about the use of
prophylactic antibiotic regimen in implant dentistry
come from retrospective analyses, with multiple operators, different antibiotic prescriptions and different
regimens. Despite the widespread use of oral implants
in dentistry, we are still lacking a clear guideline on
if, when and how to prescribe antibiotics for this type
of surgery (8). For approximately the last 15 years,
researchers have been trying to evaluate whether the
use of antibiotics may or may not influence the survival of dental implants. A review (9) on the subject
concluded that a single dose of pre-operative antibiotic therapy may slightly decrease the failure rate of
dental implants. However, as the philosophies of
treatment alter over time, a periodic review of the different concepts is necessary to refine techniques and
eliminate unnecessary procedures. Having said this,
we conducted a meta-analysis of clinical studies to
investigate whether there are any positive effects of
prophylactic antibiotic regimen on implant failure
rates and post-operative infection when performing
dental implant treatment in healthy individuals. This
study presents a more detailed and profound analysis
of the influence of prophylactic antibiotics on the
implant failure rates, previously assessed in a published systematic review (7).

Materials and methods

This study followed the PRISMA Statement guidelines
(10). A review protocol does not exist.
Objective
The purpose of this review was to test the null
hypothesis of no difference in the implant failure rates
and post-operative infection in patients receiving or
not receiving prophylactic antibiotic for dental
implant treatment against the alternative hypothesis
of a difference.
Search strategies
An electronic search without time or language restrictions was undertaken in March 2014 in the following
databases: PubMed, Web of Science and the Cochrane
Oral Health Group Trials Register. The following terms
were used in the search strategy on PubMed:
{Subject AND Adjective}
{Subject: (dental implant OR dental implant failure
OR dental implant survival OR dental implant
success [text words])
AND
Adjective: (antibiotic prophylaxis [text words])}
The following terms were used in the search strategy on Web of Science:
{Subject AND Adjective}
{Subject: (dental implant failure OR dental implant
survival OR dental implant success [title])
AND
Adjective: (antibiotic prophylaxis [title])}
The following terms were used in the search strategy on the Cochrane Oral Health Group Trials Register: (dental implant OR dental implant failure OR
dental implant survival OR dental implant success
AND (antibiotic prophylaxis))
A manual search of dental implant-related journals,
including British Journal of Oral and Maxillofacial
Surgery, Clinical Implant Dentistry and Related Research,
Clinical Oral Implants Research, European Journal of Oral
Implantology, Implant Dentistry, International Journal of Oral and Maxillofacial Implants, International
Journal of Oral and Maxillofacial Surgery, International
Journal of Periodontics and Restorative Dentistry, International Journal of Prosthodontics, Journal of Clinical Peri 2014 John Wiley & Sons Ltd

ANTIBIOTICS AND IMPLANTS: META-ANALYSIS

odontology, Journal of Dental Research, Journal of Oral
Implantology, Journal of Craniofacial Surgery, Journal of
Cranio-Maxillofacial Surgery, and Journal of Maxillofacial
and Oral Surgery, Journal of Oral and Maxillofacial Surgery, Journal of Oral Rehabilitation, Journal of Periodontology and Oral Surgery Oral Medicine Oral Pathology
Oral Radiology and Endodontology, was also performed.
The reference list of the identified studies and the
relevant reviews on the subject were also scanned for
possible additional studies. Moreover, online databases
providing information about clinical trials in progress
were checked (clinicaltrials.gov; www.centerwatch.
com/clinicaltrials; www.clinicalconnection.com).

of random assignments by preventing foreknowledge

of the forthcoming allocations), incomplete outcome
data (clear explanation of withdrawals and exclusions) and blinding (measures to blind study participants and personnel from knowledge of which
intervention a participant received). The incomplete
outcome data will also be considered addressed when
there are no withdrawals and/or exclusions. A study
that met all the criteria mentioned above was classified as having a low risk of bias, and a study that did
not meet one of these criteria was classified as having
a moderate risk of bias. When two or more criteria
were not met, the study was considered to have a
high risk of bias.

Inclusion and exclusion criteria

Eligibility criteria included clinical human studies,
either randomised or not, comparing the implant failure/survival rates in any group of patients receiving
versus not receiving antibiotic prophylactic regimen
for the placement of implants. For this review,
implant failure represents the complete loss of the
implant. Exclusion criteria were case reports, technical
reports, animal studies, in vitro studies and reviews
papers. The use of grafts in the patients was not an
exclusion criterion for the studies.
Study selection
The titles and abstracts of all reports identified through
the electronic searches were read independently by the
three authors. For studies appearing to meet the inclusion criteria, or for which there were insufficient data
in the title and abstract to make a clear decision, the
full report was obtained. Disagreements were resolved
by discussion between the authors. Studies rejected at
this or subsequent stages were recorded, together with
the reason for the exclusion.
Quality assessment
The quality assessment was performed using the recommended approach for assessing risk of bias in studies included in Cochrane reviews (11). The
classification of the risk of bias potential for each
study was based on the four following criteria:
sequence generation (random selection in the population), allocation concealment (steps that must be
taken to secure strict implementation of the schedule
2014 John Wiley & Sons Ltd

