Vol. 119 No.

4 April 2015

Associations between bacteremia from oral sources and

distant-site infections: tooth brushing versus single tooth
extraction
Farah K. Bahrani Mougeot, PhD,a Sabrina E. Saunders, BS,b Michael T. Brennan, DDS, MHS,c and
Peter B. Lockhart, DDSd

Objectives. To determine the impact of antibiotic prophylaxis (AP) on the incidence of bacteremia caused by oral bacterial
species associated with infective endocarditis (IE) and prosthetic joint infections (PJIs) and to compare the incidence of
following tooth brushing versus single tooth extraction.
Study Design. Bacterial species in blood following single tooth extraction, with or without AP, and tooth brushing1 were
compared with IE- and PJI-associated bacteria reported in the literature.
Results. Of the 98 bacterial species identified in blood following single tooth extraction and tooth brushing, 321 and 12 were
species were associated with IE and PJI, respectively. AP decreased the frequency of IE- and PJI-causing oral bacterial species
in blood; however, single tooth extraction versus brushing resulted in bacteremia with IE- and PJI-causing species with similar
frequencies: 65% versus 56% for IE, and 31% versus 28% for PJI.
Conclusions. Although AP significantly decreased the incidence of bacteremia, the similarity between the incidence of
bacteremia following brushing and extraction undermines AP as an effective strategy for the prevention of these distant-site
infections. (Oral Surg Oral Med Oral Pathol Oral Radiol 2015;119:430-435)

Over 700 bacterial species have been identified in the
human oral cavity.2 Some of these species cause local
infections, such as periodontal disease (PD). Far more
species, however, have been identified in bacteremia;
after they enter the circulation, some have the potential
to cause distant-site infections (DSI), such as infective
endocarditis (IE) or prosthetic joint infections (PJIs).1,3-5
Bacteremia is common following invasive dental pro-
cedures as well as routine dental manipulations, such as
tooth brushing.1,6 As a result, there has been a long-
standing focus on preventing or reducing bacteremia with
the use of antibiotic prophylaxis (AP) before dental pro-
cedures in patients at risk for DSIs.7-9 A previous study by
our group reported the incidence, duration, and nature of
bacteremia following tooth brushing versus single tooth
extraction, and the impact of AP on these surrogate
measures of risk for DSIs.1 The primary purpose of the
present study was to determine the impact of AP on
bacteremia caused by all oral bacterial species
following tooth brushing and single tooth extraction,1
This work was supported by Carolinas HealthCare System (CHS)
Research Fund.
a
Director of Oral Medicine Laboratory, Department of Oral Medicine,
Carolinas Medical Center, Charlotte, NC, USA.
b
Cannon Research Summer Scholar, Department of Oral Medicine,
Carolinas Medical Center, Charlotte, NC, USA.
c
Chairman, Department of Oral Medicine, Carolinas Medical Center,
Charlotte, NC, USA.
d
Professor, Department of Oral Medicine, Carolinas Medical Center,
Charlotte, NC, USA.
Received for publication Nov 3, 2014; returned for revision Dec 30,
2014; accepted for publication Jan 15, 2015.
Ó 2015 Elsevier Inc. All rights reserved.
2212-4403/$ - see front matter
http://dx.doi.org/10.1016/j.oooo.2015.01.009
as well as specific species reported to cause IE or PJI.
Given that the most prevalent periodontal pathogens are
anaerobes that thrive deep in the subgingival biofilm,3 it
is assumed that anaerobic pathogens are less likely to
be disturbed by tooth brushing than by more invasive
procedures, such as tooth extraction. Another aim of
this study was to determine if the bacterial species
identified in blood after tooth brushing are different
from those found after a single tooth extraction. These
bacterial species were also compared on their reported
causality for IE and PJI.
METHODS
The detailed methods for subject recruitment and
randomization to the three arms of this study have been
published elsewhere.1 Briefly, patients enrolled in this
study presented to our hospital-based dental clinic for
dental extraction. Two hundred and ninety patients were
randomized to one of three study arms: Tooth brushing
(98 patients), single tooth extraction with AP (96 pa-
tients), or single tooth extraction with an identical placebo
(96 patients). We collected demographic, baseline, and
Statement of Clinical Relevance
Given the similar incidence of bacteremia caused by
species associated with infective endocarditis and
prosthetic joint infections following tooth brushing
and single tooth extraction, poor oral hygiene may
put patients at greater risk compared with dental
office procedures.

