Impact of Amoxicillin Prophylaxis on the Incidence, Nature, and Duration of Bacteremia in Children After Intubation and Dental Procedures

Transcription

1 Impact of Amoxicillin Prophylaxis on the Incidence, Nature, and Duration of Bacteremia in Children After Intubation and Dental Procedures Peter B. Lockhart, DDS; Michael T. Brennan, DDS, MHS; M. Louise Kent, RN; H. James Norton, PhD; David A. Weinrib, MD Background Controversy exists about the impact of prophylactic antibiotics on bacteremia after invasive dental procedures. The purpose of this double-blind, randomized, placebo-controlled study was to determine the impact of amoxicillin prophylaxis on the incidence, nature, and duration of bacteremia from nasotracheal intubation and dental procedures in children. Methods and Results Children were randomly assigned before surgery to the American Heart Association (AHA) recommended dose of amoxicillin or to a placebo. Aerobic and anaerobic blood cultures were drawn at 8 specific time points after intubation, dental restorative and cleaning procedures, and before, during, and after dental extraction(s), to include blood drawings up to 45 minutes after the last extraction. Aerobic and anaerobic blood culture results were used to determine the incidence, nature, and duration of bacteremia from these procedures. For the 100 children enrolled (mean age, 3.5 years), the overall incidence of positive blood cultures, defined as at least 1 positive culture of the 8, was significantly higher in the placebo (84%) than the amoxicillin group (33%) (P ). Bacteremia occurrence rates after intubation and after dental restorations and cleaning were 18% and 20% in the placebo group and 4% and 6% in the amoxicillin group (P 0.05 and P 0.07, respectively). At 1.5 minutes after the initiation of dental extractions, bacteremia occurred in 76% of the placebo group versus 15% of the amoxicillin group (P 0.001). The majority of the 152 positive cultures and of the 29 different bacteria identified were Gram-positive cocci. Bacteremia persisted longer in the placebo group. Conclusions Bacteremia from these procedures occurs more often, from a wider variety of bacterial species, and for a longer duration after dental extractions than previously reported in any age group. Amoxicillin has a significant impact on the incidence, nature, and duration of bacteremia after nasal intubation, dental restorative and cleaning procedures, and dental extractions. (Circulation. 2004;109: ) Key Words: endocardium cardiovascular diseases anesthesia infection trials There are important and unresolved questions about the risks for and prevention of distant site infection (DSI) from oral pathogens. Invasive dental procedures have often been reported to cause infective endocarditis (IE) and a variety of other DSI, but such clinical events are rare in contrast to the high frequency of bacteremia from routinely occurring activities (eg, tooth brushing). Although upward of 50% of cases of community-acquired IE might result from oral bacteria, the causal role of dental office procedures and the utility of antibiotic prophylaxis are controversies that go back several decades. Thayer 1 suggested in 1926 that poor oral hygiene was the most important factor for the incidence of IE, but most reports since then have focused instead on risks associated with invasive dental procedures. 2 5 There is no definitive science about the impact of antibiotics on the incidence of DSI, and the standard of care of providing antibiotic prophylaxis before invasive procedures might be more for medicolegal rather than scientific reasons. 6,7 Attempts have been made over the past 4 decades to describe the incidence and nature (species) of bacteremia from invasive dental procedures, but variations in study design, size, and outcomes create controversy as to the clinical significance of these data. A wide range of bacteremia incidence and a wide variety of bacteria have been reported on blood cultures after different dental procedures, and bacteremia probably occurs from any manipulation of the gingival tissues. 8 The literature suggests that the incidence, nature, and duration (IND) of bacteremia are influenced most by the invasiveness of a procedure and that extractions are the most likely of dental procedures to cause bacteremia, with Received June 9, 2003; de novo received October 17, 2003; revision received February 10, 2004; accepted March 8, From the Departments of Oral Medicine (P.B.L., M.T.B., M.L.K.), Internal Medicine (D.A.W.), and Biostatistics (H.J.N.), Carolinas Medical Center, Charlotte, NC. Correspondence to Dr Peter B. Lockhart, Department of Oral Medicine, Carolinas Medical Center, 1000 Blythe Blvd, PO Box 32861, Charlotte, NC Peter.Lockhart@carolinashealthcare.org 2004 American Heart Association, Inc. Circulation is available at DOI: /01.CIR