Data extraction and meta-analysis

From the studies included in the final analysis, the
following data were extracted (when available): year
of publication, study design, unicentre or multicentre
study, number of patients, patients age, follow-up,
pre-operative antibiotics, post-operative antibiotics,
mouth rinse, failed and placed implants, post-operative infection, implant surface modification, use of
grafting procedures and presence of smokers in the
study. Contact with authors for possible missing data
was performed.
Implant failure and post-operative infection were
the outcome measures evaluated, both dichotomous
outcomes. The statistical unit for implant failure was
the implant and for post-operative infection was the
patient. Whenever outcomes of interest were not
clearly stated, the data were not used for analysis.
The I2 statistic was used to express the percentage of
the total variation across studies due to heterogeneity,
with 25% corresponding to low heterogeneity, 50%
to moderate and 75% to high. The inverse variance
method was used for random-effects or fixed-effects
model. Where statistically significant (P < 0
10) heterogeneity was detected, a random-effects model was
used to assess the significance of treatment effects.
Where no statistically significant heterogeneity was
found, analysis was performed using a fixed-effects
model (12). The estimates of relative effect were
expressed in risk ratio (RR) with 95% confidence
interval (CI). Only if there were studies with similar
comparisons reporting the same outcome measures
was meta-analysis to be attempted. Here, all patients
receiving antibiotics were considered in the group

943

944

B . R . C H R C A N O V I C et al.
use of antibiotics, to increase the sample size of individual trials, as the variation of the regimens (antibiotic type, dosage, time of administration) used by
each study could limit the number of included studies
in case any specific regimen was chosen as an inclusion criterion.
A funnel plot (plot of effect size versus standard
error) will be drawn. Asymmetry of the funnel plot
may indicate publication bias and other biases related
to sample size, although the asymmetry may also represent a true relationship between trial size and effect
size.
The data were analysed using the statistical software Review Manager (version 5.2.8, The Nordic
Cochrane Centre, The Cochrane Collaboration,
Copenhagen, Denmark, 2014).

Results

papers; 13 were cited in more than one research of

terms. The full-text reports of the remaining 15 articles led to the exclusion of 4 because they did not
meet the inclusion criteria: two articles (13, 14) did
not inform of the number of failed implants in each
group, one article (15) did not clearly inform of the
number of implants in each group, and one article
(16) did not evaluate implant failures. Orenstein et al.
(15) informed of the number of implants placed and
the number of failed implants in patients taking preand post-operative antibiotics separately. However, it
was not informed whether, for example, some
patients who did not take antibiotics pre-operatively
made use of antibiotics post-operatively. Thus, it was
not possible to clearly define the groups. Additional
hand searching of the reference lists of selected studies yielded three additional papers. Thus, a total of 14
publications were included in the review.

Literature search

Description of the studies

The study selection process is summarised in Fig. 1.

The search strategy resulted in 311 papers. The three
reviewers independently screened the abstracts for
those articles related to the focus question. The initial
screening of titles and abstracts resulted in 28 full-text

Detailed data of the 14 included studies are listed in

Table 1. Eight RCTs (8, 1723), four CCTs (2, 3, 24,
25) and two retrospective studies (26, 27) were
included in the meta-analysis. In seven studies (8,
1823), both patients and operators/outcome assessors
were blinded to the tested intervention. Nine studies
(8, 1923) had short follow-ups, six of them with a
follow-up up to 4 months, whereas one study (18)
with 5 months and two (17, 25) with a 6-month follow-up.
Four articles (2, 3, 17, 24) did not provide any kind
of information about the patients age, and two (2,
17) did not inform of the number of patients in the
study. All studies with available data of patients age
included only adult patients. Seven studies (1924,
27) did not use post-operative antibiotics in all
patients, and four (2, 3, 8, 17) made use of several
different regimens of antibiotic prophylaxis. One
study only informed that the patients were receiving
antibiotics pre- or immediately after implant surgery
(25). Seven studies (8, 1823) chose amoxicillin as
the only antibiotic used. Four studies (2, 17, 25, 27)
did not mention whether the patients performed antimicrobial mouth rinses. In all other ten studies, the
patients rinsed with antimicrobial solutions at least
pre-operatively.
All articles provided information about the number
of failed implants. However, one recent study (22) did

Fig. 1. Study screening process.

2014 John Wiley & Sons Ltd

CCT (multicentre)

CCT (multicentre)

1998

2000

2000

Gynther
et al. (26)

2014 John Wiley & Sons Ltd

Laskin
et al. (3)
Morris
et al. (24)

Zinser
et al. (27)

RCT (multicentre)

RCT (multicentre)

2010

2011

RA (unicentre)

RCT (multicentre)

2009

2013

RCT (multicentre)

2008

CCT (unicentre)

2008

Esposito
et al. (19)
Anitua
et al. (20)
Esposito
et al. (21)
Caiazzo
et al. (8)

RCT (unicentre)

2008

Abu-Taa
et al. (18)
Alsaadi
et al. (25)

CCT (multicentre)

2004

Morris
et al. (2)

RA (unicentre)

RCT (multicentre)

1997

Dent
et al. (17)

Study design

Published

Authors

Table 1. Detailed data of the included studies

1885 (49; G1)

1886 (48; G2P)
NM (4252*, G1)
NM (43, G2)

3581 (56
1)

224 (NM)

1878 (48; G1)

1976 (48; G2P)
1873 (48)

2782 (60; G1)

2688 (57; G2)
1886 (56
2)

NM

NM

NM

1886 (53; G1)

2281 (58; G2)

NM

Patients age range

(mean) (years)

316 (158, G1;

158, G2P)
105 (52, G1;
53, G2P)
506 (252, G1;
254, G2P)
100 (75 G1;
25 G2)

80 (40, G1;
40, G2)
283 (NM)

NM

702 (387, G1;

315, G2)
663 (NM)

279 (147, G1;

132, G2)

NM

Patients (n)
(number per group)

1, 3, and 6 months after

the reconstructive
procedure; at 1, 3 and 6

1 and 2 weeks
4 months
3 and 10 days
1 and 3 months
1 and 2 weeks
4 months
1, 2, 4, and 8 weeks
3 months