430
OOOO
Volume 119, Number 4
extensive oral disease and hygiene data parameters; the
mean age of patients was 40 years; and 58% of patients
were male and 73% African American. Patients ran-
domized to the extraction groups were given 2.0 grams of
amoxicillin or an identical placebo 1 hour before the
procedure. Venous blood draws were obtained at six time
points before, during, and after the procedure. Bacterial
species cultured from blood were then identified by using
the 16 S rRNA gene sequencing method.
Literature review and analysis of bacterial species
implicated in DSIs
In a previous review of the literature, we had developed a
list of 170 bacterial species that have been identified in
blood cultures after dental procedures.1 This list was
compared with another compiled list of 275 bacterial
species reported to cause IE. The bacterial species that
were common in both lists were then compared with the
98 oral bacterial species identified in blood in our prior
study.1 In the present study, we extended our review of
the literature to the past 25 years to find bacterial species
reported to cause PJI by using the key words “prosthetic
joint infections” and “bacterial species OR bacterium OR
bacteremia.” A specific PubMed search was conducted
by using the specific name of each bacterial species
along with the key words “prosthetic joint infections” to
determine whether that particular species was associated
with PJI. The list of PJI bacterial species compiled from
the published literature was used to determine which of
the 98 oral bacterial species identified in blood in our
prior study1 are associated with PJI.
First, the frequency of bacterial species present in
blood in either extraction group (i.e., with and without
AP) was calculated to find the impact of amoxicillin on
the bacterial species reported to cause IE and PJI or on
all of the 98 oral bacterial species identified in our prior
study. Second, the frequency of bacterial species in
blood in the extractioneplacebo group was compared
with that of the brushing group to determine the impact
of these two dental manipulations on bacteremia caused
by species associated with IE and PJI.
Statistical analysis
A subject was determined to be positive for bacteremia if
any of the bacterial species identified from blood cultures
matched the species reported in the published literature
to cause IE or PJI. The incidence of bacterial species in
each study arm was compared using chi-square tests.
Statistical significance (P ¼ .05) was used in all cases.
RESULTS
As reported previously, the overall incidence of any oral
bacterial species detected from blood cultures was 32%,
ORIGINAL ARTICLE
Mougeot et al. 431
56%, and 80% for tooth brushing, extractione
amoxicillin, and extractioneplacebo, respectively.1 Of
the 98 identified bacterial species, 32, and 12 are
associated with IE1,10 and PJI, respectively (Tables I
and II). The differences in the incidences of species
associated with IE and PJI, as well as all oral species, were
statistically significant for both the extractioneplacebo
and the extractioneamoxicillin groups and the extrac-
tioneplacebo and tooth brushing groups (P < .0001).
Impact of amoxicillin on bacterial species detected
in blood and associated with IE and PJI
AP with amoxicillin decreased the overall incidence of
bacteremia after a single tooth extraction by 61% and the
incidence of IE bacterial species by 68%, which is in
keeping with the findings from our previous study.4 The IE
bacterial species most significantly decreased in incidence
by amoxicillin prophylaxis were Streptococcus spp.,
Fusobacterium nucleatum, and Peptostreptococcus
micros (Table IIIA). However, F. nucleatum and
P. micros are not common causes of IE.11-13
The bacterial species identified in blood to be associ-
ated with PJI were Actinomyces naeslundii, Gran-
ulicatella adiacens, Haemophilus aphrophilus,
Peptostreptococcus micros, Prevotella melaninogenica,
Propionibacterium acnes, Staphylococcus epidermidis,
Streptococcus mitis, Streptococcus mutans, Streptococcus
oralis, Streptococcus sanguinis, and Veillonella dispar/
parvulla.1 Overall, AP decreased the frequency of these
12 bacterial species by 81% (P < .0001) (Table IIIB).
Impact of dental manipulation on bacterial species
detected in blood
We compared the effects of the two dental manipulations
and found a higher incidence of bacteremia from species
associated with IE, PJI, and other oral bacterial species in
the extractioneplacebo group than in the brushing group
(i.e., 219 vs. 43). However, we found that the relative
frequency of IE and PJI bacterial species in the extrac-
tioneplacebo group was not significantly different from
the brushing group (Table IV). That is, of the 219 total
instances in which oral bacterial species were detected in
the extractioneplacebo group, IE bacterial species were
identified 144 times (65%) and those associated with PJI
were detected 68 times (31%). Also, of the 43 incidences
of bacteremia in the brushing group, 24 (56%) were
bacterial species associated with IE and 12 (28%) with PJI.
DISCUSSION
In this study, we report for the first time the effects of
dental manipulations (tooth brushing vs. tooth extrac-
tion) on bacteremia as a surrogate risk for DSIs (e.g., IE
and PJI) and the impact of amoxicillin on PJI species.
ORAL MEDICINE OOOO
432 Mougeot et al. April 2015