2 Lockhart et al Impact of Amoxicillin on Bacteremia After Extractions 2879 incidence figures in all age groups ranging from 10% to 100%. 9 Of interest is that routinely occurring bacteremia has an incidence of 15% to 86% after common activities such as tooth brushing, flossing, and chewing food. 8 Early work in this field suggests that oral bacteremia is uncommon in children, even after dental procedures, 10 but the incidence ranges as high as 65% in children undergoing extraction of teeth under general anesthesia. 4,11,12 Our review of the literature identified 38 different oral bacteria reported in blood cultures after various dental procedures in children. Although the duration of bacteremia after dental procedures may have a role in the risk of development of a DSI, there are no published studies large enough to define duration, with or without antibiotic prophylaxis. The literature suggests that the retrieval of organisms drops off sharply between 10 and 30 minutes after invasive dental procedures. 2 Although several studies address the impact of antibiotic prophylaxis on the IND of bacteremia after different dental procedures, the results are conflicting There are no data from large studies on duration of bacteremia from specific organisms or the impact of amoxicillin on individual species. The major objective of this prospective, double-blind, randomized, placebo-controlled study was to determine the IND of bacteremia from dental procedures in children and the impact of the American Heart Association (AHA) recommended dosing for amoxicillin on these surrogate measures of risk for DSI. Methods Children enrolled required dental treatment in the operating room setting because of behavior, young age, and/or the scope of treatment needs. Exclusion criteria included poorly controlled systemic illness, American Society of Anesthesiologists Classification of Physical Status level III or IV, 16 medical conditions requiring antibiotic prophylaxis, allergy to penicillin-type drugs, weight 12 kg, and exposure to systemic antibiotics within the past 2 weeks. After establishing patient eligibility, the study was described to the parent or guardian, who signed an institutional review board approved consent form. Routine preoperative history and physical examination, anesthesia consultation, and laboratory tests were completed on all subjects. A computer-generated random number scheme was used by our pharmacy to assign identically appearing syringes containing placebo or study drug. All investigators were blinded as to the assigned treatment. One hour before the anticipated time of intubation, children were given amoxicillin elixir (50 mg/kg) or placebo, and the time and volume ingested were recorded (Table 1). Subjects were sedated with midazolam and brought to the operating room, where a mask induction was performed and an IV line was placed for anesthetic purposes (Figure 1). For blood culture collections, a large-bore (18 to 22 g) angiocath needle was placed in a second puncture site in the antecubital fossa or dorsum of the hand, after scrubbing the site in the usual manner with alcohol, followed by povidone-iodine. The first 6-mL blood draw occurred 2 minutes after the initiation of intubation. After intubation and placement of a throat pack, dental radiographs were taken and a thorough oral examination was conducted, to include several standard measures for severity of dental disease. Dental restorations, pulp therapy, and cleaning were then completed, and a second 6-mL blood sample was drawn. Ten minutes later, a third 6-mL of blood was drawn for a baseline culture before dental extraction(s). A stopwatch was started and a straight elevator was used for 5 to 10 seconds, followed by forceps extraction of a tooth in the usual manner. Ninety seconds after the initiation of the first extraction, a fourth blood draw of 6 ml was taken. The remaining teeth indicated for extraction were then removed, and the TABLE 1. Baseline Characteristics of 100 Subjects Randomly