35 years
(depending
on the research
centre)
710 days
2 and 5 months
6 months

36 months

Until uncovering
surgery (maxilla:
6 months;
mandible:
4 months)
16 years
(mean 3; G1)
15 years
(mean 3; G2)
36 months

Follow-up visits
(or range)

1 g phenoxymethylpenicillin,
39/day for 10 days

1 g phenoxymethy
lpenicillin,
1 h before

(1) ampicillin/
sulbactam
3 9 3 g, 1 h before

1 g amoxicillin,
1 h before
pre- or immediately
after
implant surgery
2 g amoxicillin,
1 h before
2 g amoxicillin,
1 h before
2 g amoxicillin,
1 h before
(1) 2 g amoxicillin
(2) 2 g amoxicillin
(3) NP

(1) NP
(2) 1 g amoxicillin,
29/day for 7 days
(3) 1 g amoxicillin,
29/day for 7 days
(1) ampicillin/
sulbactam
3 9 374 mg,

NP

NP

NP

500 mg amoxicillin,
49/day for 2 days
pre- or immediately
after implant surgery

Several different regimens

NP

Several different regimens

Several different
regimens

Several different
regimens

Several different
regimens
Used, but details
were not
informed
Several different
regimens

Post-operative
antibiotics (G1)

Pre-operative
antibiotics (G1)

ANTIBIOTICS AND IMPLANTS: META-ANALYSIS

945

2014

Mouth rinse

NM

Tan
et al. (23)

Authors

Dent

et al. (24)

Morris

et al. (3)

Laskin

et al. (26)

Gynther

(n = 1615)

(chlorhexidine)

were not informed

Used, but details

(chlorhexidine)

following surgery

For 2 weeks

solution)

123/1254 (G2)

77/1663 (G1)

128/1287 (G2)

56/1293 (G1)

34/664 (G2)

49/790 (G1)

48/1193 (G2)

21/1448 (G1)

implants (n)

9
81 (G2)

4
63 (G1)

9
9 (G2)

4
3 (G1)

5
1 (G2)

6
2 (G1)

4
0 (G2)

1
5 (G1)

rate (%)

failure

Implant

329 (249,
G1; 80, G2)

55 (27, G1;
28 G2P)

NM

<0
05

0
395 (mandible)

0
123 (maxilla)

0
001

rate)

(for failure

NM

NM

7 (G2)

9 (G1)

NM

infection

NM

NM

0
906

NM

infection)

operative

(for post-

P-value

Vent Corporation,

(Spectra System, Core-

Turned and HA-coated

Encino, USA)

DBA Paragon Company,

Core-Vent Corporation,

(Spectra System,

Turned and HA-coated

G
oteborg, Sweden)

System, Nobel Biocare,

Turned (Br

anemark

various designs

modification (brand)

NM

NM

NM

NM

Grafting

(1) 2 g amoxicillin,
1 h before (n = 81)
(2) NP (n = 82)
(3) 2 g amoxicillin,
1 h before (n = 86)

3 g amoxicillin,
1 h before

(2) 3 g amoxicillin,
0
5 h before

Pre-operative
antibiotics (G1)

Implant surface

1, 2, 4 and 8 weeks

months after implant

placement; at 1 and 3
months after the start
of prosthetic loading;
and annually thereafter
2 and 7 days
4 months

Follow-up visits
(or range)

operative

Post-

<40 years; n = 31
4060 years;
n = 17
>60 years; n = 7
NM (47
1)

Patients age range

(mean) (years)

P-value

Patients (n)
(number per group)

Failed / placed

RCT (multicentre)

RCT (unicentre)

Study design

(povidone-iodine

Before surgery

2014

Nolan
et al. (22)

et al. (17)

Published

Authors

Table 1. (continued)

the patients

Smokers: 1/3 of

NM

Observations

(1) NP
(2) 2 g amoxicillin
immediately
following
surgery(3) 500 mg
amoxicillin,
39/day
for 3 days

NP

for 6 days
(2) NP

Post-operative
antibiotics (G1)

946

B . R . C H R C A N O V I C et al.

2014 John Wiley & Sons Ltd

 2014 John Wiley & Sons Ltd

et al. (20)

Anitua

et al. (19)

Esposito

et al. (25)

Alsaadi

et al. (18)

Abu-Taa

(chlorhexidine)

Before surgery

(chlorhexidine)

7 days

and 29/day for

Before surgery

NM

days (chlorhexidine)

29/day for 710

Before surgery and

NM

Morris

et al. (2)

Mouth rinse

Authors

Table 1. (continued)

2/53 (G2P)

2/52 (G1)

9/355 (G2P)

2/341 (G1)

6/336 (G2)

7/378 (G1)

5/119 (G2)

0/128 (G1)

17/354 (G2)

43/1175 (G1)

implants (n)

Failed / placed

3
8 (G2P)

3
8 (G1)

2
5 (G2P)

0
6 (G1)

1
79 (G2)

1
85 (G1)

4
2 (G2)

0
0 (G1)

4
8 (G2)

3
7 (G1)

rate (%)

failure

Implant

0
083

difference

significant

no statistically

0
104

1
00

NM

0
21

rate)

(for failure

P-value

6 (G2P)

6 (G1)

2 (G2P)

2 (G1)

NM

4 (G2)

1 (G1)

NM

infection

operative

Post-

0
69

0
960

0
623

NM

NM

NM

infection)

operative

(for post-

P-value

sockets

Vitoria-Gasteiz, Spain)