Table I. Comparison of infective endocarditis (IE) and prosthetic joint infections (PJIs) bacterial species detected in
blood following tooth brushing versus single tooth extraction with or without antibiotic prophylaxis
IE species detected in blood PJI species detected in blood
Single tooth Single tooth
Single tooth extraction Single tooth extraction Tooth extraction with extraction with Tooth
Bacterial species with amoxicillin with placebo brushing amoxicillin placebo brushing
Acinetobacter lwofii/calcoaceticus O
Actinobacillus actinomycetemcomitans O O
Actinomyces meyeri/odontolyticus O O
Actinomyces naeslundii O
Bifidobacterium longum/infantis O
Capnocytophaga oral clone X089 O
Eikonella corrodens O O
Fusobacterium nucleatum O O
Granulicatella adiacens O O
Haemophilus aphrophilus O O O O
Lactobacillus rhamnosus/casei O
Lactobacillus salivarius O O
Neisseria elongate O O
Neisseria flavescens O
Neisseria mucosa/sicca O O
Peptostreptococcus micros O O O O O
Prevotella buccae O
Prevotella denticola O
Prevotella melaninogenica O O O
Prevotella oralis O
Propionibacterium acnes O O O
Staphylococcus epidermidis O O O O O
Staphylococcus warneri O O
Stenotrophomonas maltophilia O
Streptococcus anginosus O O O
Streptococcus constellatus O O
Streptococcus cristatus O O O
Streptococcus gordonii O O
Streptococcus intermedius O O O
Streptococcus mitis/pneumoniae O O O
Streptococcus mutans O
Streptococcus oralis O
Streptococcus salivarius O
Streptococcus sanguinis O O O O O
Veillonella dispar/parvula O O O O O

Table II. Comparison of the incidences of bacterial species associated with infective endocarditis (IE), prosthetic
joint infections (PJIs), and all oral bacterial species detected in blood following tooth brushing and single tooth
extractions with amoxicillin or placebo
Incidence
Bacterial species Tooth brushing Extractioneamoxicillin Extractioneplacebo Total species
IE (P < .0001) *
23% 33% 60% N ¼ 32
PJI (P < .0001) 12% 14% 52% N ¼ 12
All oral species (P < .0001)* 32% 56% 80% N ¼ 98
*Results previously reported.1

Infective endocarditis would expect that the proportions of bacterial species

IE-associated bacteria are often found in bacteremia
resulting from invasive dental procedures, such as tooth
extraction. Since tooth brushing disrupts only the
(shallow) gingival pocket and tooth extraction disrupts
both gingival and periodontal (deeper) pockets, one
found in blood after these two manipulations would be
different. However, our data suggest that these two very
different manipulations, in terms of perceived inva-
siveness, disrupted similar percentages of bacterial
species that could cause IE. Although the load, or
OOOO ORIGINAL ARTICLE
Volume 119, Number 4 Mougeot et al. 433