Assigned to Amoxicillin Versus Placebo Characteristic Amoxicillin (n 49) Placebo (n 51) Mean age, y Weight, kg Male 30 (61) 26 (51) Ethnic group Asian 1 (2) 4 (8) Black 29 (59) 30 (59) White 16 (33) 14 (27) Hispanic 3 (6) 3 (6) Dental disease Radiographic abscess* 25 (51) 36 (73) Swelling 10 (20) 15 (29) Suppuration 19 (39) 23 (45) Clinical caries 49 (100) 51 (100) Gingivitis (0 3) No. of extractions Total time for extractions, min % of drug ingested Minutes from study drug to intubation Minutes from study drug to draw Values are mean SD or n (%). *P fifth blood draw occurred at 90 seconds after the final extraction. Additional 6-mL blood drawings were taken at 15, 30, and 45 minutes after the end of extraction(s). The angiocath needle and line were flushed with 3 ml of saline after each blood draw, and 2 ml of blood was drawn and discarded just before each draw. Microbial Analysis Each blood sample was immediately divided into 2-mL BACTEC pediatric aerobic (Peds Plus) and 4-mL BACTEC adult anaerobic bottles, on recommendation from the manufacturer (Becton Dickinson). After the eighth blood sample was drawn, the 16 aerobic and anaerobic bottles were transported immediately to the microbiology laboratory for incubation and processing according to standard methods. 17 Cultures with bacterial growth were Gram stained and subcultured onto appropriate media. Blood cultures were continuously monitored for growth with the use of an automated Microscan (Baxter) system, and standard biochemical tests were done manually to complete the identity. Blood cultures were incubated for up to 14 days before being considered no growth to avoid missing more slow-growing oral pathogens. Outcome Measures We analyzed the incidence and nature of bacteremia from intubation, dental restorative and cleaning procedures, and dental extractions, as well as the duration of bacteremia after dental extractions (ie, draws 4 to 8) in the treatment and placebo groups. Incidence Overall incidence of bacteremia was defined as the occurrence of a positive culture at any of the 8 blood draws. Separate incidence measures were determined for each blood draw as well. A preextraction baseline blood draw (No. 3) was taken, recognizing that there might be residual bacteremia 10 minutes after the dental restorative and cleaning procedures.

3 2880 Circulation June 15, 2004 Figure 1. Timeline. *In the case of single tooth extractions, the 5th blood draw took place 3 minutes after the initiation of the first extraction. Nature We categorized all bacteria isolated by BACTEC. Only bacteria considered as likely or possibly from the oral cavity were included in the analysis of draws 2 to 8. The following bacteria were considered as unlikely or rarely from the oral cavity: Staphylococcus coagulase negative, Streptococcus bovis, Streptococcus group D, Enterococcus, Micrococcus, Aeromonas species, Aerococcus, and Moraxella species. 18,19 All bacteria were considered in the analysis of the intubation blood draw (No. 1) because the skin and nasopharynx are more likely to harbor the bacteria listed above. Duration We determined the time (in minutes) that oral bacteria were identified after completion of dental extraction(s). Duration of bacteremia was calculated only for subjects with a positive culture at draw 4 and/or draw 5, which were combined and defined as an extraction incidence (EI) draw. For those with a positive EI draw, the time difference from the end of extractions to the last positive draw (ie, draws 6, 7, or 8) represents the duration of bacteremia. Negative cultures between positive cultures were considered spurious false-negatives. Data Analysis Descriptive statistics, including means and standard deviations or counts and percentages were calculated. For continuous data, a Student s t test or a Wilcoxon rank-sum test was used. A 2 or Fisher s exact test was used for nominal data. A logistic regression analysis was used to adjust for possible demographic and baseline differences, with the dependent variable a positive culture on blood draws 4 to 8. SAS software (SAS Institute, Inc) was used for all analyses. A probability value of 0.05 was considered statistically significant. The sample size was based on the proportion of subjects who had development of a bacteremia. To detect a difference of 30% in bacteremia incidence between treatment groups (ie, 85% in the placebo versus 55% in the control group), with a power of 