Biotechnology Institute,

Acid-etched (BTI,

NP

growth factors

(10, G1; 8, G2)

18 smokers

before insertion.

plasma rich in
were provided

humidified with

Implants were

74, G2

sockets: 62, G1;

fresh extraction

implants inserted in

23, G1; 22, G2P),

(>10 cigarettes/day;

but no details

some patients,

36, G1; 28, G2P),

45 heavy smokers

(10 cigarettes/day;

64 light smokers
operators

Performed in

NP

in fresh extraction

implants inserted

left to the individual

Several: Choice was

implants in
diabetics, 9

smokers, 26

95 implants in

were provided

but no details

were smokers,

Some patients

NM

Observations

G
oteborg, Sweden)

NM

NM

NM

Grafting

TiUnite, Nobel Biocare,

Oxidised (Mk III,

NM

Mannheim, Germany)

Dentsply-Friadent,

etched (Ankylos,

Sandblasted and acid-

Encino, USA)

DBA Paragon Company,

modification (brand)

Implant surface

ANTIBIOTICS AND IMPLANTS: META-ANALYSIS

947

et al. (22)

Nolan

et al. (27)

Zinser

et al. (8)

Caiazzo

2 (G2P)

0 (G1)

NM

0 (G2)

0 (G1)

7 (G2P)

3 (G1)

infection

0
49

NM

NM

infection)

operative

(for post-

Biomet3i, Palm Beach

Acid-etched (Osseotite,

NM

NM

operators

left to the individual

Several: Choice was

modification (brand)

Implant surface

Switzerland)

SLA, Straumann, Basel,

Mannheim, Germany;

Dentsply-Friadent,

etched (Ankylos,

sandblasted and acid-

G
oteborg, Sweden),

(TiUnite, Nobel Biocare,

0
0515

0
19

rate)

operative

P-value

(chlorhexidine)

(G2P)

(G1)

12
9 (G2)

8
4 (G1)

6
9 (G2)

0 (G1)

2
7 (G2P)

1
4 (G1)

rate (%)

(for failure

failure

Post-

Gardens, USA), anodised

5/ (G2P)

0/ (G1)

8/62 (G2)

40/478 (G1)

2/29 (G2)

0/119 (G1)

13/483 (G2P)

7/489 (G1)

implants (n)

Failed / placed

P-value

Implant

for 7 days

and 459/day

Before surgery

NM

(chlorhexidine)

15 days

and 29/day for

Before surgery

(chlorhexidine)

7 days

and 29/day for

Before surgery

Esposito

et al. (21)

Mouth rinse

Authors

Table 1. (continued)

60, G1; 76, G2

NP

implants, 470 were

(7, G1; 6 G2)

13 smokers

unknown)

antibiotics is

using vs. not using

implants in smokers

(the number of

inserted in smokers

In the total of 1045

sinus

Only in maxilla

extraction sockets:

inserted in fresh

28, G2P), implants

day; 26, G1;

(>10 cigarettes/

heavy smokers

60, G2P), 54

day; 55, G1;

(10 cigarettes/

115 light smokers

Observations

grafted maxillary

All patients had

NM

Grafting

948

B . R . C H R C A N O V I C et al.

2014 John Wiley & Sons Ltd

 2014 John Wiley & Sons Ltd

Before surgery

Tan
1/80 (G2P)

0/249 (G1)

implants (n)

Failed / placed

NM

0 (G1)
1
25 (G2P)

rate)

(for failure

P-value

rate (%)

failure

Implant

0 (G2P)

2 (G1)

infection

operative

Post-

NM

infection)

operative

(for post-

P-value

of the patients were

smoker (1
5%)

day; 9
1%) or

(<20 cigarettes/

light smoker

8
8%),

(quit 5 years;

smoker

either previous

restorations, 19
4%

Only single-tooth

Observations

Basel, Switzerland)

NP

Grafting

etched (SLA, Straumann,

Sandblasted and acid-

modification (brand)

Implant surface

NM, not mentioned; CCT, controlled clinical trial; RCT, randomised controlled trial; RA, retrospective analysis; G1, group receiving antibiotics; G2, group not receiving antibiotics; G2P, group receiving placebo; NP, not performed.
*The mean age of the patients varied depending on which prophylactic antibiotic regimen they were allocated.

Two or three (depending on the study) different prophylactic antibiotic regimens were adopted for G1.

The total number of patients were 224, and 1045 implants were evaluated. However, the authors informed the number of implant failures comparing the use or non-use of
antibiotics only for 540 implants (unknown number of patients).

The comparison was made between the four different groups in the study (and the difference was not statistically significant), not between patients receiving and not receiving antibiotics.

The total number of implants inserted was 82, when calculated from a table in the article informing the influence of number of implants placed on osseointegration of
implants, but the authors did not inform how many implants were in the antibiotic group and in the placebo group. The authors of the publication were consulted by e-mail.
Only one replied, stating that the list identifying participants was deleted recently so I cant search their records to calculate what you need.

et al. (23)

Mouth rinse

Authors

Table 1. (continued)