Table III. Impact of amoxicillin on bacterial species associated with distant-site infections detected in blood
following dental extraction
A. Most prevalent infective endocarditisecausing bacterial species
Infective endocarditis (IE)ecausing Counts Counts Percentage
bacterial species (extractioneplacebo) (extractioneamoxicillin) change P value
Streptococcus mutans 10 0 100% .0015
Fusobacterium nucleatum 10 1 90% .0096
Peptostreptococcus micros 13 2 85% .0054
Streptococcus intermedius 12 2 83% .0183
Streptococcus mitis/pneumoniae 18 4 78% .0138
Streptococcus anginosus 19 7 63% .0114
B. Prosthetic joint infectionecausing bacterial species
Counts Counts Percentage
Bacterial species (extractioneplacebo) (extractioneamoxicillin) change P value
Facultative anaerobes (Actinomyces 8 1 88% .0168
naeslundi, Granulicatella adiacens,
Propionibacterium acnes,
Staphylococcus epidermidis)
Viridans streptococci 37 6 84% <.0001
Aerobes (Haemophilus aphrophilus) 4 1 75% .3122
Strict anaerobes (Peptostreptococcus 19 6 68% .0088
micros, Prevotella melaninogenica,
Veillonella dispar/parvulla)
Total 68 14 79% <.0001