80% and an of 0.05, we projected that 100 subjects would be necessary to allow for possible study dropouts. Results Study Population One hundred children between the ages of 1 and 8 years were enrolled in this study between March 1994 and June Of these, 51 were randomly assigned to placebo versus 49 to amoxicillin. On examination in the operating room, 12 subjects did not require dental extraction(s), and they were excluded from the analysis of draws 4 to 8. Three subjects did not have restorative procedures done, and draws 2 and 3 were not analyzed. There was no statistically significant difference in the baseline characteristics for all subjects, stratified by treatment group (Table 1). Incidence The overall incidence of positive cultures from all 8 draws was greater in the placebo group (n 43, 84%) than the amoxicillin group (n 16, 33%) (P ). The highest incidence of positive cultures at a single time point occurred 1.5 minutes (draw No. 5) after completion of tooth extraction(s) in the placebo group (n 34, 76%) versus the amoxicillin group (n 6, 15%) (P ) (Figure 2). The incidence of bacteremia after intubation (draw No. 1) and restorative and cleaning procedures (draw No. 2) was 18% and 20% in the placebo groups and 4% and 6% in the amoxicillin groups (P 0.05 and 0.07, respectively). Subjects in the placebo group had a bacteremia incidence of 18% (n 7) at 15 minutes, 16% (n 6) at 30 minutes, and 14% (n 5) at 45 minutes. By contrast, only 1 subject in the amoxicillin group had bacteremia at 15 minutes after extraction(s). There was a statistically significant decrease in the incidence of bacteremia from amoxicillin at all but one (draw No. 2) of the blood draws: D1 (P 0.05); D3 (P 0.03); D4 (P ); D5 (P ); D6 (P 0.04); D7 (P 0.01); and D8 (P 0.03). The logistic regression analysis suggests that the incidence of bacteremia associated with extraction (EI) draws increases with the age of the subject (P 0.025) and number of teeth extracted (P 0.002) and that the use of amoxicillin significantly reduced the incidence of bacteremia (P ). No subject had a positive culture at draws 6, 7, or 8 who did not have a positive EI blood draw. For the intubation draw, a logistic regression analysis demonstrated that amoxicillin significantly reduced the incidence of bacteremia (P 0.03), and the following factors were not associated with an increased incidence of bactere- Figure 2. Incidence of bacteremia in the amoxicillin and placebo groups.

4 Lockhart et al Impact of Amoxicillin on Bacteremia After Extractions 2881 TABLE 2. Bacteria Identified in Blood Cultures Placebo Amoxicillin D1 D2 D3 D4 D5 D6 D7 D8 D1 D2 D3 D4 D5 D6 D7 D8 Totals Gram( ) cocci* 1 1 Viridans streptococcus* S oralis group* 1 S mitis S sanguis I S anginosus group* S constellatus S intermedius S mutans S salivarius group S salivarius 1 1 Peptostreptococcus* 1 1 P micros Group F -hemolytic strep 1 Gemella morbillorum Enterococcus* 1 Abiotrophia Staphylococcus coag neg 2 2 Subtotal Gram( ) bacilli or rods* Lactobacillus Eubacterium aerofacies 2 1 Propionibacterium acnes Actinomyces* 2 1 A odontolyticus 1 A israelii 2 1 A myeri 1 A viscosus 1 Bifidobacterium 3 Corynebacterium (diphtheroids) Subtotal Gram( ) bacilli (or rods)* 1 1 Actinobacillus actinomycetemcomitans 1 1 Eikenella 1 2 Haemophilus parainfluenza 1 Prevotella* P melaninogenica 1 Subtotal Gram( ) cocci Veillonella Neisseria subflava 2 2 Moraxella nonliquefaciens 1 Moraxella catarrhalis 1 Subtotal Overall total *Not further speciated or otherwise specified. Bacterial species not reported previously in blood cultures from children after dental procedures. Bacteria reported to cause distant site infections. 37

5 2882 Circulation June 15, 2004 TABLE 3. Microbiology of Bacteremia From Extraction(s) Subject No. Draw 4 Draw 5 Draw 6 Draw 7 Draw 8 Antibiotic group 1 Propionibacterium acnes S salivarius S intermedius Placebo group 1* -Hemolytic strep -Hemolytic strep 2* Actinomyces Actinomyces Corynebacterium 3* S constellatus Gram( ) cocci S mitis S constellatus S constellatus Aerococcus 4 Corynebacterium Bacillus 5 Prevotella -Hemolytic strep Actinobacillus actinomycetemcomitans 6* Eubacterium aerofaciens Eubacterium aerofaciens 7* S mitis S mitis Gemella morbillorum 8 S mutans S mutans S mitis S sanguis Actinomyces israeli 9 Diphtheroids S mutans Propionibacterium acnes Veillonella S milleri 10 Gemella morbillorum Gemella morbillorum S mitis Prevotella intermedius