ANTIBIOTICS AND IMPLANTS: META-ANALYSIS

949

950

B . R . C H R C A N O V I C et al.
not inform of the total number of implants in each
group. The authors of the publication were consulted
by e-mail and only one replied, stating that the list
identifying participants was deleted recently so I cant
search their records to calculate what you need. Two
studies (3, 17) showed that the use of prophylactic
antibiotics significantly increased the survival rates of
dental implants, and eight studies (2, 1922, 2527)
did not show significant difference, whereas one study
(8) did not observe statistically significant difference
between the four different groups (three receiving
and one not receiving antibiotics), but the difference
between patients receiving and not receiving antibiotics was not informed. Two studies (18, 23) did not
inform of such analyses, but showed an implant failure rate of 4
2% and 1
25%, respectively, in patients
not receiving antibiotics and no failure in patients
receiving antibiotics, whereas one study (24) showed
a failure rate of 4
63% and 9
81% for patients receiving and not receiving antibiotics, respectively. From
the fourteen studies, a total of 8603 implants were
placed [not counting the unknown number of
implants placed in the study of Nolan et al. (22)] in
patients receiving antibiotics, with 304 failures
(3
53%), and 6269 implants were placed in patients
not receiving antibiotics or receiving placebo, with
396 failures (6
32%).
Six articles (2, 3, 17, 24, 25, 27) did not inform of
the incidence of post-operative infections. From the
eight articles (8, 1823, 26) that provided this infor-

mation, it was observed 25 occurrences of infection in

1000 patients receiving antibiotics (2
5%) and 29 episodes of post-operative infection in 770 patients not
receiving antibiotics (3
8%).
Some studies performed grafting procedures in part
of the patients (20) or in all patients (27). Part of the
patients in nine studies (1823, 2527) were smokers.
The insertion of some implants in fresh extraction
sockets was performed in three studies (19, 21, 25),
whereas the humidification of the implants with
plasma rich in growth factors before insertion was
performed in one study (20).
Quality assessment
Each trial was assessed for risk of bias, and the scores
are summarised in Table 2. Seven studies (2, 3, 17, 24
27) were judged to be at high risk of bias, whereas one
study (23) was considered at moderate risk, and six
studies (8, 1822) were considered at low risk of bias.
Meta-analysis
In this study, a random-effects model was used to
evaluate the implant failure, as statistically significant
heterogeneity was found (P = 0
004; I2 = 58%). The
fixed-effects model was used when the post-operative
infection outcomes were evaluated, as statistically significant heterogeneity was not found (P = 0
690;
I2 = 0%).

Table 2. Results of quality assessment

Authors

Published

Sequence
generation
(randomised?)

Dent et al. (17)

Gynther et al. (26)
Laskin et al. (3)
Morris et al. (24)
Morris et al. (2)
Abu-Taa et al. (18)
Alsaadi et al. (25)
Esposito et al. (19)
Anitua et al. (20)
Esposito et al. (21)
Caiazzo et al. (8)
Zinser et al. (27)
Nolan et al. (22)
Tan et al. (23)

1997
1998
2000
2000
2004
2008
2008
2008
2009
2010
2011
2013
2014
2014

Yes
No
No
No
No
Yes
No
Yes
Yes
Yes
Yes
No
Yes
Yes

Allocation
concealment

Incomplete
outcome data
addressed

Blinding

Estimated
potential
risk of bias

Inadequate
Inadequate
Inadequate
Inadequate
Inadequate
Adequate
Inadequate
Adequate
Adequate
Adequate
Adequate*
Inadequate
Adequate
Adequate

No
No
No
No
No
Yes
Yes
Yes
Yes
Yes
Yes
No
Yes
No

No
No
No
No
No
Yes
No
Yes
Yes
Yes
Yes*
No
Yes
Yes

High
High
High
High
High
Low
High
Low
Low
Low
Low
High
Low
Moderate

*Unpublished information obtained by personal communication with one of the authors.

2014 John Wiley & Sons Ltd

ANTIBIOTICS AND IMPLANTS: META-ANALYSIS

The test for overall effect showed that the difference
between the procedures (use versus non-use of antibiotics) significantly affected the implant failure rates
(P = 0
0002; Fig. 2). A RR of 0
55 (95% CI 0
410
75)
for the prophylactic use of antibiotic implies that
implant failures in patients receiving prophylactic
antibiotic treatment were 0
55 times likely to happen
than implant failures in patients not receiving prophylactic antibiotic treatment or receiving placebo. Thus,
the relative risk reduction (RRR) is 45%. In other
words, the prophylactic use of antibiotic decreases the
risk of implant failure by 45%. The number needed
to treat (NNT) to prevent one patient having an

implant failure was 50 (95% CI 33100), based on an

implant failure rate of 6
32% in patients not receiving
antibiotics.
On the other side, the meta-analysis showed that
there were no apparent significant effects of prophylactic antibiotics on the occurrence of post-operative
infections in healthy patients receiving implants (RR
0
84, 95% CI 0
491
44; P = 0
520; Fig. 3).
As the RR could differ depending on the risk of bias
of the studies, a sensitivity analysis was performed.
The RR was examined when studies judged as having
high risk of bias were not considered. For the event
implant failure, six studies (8, 1821, 23) with

Fig. 2. Forest plot of comparison of use versus non-use of antibiotics/use of placebo for the event implant failure.

Fig. 3. Forest plot of comparison of use versus non-use of antibiotics/use of placebo for the event post-operative infection.
2014 John Wiley & Sons Ltd

951

952

B . R . C H R C A N O V I C et al.
moderate and low risk of bias were considered, still
resulting in a statistically significant difference
(P = 0
003; RR 0
37, 95% CI 0
190
72; heterogeneity: I2 = 6%, P = 0
38, fixed-effects model; Fig. 4). For
the event post-operative infection, seven studies (8,
1823) with moderate and low risk of bias were considered, also not resulting in a statistically significant
difference (P = 0
33; RR 0
72, 95% CI 0
381
39; heterogeneity: I2 = 0%, P = 0
66, fixed-effects model;
Fig. 5).

small-study effects. Small-study effects may be due to

reasons other than publication bias (28). Moreover,
the studies causing asymmetry (8, 18, 23) have a
good methodological quality (RCT, double blinded). It
can be observed that the effect estimates in the larger
studies were close to the true intervention RR of 0
55.
The funnel plot did not show asymmetry when the
studies reporting the outcome post-operative infection (Fig. 7) were analysed.