Table IV. Incidence of total bacterial species associ-
ated with infective endocarditis and prosthetic joint
infections detected in blood following tooth brushing or
single tooth extraction with placebo
Total incidence
Bacterial species Randomization (N)
Infective endocarditis Tooth brushing (43)
Extractioneplacebo (219)
Prosthetic joint infection Tooth brushing (43)
Extractioneplacebo (219)
magnitude, of bacterial species in blood and the host
immune response likely play important roles in the
outcomes of bacteremia following dental office pro-
cedures, the much higher frequency of tooth brushing
may put individuals at greater risk for DSIs annually
compared with dental office procedures, which are far
less frequent.
The most prevalent bacteria found in blood following
both tooth extraction and brushing were Streptococcus
spp.,1 and most of the detected IE-causing bacterial
species were Streptococcus spp. as well (see
Table IIIA). For example, S. mitis, accounted for
approximately 15% of the total frequency of IE
bacterial species, and approximately 28% of the
Streptococcus spp. (data not shown). Although the
incidence of S. mitis in blood was reduced by 78% in
the extractioneamoxicillin group (see Table IIIA),
amoxicillin did not completely eliminate these
bacteria. Other studies have also reported the failure
of AP to prevent IE caused by S. mitis and pointed
out the increasing resistance of this species to b-
lactam antibiotics, such as amoxicillin.14-17 Some
of bacterial early studies have suggested that amoxicillin has a
species (%) lower, or no, significant effect on the incidence or
24 (56) magnitude of bacteremia during and up to 10 minutes
144 (65) after dental extractions.18
12 (28) Non-streptococcal species that were significantly
68 (31)
impacted by amoxicillin are rare causes of IE. For
example, the frequency of bacteremia from Fusobacte-
rium nucleatum was decreased by 90% (see Table IIIA),
but we could find only six reported cases of IE caused by
this organism.11,12 Another rare cause of IE is Peptos-
treptococcus micros, which was reduced by 85% by
amoxicillin, but we found only 21 reported cases of IE
caused by this species (see Table IIIA).13 Similar trends
hold true for other anaerobic species detected in our
study but which rarely cause IE.13,19-21 Therefore, the
reduction in frequency of these bacterial species by
amoxicillin may not be clinically significant in the pre-
vention of IE.
Prosthetic joint infections
Our study suggests that amoxicillin decreased the fre-
quency of PJI bacterial species in blood after single tooth
extractions by 81% (see Table IIIB). These species
belong to the diverse group of aerobic and anaerobic
bacterial species inhabiting the oral cavity. However,
ORAL MEDICINE
434 Mougeot et al.
the majority of these bacterial species was identified as
viridans group streptococci rarely associated with
PJIs.22 The longstanding practice of AP for PJIs is
controversial for several reasons, largely because PJIs
are rarely caused by species found in the mouth,23 for
example, Staphylococcus aureus, which was not
detected in our original study.1,10 Even if S. aureus
entered the bloodstream through the oral cavity, this
species is not significantly affected by amoxicillin.24
Additionally, the percentage of the total detected
bacterial species that are associated with PJI is essen-
tially the same between the extractioneplacebo and the
brushing groups (see Table IV) (31% vs. 28%,
respectively). Since tooth extraction and tooth
brushing disrupted similar percentages of bacterial
species that can cause PJIs, patients with prosthetic
joints might also be at greater risk over time from
such naturally occurring bacteremia.
There are some limitations to this study. First, as we
pointed out in our original paper,1 all participants in this
study presented to our hospital-based urgent care dental
clinic in need of an extraction and therefore may not
represent the general population with regard to de-
mographic characteristics and severity of dental disease.
Additionally, the subanalysis in the present study relies
heavily on published studies that used conventional
culture-based approach, rather than the more sensitive
16 S rRNA gene sequencing approach, to identify
bacterial species. Furthermore, although we have
investigated the nature and incidence of IE- and PJI-
causing bacteremia, we did not address such factors
as bacterial load and host immune responses, which are
likely to be important factors in the outcome of IE- and
PJI- associated bacteremia. The molecular microbi-
ology technology available at the time of our original
study did not allow for determining the bacterial load
for bacteremia.
Although the use of AP before dental extraction
decreased the frequency of bacterial species associated
with IE and PJIs, clearly bacteremia incidence is a sur-
rogate measure of risk for IE or PJIs. Our data support
the current emphasis on the fact that tooth brushing
likely causes a far greater incidence of bacteremia
annually compared with office-based dental procedures.
From a practical standpoint, since AP is not feasible
before tooth brushing and the efficacy of AP has yet to
be demonstrated, the benefits of AP before dental pro-
cedures may be outweighed by the risks, which include
cost, antibiotic resistance, and drug reactions.
CONCLUSIONS
Antibiotic prophylaxis is currently recommended by the
American Heart Association for patients at higher risk
for developing IE and is often given to those with
prosthetic joints. Although AP decreases the frequency
OOOO
April 2015
of all oral bacterial species, including species associated
with IE and PJIs, both tooth brushing and single tooth
extractions disrupt similar bacterial species in similar
proportions. Although more invasive dental procedures,
such as multiple extractions, could yield different in-
cidences of bacteremia compared with single tooth
extractions, our data support the current trend away
from the routine use of AP in these patient populations.
Considering the much higher frequency of tooth
brushing compared with dental office procedures and
the significant incidence of bacteremia caused by bac-
terial species associated with IE and PJIs that results
from tooth brushing, such routine daily activities in the
presence of poor oral hygiene and periodontal disease
may put patients at a greater risk for these two DSIs.
This work was supported by Carolinas Health Care System
(CHS) Research Fund. SS was the recipient of the CHS
Summer Scholar Program Award.
REFERENCES
1. Lockhart PB, Brennan MT, Sasser HC, et al. Bacteremia associ-
ated with toothbrushing and dental extraction. Circulation.
2008;117:3118-3125.
2. Dewhirst FE, Chen T, Izard J, et al. The human oral microbiome.
J Bacteriol. 2010;192:5002-5017.
3. Castillo DM, Sanchez-Beltran MC, Castellanos JE, et al. Detec-
tion of specific periodontal microorganisms from bacteraemia
samples after periodontal therapy using molecular-based di-
agnostics. J Clin Periodontol. 2011;38:418-427.
4. Hoen B, Duval X. Clinical practice. Infective endocarditis. N Engl
J Med. 2013;368:1425-1433.
5. Rethman MP, Watters W III, Abt E, et al. The American Acad-
emy of Orthopaedic Surgeons and the American Dental Associ-
ation clinical practice guideline on the prevention of orthopaedic
implant infection in patients undergoing dental procedures.
J Bone Joint Surg Am. 2013;95:745-747.
6. Lockhart PB, Brennan MT, Thornhill M, et al. Poor oral hygiene
as a risk factor for infective endocarditis-related bacteremia. J Am
Dent Assoc. 2009;140:1238-1244.
7. Wilson W, Taubert KA, Gewitz M, et al. Prevention of infective
endocarditis. Guidelines from the American Heart Association. A
guideline from the American Heart Association Rheumatic Fever,
Endocarditis, and Kawasaki Disease Committee, Council on
Cardiovascular Disease in the Young, and the Council on Clinical
Cardiology, Council on Cardiovascular Surgery and Anesthesia,
and the Quality of Care and Outcomes Research Interdisciplinary
Working Group. Circulation. 2007;116:1736-1754.
8. Watters W III, Rethman MP, Hanson NB, et al. Prevention of
orthopaedic implant infection in patients undergoing dental pro-
cedures. J Am Acad Orthop Surg. 2013;21:180-189.
9. Lockhart PB, Loven B, Brennan MT, Fox PC. The evidence base
for the efficacy of antibiotic prophylaxis in dental practice. J Am
Dent Assoc. 2007;138:458-474.
10. Bahrani-Mougeot FK, Paster BJ, Coleman S, et al. Diverse and
novel oral bacterial species in blood following dental procedures.
J Clin Microbiol. 2008;46:2129-2132.
11. Shammas NW, Murphy GW, Eichelberger J, et al. Infective
endocarditis due to Fusobacterium nucleatum: case report and
review of the literature. Clin Cardiol. 1993;16:72-75.
OOOO ORIGINAL ARTICLE
Volume 119, Number 4 Mougeot et al. 435