Peptostreptococcus micros 11 S mitis S milleri S milleri 12 Actinomyces odontolyticus Propionibacterium acnes S mutans *Five subjects had the same bacteria identified in an EI blood draw (draws 4 and 5) and at least one follow-up blood draw (draws 6 through 8). mia: number of attempted IV needle sticks, complications with nasal intubation (yes, no), time period from antibiotic to intubation or to draw 2, total time period for intubation (mean, 1.96 minutes; range, 1 to 10), or difference in total time from intubation to draw 1. Nature Of the total of 152 positive cultures in both groups, there were 29 different oral bacteria cultured, although there were multiple reports of a bacteria genus that was not further speciated (Table 2). Overall, Gram-positive cocci were the most frequent group of bacteria identified (n 85, 56%), accounting for 56% (n 72) and 54% (n 13) of bacteria isolated by BACTEC in the placebo and amoxicillin groups, respectively. Gram-positive bacilli (rods) were the next most common group (n 45) of bacteria isolated, followed by Gram-negative bacilli (rods) (n 12) and Gram-negative cocci (n 10). There was a 5-fold difference in the number of positive blood cultures recovered from the placebo (n 128) versus the amoxicillin (n 24) groups. Viridans streptococci made up 45% (n 57) of the total bacteria cultured in the placebo group versus 33% (n 8) of the amoxicillin group. On one occasion, a type of bacteria (S sanguis) from draws 4 to 8 was reported in an earlier draw (draw 2) from the same subject. Twelve subjects had a positive EI blood culture at draw 4 and/or 5, with a variety of bacteria in draws 6 to 8 (Table 3). Duration There was a consistent effect of amoxicillin on the incidence of bacteremia at each blood draw and therefore on the duration of bacteremia (Figure 3). Twelve subjects (27%) in the placebo group had a positive EI draw with at least 1 positive successive blood draw and were therefore evaluable for duration, versus 1 subject (2%) in the antibiotic group. In the placebo group, 12 subjects (27%) had a bacteremia at 15 minutes (draw 6), and 5 (11%) were still bacteremic at 45 minutes (draw 8). Discussion This study provides the first comprehensive data on the impact of the AHA guidelines for amoxicillin on the incidence, nature, and duration of bacteremia after nasal intubation, dental restorations and cleaning, and dental extractions in children. Our finding of an 84% overall incidence from dental extraction(s) in the placebo group is the highest reported in a large, controlled study of children, although this Figure 3. Duration of bacteremia in the amoxicillin and placebo groups. *EI (extraction incidence) positive culture at draw 4 and/or draw 5.

6 Lockhart et al Impact of Amoxicillin on Bacteremia After Extractions 2883 is lower than what we reported for adults (94%). 9 This is consistent with our finding of a higher incidence of bacteremia in older children, which probably reflects the larger number of teeth and volume and species of bacterial flora as well as the increased prevalence and severity of periodontal disease with age. The nature and variety of the bacteria in the present study overlaps what others have reported after dental extractions in children. However, we found 15 additional oral bacteria not previously reported, representing a 40% increase in the number of different bacteria reported in this setting (Table 2). At least 7 of these 15 bacteria are reported to cause IE. The duration of bacteremia probably reflects the nature and volume of organisms that enter the circulation as well as multiple other host factors (eg, immunity), but it is not clear what role duration has on the risk for bacterial seeding of distant sites. Berry et al 20 reported a 14% incidence of bacteremia in children in the recovery room 15 to 20 minutes after dental extractions under general anesthesia, compared with our finding of 27% at 15 minutes. The present study provides the only human data to use modern culturing methodology that addresses duration of bacteremia beyond 30 minutes. Clearly, amoxicillin had a significant impact on duration, virtually eliminating bacteremia well within 30 minutes. Our finding of an 18% incidence of bacteremia after nasal intubation was the highest we could find in a pediatric population. Some authors have reported negative cultures after nasal intubation, 21,22 and others have found a varying incidence (5.5% to 17%) from nasotracheal intubation in all age groups. 