Discussion
Publication bias
The funnel plot showed asymmetry when the studies
reporting the outcome implant failure (Fig. 6) were
analysed. However, there is a tendency for the intervention effects estimated in smaller studies to differ
from those estimated in larger studies, which is called

In a meta-analysis, homogeneity implies a mathematical compatibility between the results of each individual trial. Potential biases are likely to be greater for
non-randomised studies compared with RCTs, so
results should always be interpreted with caution
when they are included in reviews and meta-analyses

Fig. 4. Forest plot of comparison of use versus non-use of antibiotics/use of placebo for the event implant failure, excluding studies
judged as having high risk of bias.

Fig. 5. Forest plot of comparison of use versus non-use of antibiotics/use of placebo for the event post-operative infection, excluding
studies judged as having high risk of bias.
2014 John Wiley & Sons Ltd

ANTIBIOTICS AND IMPLANTS: META-ANALYSIS

Fig. 6. Funnel plot for the studies

reporting
the
outcome
event
implant failure.

Fig. 7. Funnel plot for the studies reporting the outcome event
post-operative infection.

(11). However, narrowing the inclusion criteria

increases homogeneity but also excludes the results of
more trials and thus risks the exclusion of significant
data (29). This was the reason to include non-randomised studies in the present meta-analysis. The issue
is important because meta-analyses are frequently
conducted on a limited number of RCTs. In metaanalyses such as these, adding more information from
observational studies may aid in clinical reasoning
and establish a more solid foundation for causal inferences (29).
Six of the included studies (2, 8, 20, 2527)
appeared to be underpowered to detect a clinically
significant difference, and four others showed trends
favouring antibiotics (18, 19, 21, 22), but without a
2014 John Wiley & Sons Ltd

statistically significant difference. The previous review

of Sharaf et al. (9) evaluated eight studies separately
and came to the conclusion that a single dose of preoperative antibiotic therapy may slightly decrease the
failure rate of dental implants. However, the study
did not evaluate the overall effect. In the present
review, a statistically and clinically significant
difference in implant failures was found after the
meta-analysis of the included studies, stressing the
importance of meta-analyses to increase sample size
of individual trials to reach more precise estimates of
the effects of interventions. The meta-analysis suggested that the use of antibiotics in healthy patients
significantly decreases early implant failures. With a
NNT equal to 50, it is expected that one additional (or
less) person will incur an implant failure for every 50
participants receiving prophylactic antibiotics rather
than not receiving. One the other hand, there were
no apparent significant effects of antibiotics on the
occurrence of post-operative infections. This may be
explained by the fact that asymptomatic infections
without apparent clinical signs may determine the
loss of some implants (30). Thus, the actual occurrence of post-operative infection can be difficult to
verify clinically. It is important to stress that the infection rates remained low even without the use of antibiotics. A sensitivity analysis did not reveal difference
when studies judged as having high risk of bias were
not considered in the meta-analysis.
In the present meta-analysis, the statistical unit of
analysis for implant failure was the implant. It would

953

954

B . R . C H R C A N O V I C et al.
be technically more correct to adjust for the effect of
clustered, correlated observations; however, it is a
challenging analytic method and the implant survival
is so high that failing to adjust for clustered, correlated observations would have little effect on the estimate and deviation of survival (31).
The drawback of some studies is the fact the regimens used by each study varied somewhat by antibiotic type, dosage and time of administration. Another
drawback found in nine studies (8, 1723, 25) is the
fact that the patients were followed for only few
months (36 months). Thus, only early failures could
be assessed. A longer follow-up period can lead to an
increase in the failure rate, especially if it extended
beyond functional loading, because other prosthetic
factors can influence implant failure from that point
onward. The efficacy of prophylactic antibiotic use
will have different implications when implant failure
is reported at prosthetic rehabilitation or several years
after loading (9). Moreover, the results found in the
studies differed from each other, and this difference
could be due to factors such as differences in the
patients included in the study or the clinicians placing
and restoring the implants.
The use of grafting in some studies (20, 27) is a
confounding factor, as well as the presence of smokers
among the patients in several trials (1823, 2527),
the insertion of some implants in fresh extraction
sockets (19, 21, 25), the humidification of the
implants with plasma rich in growth factors before
insertion (20), different prosthetic configurations and
the insertion of implants from different brands and
surface treatments. Titanium with different surface
modifications shows a wide range of chemical, physical properties and surface topographies or morphologies, depending on how they are prepared and handled
(32, 33), and it is not clear whether, in general, one
surface modification is better than another (34).
Moreover, in some studies (8, 1921, 23), the patients
received adequate oral hygiene instructions and were
treated periodontally prior to the implant installation,
which constituted a second intervention that might
have masked the effect of antibiotic use on the
implant failure rate.
It is also important to mention another variable, the
surgeons surgical experience, which was evaluated
together with the use of prophylactic antibiotics in one
study (3). When implant survival rates were compared
according to surgeons previous implant surgery experi-

ence and use or non-use of pre-operative antibiotics,

those surgeons with greater than 50 implant placements prior to the study had a slightly higher implant
survival rate (2
9%) when pre-operative antibiotics
were used. There was an even greater increase in survival rate (7
3%) when less experienced surgeons (<50
previous implant placements) used pre-operative antibiotic coverage, which would suggest that as surgical
skill increases with experience, there is less reliance on
antibiotics for survival of implants. However, this
should not be interpreted that antibiotics will make up
for lax and poor surgical techniques.
This paper illustrates the utility meta-analysis in
evaluating medical technologies described in nonrandomised studies, if proper attention is given to
biases in those studies.
The results of the present study have to be interpreted with caution because of its limitations. First of
all, all confounding factors may have affected the
long-term outcomes and not just the use or not use
of prophylactic antibiotics, and the impact of these
variables on the implant survival rate and post-operative infection is difficult to estimate if these factors are
not identified separately between the two different
procedures to perform a meta-regression analysis. The
lack of control of the confounding factors limited the
potential to draw robust conclusions. Second, some of
the included studies had a retrospective design, and
the nature of a retrospective study inherently results
in flaws. These problems were manifested by the gaps
in information and incomplete records. Furthermore,
all data rely on the accuracy of the original examination and documentation. Items may have been
excluded in the initial examination or not recorded in
the medical chart (35, 36).
Although the exact mechanism by which prophylactic antibiotics has a significant effect on implant
survival is not known, it may be that a more aseptic
local environment during the time of implant placement and in the immediate perioperative period
results in improved healing and, ultimately, a better
state of osseointegration (3).
The authors believe that new research efforts
should be concentrated in a comparison between the
use versus non-use of prophylactic antibiotics in a
large double-blinded randomised controlled trial
where the patients are categorised according to possible confounding factors. Another idea would be the
comparison of the use versus non-use of prophylactic
2014 John Wiley & Sons Ltd

ANTIBIOTICS AND IMPLANTS: META-ANALYSIS

antibiotics in prolonged surgical procedures for the
insertion of multiple dental implants or in surgical
procedures for the insertion of a single implant.

Conclusion
There is evidence suggesting that a prophylactic antibiotic regimen significantly reduces failures of dental
implants placed in ordinary conditions. However,
there were no apparent significant effects of prophylactic antibiotics on the occurrence of post-operative
infections in healthy patients receiving implants. A
sensitivity analysis did not reveal difference when
studies judged as having high risk of bias were not
considered in the meta-analysis. The results of the
present meta-analysis have to be interpreted with
caution due to the presence of several confounding
factors in the studies.

Acknowledgments
The authors would like to thank Dr. Alfonso Caiazzo,
who provided us some missing information about his
study.

Ethical approval
None.

Source of funding
This work was supported by CNPq, Conselho Nacional
de Desenvolvimento Cientfico e Tecnol
ogico Brazil.

Conflict of interests
There are no conflict of interests.

References
1. Peterson LJ. Antibiotic prophylaxis against wound infection
in oral and maxillofacial surgery. J Oral Maxillofac Surg.
1990;48:617620.
2. Morris HF, Ochi S, Plezia R, Gilbert H, Dent CD, Pikulski J,
et al. AICRG, Part III: the influence of antibiotic use on the
survival of a new implant design. J Oral Implantol.
2004;30:144151.
3. Laskin DM, Dent CD, Morris HF, Ochi S, Olson JW. The influence of preoperative antibiotics on success of endosseous
implants at 36 months. Ann Periodontol. 2000;5:166174.

2014 John Wiley & Sons Ltd

4. Chrcanovic BR, Martins MD, Wennerberg A. Immediate

placement of implants into infected sites: a systematic
review. Clin Implant Dent Relat Res. 2013. doi:10.1111/cid.
12098. [Epub ahead of print].
5. Isla A, Canut A, Gasc
on AR, Labora A, Ardanza-Trevijano B,
Solins MA, et al. Pharmacokinetic/pharmacodynamic evaluation of antimicrobial treatments of orofacial odontogenic
infections. Clin Pharmacokinet. 2005;44:305316.
6. Smith A, Bragg J. An update on antimicrobial chemotherapy: antimicrobial resistance and the oral cavity. Dent
Update. 1998;25:230234.
7. Chrcanovic BR, Albrektsson T, Wennerberg A. Reasons for
failures of oral implants. J Oral Rehabil. 2014;41:443476.
8. Caiazzo A, Casavecchia P, Barone A, Brugnami F. A pilot study
to determine the effectiveness of different amoxicillin regimens in implant surgery. J Oral Implantol. 2011;37:691696.
9. Sharaf B, Jandali-Rifai M, Susarla SM, Dodson TB. Do perioperative antibiotics decrease implant failure? J Oral Maxillofac Surg. 2011;69:23452350.
10. Moher D, Liberati A, Tetzlaff J, Altman DG, PRISMA Group.
Preferred reporting items for systematic reviews and metaanalyses: the PRISMA statement. Ann Intern Med.
2009;151:W64.
11. Higgins JPT, Green S (editors). Cochrane handbook for systematic reviews of interventions version 5.1.0. [updated
March 2011]. The Cochrane Collaboration; 2011. www.
cochrane-handbook.org. Accessed on 03/03/2014.
12. Egger M, Smith GD. Principles of and procedures for systematic reviews. In: Egger M, Smith GD, Altman DG, eds.
Systematic reviews in health care: meta-analysis in context.
London: BMJ books; 2003:2342.
13. Zupnik J, Kim SW, Ravens D, Karimbux N, Guze K. Factors
associated with dental implant survival: a 4-year retrospective analysis. J Periodontol. 2011;82:13901395.
14. Alissa R, Oliver RJ. Influence of prognostic risk indicators
on osseointegrated dental implant failure: a matched casecontrol analysis. J Oral Implantol. 2012;38:5161.
15. Orenstein IH, Petrazzuolo V, Morris HF, Ochi S. Variables
affecting survival of single-tooth hydroxyapatite-coated
implants in anterior maxillae at 3 years. Ann Periodontol.
2000;5:6878.
16. Khoury SB, Thomas L, Walters JD, Sheridan JF, Leblebicioglu B. Early wound healing following one-stage dental
implant placement with and without antibiotic prophylaxis:
a pilot study. J Periodontol. 2008;79:19041912.
17. Dent CD, Olson JW, Farish SE, Bellome J, Casino AJ, Morris HF, et al. The influence of preoperative antibiotics on
success of endosseous implants up to and including stage II
surgery: a study of 2,641 implants. J Oral Maxillofac Surg.
1997;55:1924.
18. Abu-Taa M, Quirynen M, Teughels W, van Steenberghe D.
Asepsis during periodontal surgery involving oral implants
and the usefulness of peri-operative antibiotics: a prospective, randomized, controlled clinical trial. J Clin Periodontol.
2008;35:5863.
19. Esposito M, Cannizzaro G, Bozzoli P, Consolo U, Felice P,
Ferri V, et al. Efficacy of prophylactic antibiotics for dental

955

956

B . R . C H R C A N O V I C et al.

20.

21.

22.

23.

24.

25.

26.

27.

28.

implants: a multicentre placebo-controlled randomised clinical trial. Eur J Oral Implantol. 2008;1:2331.
Anitua E, Aguirre JJ, Gorosabel A, Barrio P, Errazquin JM,
Rom
an P, et al. A multicentre placebo-controlled randomised
clinical trial of antibiotic prophylaxis for placement of single
dental implants. Eur J Oral Implantol. 2009;2:283292.
Esposito M, Cannizzaro G, Bozzoli P, Checchi L, Ferri V,
Landriani S, et al. Effectiveness of prophylactic antibiotics at
placement of dental implants: a pragmatic multicentre placebo-controlled randomised clinical trial. Eur J Oral Implantol. 2010;3:135143.
Nolan R, Kemmoona M, Polyzois I, Claffey N. The influence
of prophylactic antibiotic administration on post-operative
morbidity in dental implant surgery. A prospective double
blind randomized controlled clinical trial. Clin Oral Implants
Res. 2014;25:252259.
Tan WC, Ong M, Han J, Mattheos N, Pjetursson BE, Tsai
AY, et al. Effect of systemic antibiotics on clinical and
patient-reported outcomes of implant therapy a multicenter randomized controlled clinical trial. Clin Oral Implants
Res. 2014;25:185193.
Morris HF, Ochi S, Winkler S. Implant survival in patients
with type 2 diabetes: placement to 36 months. Ann Periodontol. 2000;5:157165.
Alsaadi G, Quirynen M, Michiles K, Teughels W, Komarek
A, van Steenberghe D. Impact of local and systemic factors
on the incidence of failures up to abutment connection with
modified surface oral implants. J Clin Periodontol.
2008;35:5157.
Gynther GW, K
ondell P

A, Moberg LE. Dental implant
installation without antibiotic prophylaxis. Oral Surg Oral
Med Oral Pathol Oral Radiol Endod. 1998;85:509511.
Zinser MJ, Randelzhofer P, Kuiper L, Z
oller JE, De Lange
GL. The predictors of implant failure after maxillary sinus
floor augmentation and reconstruction: a retrospective study
of 1045 consecutive implants. Oral Surg Oral Med Oral
Pathol Oral Radiol. 2013;115:571582.
Sterne JA, Gavaghan D, Egger M. Publication and related
bias in meta-analysis: power of statistical tests and preva-

29.

30.

31.

32.

33.

34.

35.

36.

lence in the literature. J Clin Epidemiol. 2000;53:1119

1129.
Shrier I, Boivin JF, Steele RJ, Platt RW, Furlan A, Kakuma
R, et al. Should meta-analyses of interventions include
observational studies in addition to randomized controlled
trials? A critical examination of underlying principles. Am J
Epidemiol. 2007;15:12031209.
Esposito M, Thomsen P, Ericson LE, Lekholm U. Histopathologic observations on early oral implant failures. Int J Oral
Maxillofac Implants. 1999;14:798810.
Chuang SK, Tian L, Wei LJ, Dodson TB. Kaplan-Meier
analysis of dental implant survival: a strategy for estimating
survival with clustered observations. J Dent Res.
2001;80:20162020.
Chrcanovic BR, Martins MD. Study of the influence of acid
etching treatments on the superficial characteristics of Ti.
Mater Res. 2014;17:373380.
Chrcanovic BR, Le~ao NLC, Martins MD. Influence of different acid etchings on the superficial characteristics of Ti
sandblasted with Al2O3. Mater Res. 2013;16:10061014.
Wennerberg A, Albrektsson T. On implant surfaces: a
review of current knowledge and opinions. Int J Oral Maxillofac Implants. 2010;25:6374.
Chrcanovic BR, Abreu MH, Freire-Maia B, Souza LN. Facial
fractures in children and adolescents: a retrospective study
of 3 years in a hospital in Belo Horizonte, Brazil. Dent Traumatol. 2010;26:262270.
Chrcanovic BR, Abreu MH, Freire-Maia B, Souza LN. 1,454
mandibular fractures: a 3-year study in a hospital in Belo
Horizonte, Brazil. J Cranio-Maxillofac Surg. 2012;40:116
123.

Correspondence: Bruno Ramos Chrcanovic, Department of Prosthodontics, Faculty of Odontology, Malm

o University, Carl Gustafs v
ag
34, SE-205 06, Malm
o, Sweden.
E-mails: bruno.chrcanovic@mah.se; brunochrcanovic@hotmail.com

2014 John Wiley & Sons Ltd