12. Weber G, Borer A, Riesenberg K, Schlaeffer F. Infective endo- 21. Haddow LJ, Mulgrew C, Ansari A, et al. Neisseria elongata
carditis due to Fusobacterium nucleatum in an intravenous drug endocarditis: case report and literature review. Clin Microbiol
abuser. Eur J Clin Microbiol Infect Dis. 1999;18:655-657. Infect. 2003;9:426-430.
13. Pathak R, Ngwe A, Enuh H, Saverimuttu J. Peptostreptococcus 22. Sollecito TP, Abt E, Lockhart PB, et al. The use of prophylactic
induced native-valve endocarditis. 2013. Available at: http:// antibiotics prior to dental procedures in patients with prosthetic
ispub.com/IJID/12/1/1478. Accessed September 3, 2014. joints: evidence-based clinical practice guideline for dental
14. Masuda K, Nemoto H, Nakano K, et al. Amoxicillin-resistant oral practitionersda report of the American Dental
streptococci identified in dental plaque specimens from healthy Association Council on Scientific Affairs. J Am Dent Assoc.
Japanese adults. J Cardiol. 2012;59:285-290. 2015;146:11-16.
15. Chunduri NS, Madasu K, Goteki VR, Karpe T, Reddy H. Eval- 23. Berbari EF, Osmon DR, Carr A, et al. Dental procedures as risk
uation of bacterial spectrum of orofacial infections and their factors for prosthetic hip or knee infection: a hospital-based
antibiotic susceptibility. Ann Maxillofac Surg. 2012;2:46-50. prospective case-control study. Clin Infect Dis. 2010;50:8-16.
16. Smith A, Jackson MS, Kennedy H. Antimicrobial susceptibility 24. Peel TN, Cheng AC, Buising KL, Choong PF. Microbiological
of viridans group streptococcal blood isolates to eight antimi- aetiology, epidemiology, and clinical profile of prosthetic joint
crobial agents. Scand J Infect Dis. 2004;36:259-263. infections: are current antibiotic prophylaxis guidelines effective?
17. Hall GE, Baddour LM. Apparent failure of endocarditis prophy- Antimicrob Agents Chemother. 2012;56:2386-2391.
laxis caused by penicillin-resistant Streptococcus mitis. Am J Med
Sci. 2002;324:51-53.
18. Hall G, Hedström SA, Heimdahl A, Nord CE. Prophylactic
administration of penicillins for endocarditis does not reduce the
Reprint requests:
incidence of postextraction bacteremia. Clin Infect Dis. 1993;17:
188-194. Peter B. Lockhart, DDS
19. Michaux-Charachon S, Lavigne JP, Le FA, et al. Endocarditis due Department of Oral Medicine
to a new rod-shaped Neisseria sp. J Clin Microbiol. 2005;43:886- Carolinas Medical Center
889. PO Box 32861
20. Oh S, Havlen PR, Hussain N. A case of polymicrobial endo- Charlotte
carditis caused by anaerobic organisms in an injection drug user. NC 28232-2861
J Gen Intern Med. 2005;20:C1-C2. Peter.lockhart@carolinashealthcare.org