23,24 The current AHA guidelines do not address the issue of intubation and risk for IE. There is ongoing debate about the efficacy, health risks, and cost-effectiveness of the routine use of prophylactic antibiotics for dental procedures Of the many species of oral bacteria that can enter the circulation, only a small number are reported to cause DSI. 37 Because some of these pathogens also colonize the skin and the upper and lower aerodigestive tracts, it is speculative as to the origin of a given case of DSI. A single antibiotic cannot be expected to cover all oral pathogens, and the degree to which aminopenicillins cover anaerobes as well as aerobes is of concern. 38 The use of amoxicillin in the present study reduced the incidence of bacteremia by pathogens reported to cause DSI, but both known pathogens and relative nonpathogens were recovered in the amoxicillin group (Table 2). A legitimate concern in studies that evaluate antibiotic efficacy and that use a cultivable methodology such as BACTEC is the issue of the continued elimination of bacteria by the antibiotic after the blood is transferred to the BACTEC bottle. However, BACTEC growth media contain resins that enhance microbial recovery in subjects receiving antimicrobial therapy. 17,39 Other investigators report a varying impact of antibiotics at various times after dental procedures Given the significant impact of amoxicillin on the incidence and duration of bacteremia in our study, it is tempting to assume that antibiotic prophylaxis is effective and desirable. However, amoxicillin does not prevent oral pathogens from entering the systemic circulation, and epidemiological evidence does not suggest that dental procedures are a source of IE. 7 Clearly, bacterial pathogens from the oral cavity do cause some cases of IE, but dental procedures are relatively rare events by comparison with routinely occurring bacteremia (eg, tooth brushing). There are no prospective, randomized clinical trials of patients thought to be at risk for DSI with IE as an end point because of ethical, logistic, and cost issues. We must therefore rely on the extrapolation of data from prospective studies of surrogate measures, such as IND of bacteremia from routinely occurring activities versus dental procedure related sources. Acknowledgments Grant funding was provided by Health Services Foundation, Inc, Carolinas HealthCare System, Charlotte, NC. The authors would like to acknowledge the following clinicians for their participation in this study: Margaret Lochary, DDS, MS; Kelly Zukaitis, DDS; and Carlton V. Winter, DDS, MS. In addition, we thank Philip C. Fox, DDS, for his advice on the manuscript; David Rupar, MD, for his help with the microbiology data; and William T. Williams, MD, for his instrumental and long-standing support of this effort. References 1. Thayer W. Studies on bacterial (infective) endocarditis. Hopkins Hosp Rep. 1926;22: Baltch AL, Pressman HL, Schaffer C, et al. Bacteremia in patients undergoing oral procedures: study following parenteral antimicrobial prophylaxis as recommended by the American Heart Association, Arch Intern Med. 1988;148: Guntheroth WG. How important are dental procedures as a cause of infective endocarditis? Am J Cardiol. 1984;54: Coulter WA, Coffey A, Saunders IDF, et al. Bacteremia in children following dental extraction. J Dent Res. 1990;69: Lacassin F, Hoen B, Leport C, et al. Procedures associated with infective endocarditis in adults: a case control study. Eur Heart J. 1995;16: Lockhart PB, Brennan MT, Fox PC, et al. Decision-making on the use of antimicrobial prophylaxis for dental procedures: a survey of infectious disease consultants and review. Clin Infect Dis. 2002;34: Strom BL, Abrutyn E, Berlin JA, et al. Dental and cardiac risk factors for infective endocarditis: a population-based case-control study. Ann Intern Med. 1998;129: Lockhart PB, Schmidtke MA. Antibiotic considerations in medically compromised patients. Dent Clin North Am. 1994;38: Lockhart PB. An analysis of bacteremias during dental extractions: a double-blind, placebo-controlled study of chlorhexidine. Arch Intern Med. 1996;156: Faigel HC, Gaskill WF. Bacteremia in pediatric patients following dental manipulations. Clin Pediatr. 1975;14: Peterson LJ, Peacock R. The incidence of bacteremia in pediatric patients following tooth extraction. Circulation. 1976;53: Roberts GJ, Watts R, Longhurst P, et al. Bacteremia of dental origin and antimicrobial sensitivity following oral surgical procedures in children. Pediatr Dent. 1998;20: Baltch AL, Schaffer C, Hammer MC, et al. Bacteremia following dental cleaning in patients with and without penicillin prophylaxis. Am Heart J. 1982;104: Hall G, Hedström SA, Heimdahl A, et al. Prophylactic administration of penicillins for endocarditis does not reduce the incidence of postextraction bacteremia. Clin Infect Dis. 1993;17: Hall G, Nord CE, Heimdahl A. Elimination of bacteraemia after dental extraction: comparison of erythromycin and clindamycin for prophylaxis of infective endocarditis. J Antimicrob Chemother. 1996;37: Keats AS. The ASA classification of physical status: a recapitulation. Anesthesiology. 1978;49: Chokephaibulkit K, Wanprapa N, Srifuengfung S, et al. Comparison of BACTEC automated blood culture system and conventional system in hospitalized pediatric patients. J Med Assoc Thai. 1999;82: von Eiff C, Becker K, Machka K, et al. Nasal carriage as a source of Staphylococcus aureus bacteremia. N Engl J Med. 2001;344: Isenberg HD, D Amato RF. Indigenous and pathogenic microorganisms of humans. In: Murray PR, Baron EJ, Pfaller MA, et al, eds. Manual of Clinical

7 2884 Circulation June 15, 2004 Microbiology. Washington, DC: American Society for Microbiology; 1995: Berry FA Jr, Yarbrough S, Yarbrough N, et al. Transient bacteremia during dental manipulation in children. Pediatrics. 1973;51: Otten J-E, Pelz K, Christmann G. Anaerobic bacteremia following tooth extraction and removal of osteosynthesis plates. J Oral Maxillofac Surg. 1987;45: Gerber MA, Gastanaduy AS, Buckley JJ, et al. Risk of bacteremia after endotracheal intubation for general anesthesia. South Med J. 1980;73: Dinner M, Tjeuw M, Artusio JF Jr. Bacteremia as a complication of nasotracheal intubation. Anesth Analg. 1987;66: McShane AJ, Hone R. Prevention of bacterial endocarditis: does nasal intubation warrant prophylaxis? BMJ. 1986;292: Kaye D. Prophylaxis for infective endocarditis: an update. Ann Intern Med. 1986;104: Roberts GJ. Dentists are innocent: everyday bacteremia is the real culprit: a review and assessment of the evidence that dental surgical procedures are a principal cause of bacterial endocarditis in children. Pediatr Cardiol. 1999;20: Gould IM, Buckingham JK. Cost effectiveness of prophylaxis in dental practice to prevent infective endocarditis. Br Heart J. 1993;70: Durack DT. Antibiotics for prevention of endocarditis during dentistry: time to scale back? Ann Intern Med. 1998;129: Durack DT, Kaplan EL, Bisno AL. Apparent failures of endocarditis prophylaxis: analysis of 52 cases submitted to a national registry. JAMA. 1983;250: Imperiale TF, Horwitz RI. Does prophylaxis prevent postdental infective endocarditis? A controlled evaluation of protective efficacy. Am J Med. 1990;88: van der Meer JTM, van Wijk W, Thompson J, et al. Efficacy of antibiotic prophylaxis for prevention of native-valve endocarditis. Lancet. 1992; 339: Lockhart PB, Crist D, Stone PH. The reliability of the medical history in the identification of patients at risk for infective endocarditis. J Am Dent Assoc. 1989;119: Jones RN, Sader HS, Erwin ME, et al. Emerging multiply resistant Enterococci among clinical isolates, I: prevalence data from 97 medical center surveillance study in the United States. Diagn Microbiol Infect Dis. 1995;21: Helovuo H, Hakkarainen K, Paunio K. Changes in the prevalence of subgingival enteric rods, staphylococci and yeasts after treatment with penicillin and erythromycin. Oral Microbiol Immunol. 1993;8: Kaneko A, Sasaki J, Yamazaki J, et al. Intravenous administration of vancomycin is ineffective against bacteremia following tooth extraction. Tokai J Exp Clin Med. 1995;20: Parrillo JE, Borst GC, Mazur MH, et al. Endocarditis due to resistant viridans streptococci during oral penicillin chemoprophylaxis. N Engl J Med. 1979;300: Lockhart PB, Durack DT. Oral microflora as a cause of endocarditis and other distant site infections. Infect Dis Clin North Am. 1999;13: Head TW, Bentley KC, Millar EP, et al. A comparative study of the effectiveness of metronidazole and penicillin V in eliminating anaerobes from postextraction bacteremias. Oral Surg Oral Med Oral Pathol. 1984; 58: Morello JA, Matushek SM, Dunne WM, et al. Performance of a BACTEC nonradiometric medium for pediatric blood cultures. J Clin Microbiol. 1991;29: