Appraisal of the recommendations for
antibiotic prophylaxis of infective endocarditis
John L Brusch, MD
Cambridge Hospital, Cambridge, Massachusetts, USA
While the critical pathogenic steps are known for all types of infective endocarditis (IE), including acute bacterial
endocarditis (ABE) and subacute bacterial endocarditis (SBE), many clinicians approach the use of antimicrobials
for its prevention with uncertainty. The fundamental principles of primary antibiotic prophylaxis are often neglected,
even though recommendations for their implementation have developed over the last 45 years. The first of these
was established in 1955 for the prevention of rheumatic fever. Since then, there have been numerous protocols for
the prevention of IE; including those of several European consensus panels, the Medical Letter and most prominently
the recommendations of the American Hospital Association (AHA). Over time, all of these have become more flexible
in order to improve their use. For example, current 1997 AHA guidelines, essentially a modification of 1990 guidelines,
have rejected the parenteral route except for the highest risk patients undergoing genitourinary or gastrointestinal
Though most cases of IE are secondary to bacteremia that arise spontaneously from the routine activities of daily
living, such as brushing one's teeth or having a bowel movement, any procedure that causes injuries to a mucosal
surface colonized by microbes can lead to bacteremia. Dental work, suction abortions and tonsillectomies, among
other routine medical procedures, can lead to bacteremia. Thus, it is important to understand the benefits of antibiotic
prophylaxis, though there is evidence that clinicians often either do not know enough about its proper role, or
are unmotivated to follow its procedure. In a 1992 survey of Israeli dentists, only 58% used an acceptable regimen
of prophylaxis prior to dental work.(2) Furthermore, while these practitioners were aware of the appropriate antibiotic,
only about 50% of the respondents had basic knowledge of the proper timing of administration. About 29% knew the
cardiac conditions that constituted a high-risk patient for whom prophylaxis is recommended (Table 1), while 64%
were aware of those dental procedures that required prophylaxis (Table 2).
Need for prophylaxis
Experimental models have allowed better definition of the most important sites of antibiotic action. Findings indicate
that their principal prophylactic effect is to suppress bacterial growth on the valvular thrombus, allowing the
host's defenses to take over. Antiinfective agents given within two hours following a procedure will successfully
block infection. Additionally, adding penicillinase to the system after the initial infection of the thrombus blocks
the prophylactic actions of ampicillin.(3) Antibiotics sterilize the infective vegetation in over 90% of cases
of SBE and in 70% of cases of ABE. However, 15 to 25% of all patients eventually will require cardiac surgery to
correct the consequences of valvular damage. This rate is significantly higher (70%) for prosthetic valve endocarditis
(PVE). In addition, there is the increasing challenge of treating the IE of resistant enterococci and methicillin-resistant
Clearly, it would be better to prevent valvular infection to avoid its immediate and long-term complications.
Understanding the prophylactic benefits of antimicrobials is critical, especially given that the incidence of ABE
is increasing.(4) This type of IE often attacks previously normal valves, and because many patients have no previous
cardiac disease, prophylaxis is often not ordered. Currently only about 15 to 25% of cases of all types of IE are
secondary to invasive procedures. Only 50% of these patients have had a previous known valvular disorder that would
lead the practitioner to provide antibiotic prophylaxis. Thus, only about 10% of IE may be prevented by the current
The agents of infection vary, depending on the type of IE. In general, Staphylococcus
aureus has become the most frequent pathogen. S. viridans is the
infective agent in approximately 50% of SBE cases, and is the most common pathogen in cases involving secondary
procedures of the mouth, respiratory tract or esophagus. Amoxicillin is the mainstay of prophylaxis used in these
procedures. It is preferred to penicillin VK because of its superior gastrointestinal absorption. According to
1990 AHA guidelines, the recommended oral dose of amoxicillin was 3 grams administered one hour before a procedure
and 1.5 grams of amoxicillin 6 hours after the first dose. This has been reduced to a single dose of 2 grams given
one hour before the procedure. Although the dose schedule provides levels in serum adequate to prevent infection,
there is great variation in the absorption of amoxicillin among individuals. Therefore, it may be prudent to adhere
to the previous guidelines.
Five to 10% of the population has indications of mitral valve prolapse (MVP), a preexisting condition that sometimes
warrants the prophylactic administration of antibiotics. While not all MVP cases are at increased risk of developing
IE, when compared to the general population, the rate of IE is 5 to 10 times greater in MVP patients. Mitral valves
that show minimal leakage are not at increased risk. Individuals with valves that prolapse and leak (5) require
antibiotic prophylaxis. Mitral valves with evidence of degeneration and regurgitation are also candidates for a
preventative approach. About 3% of patients with regurgitation and/or thickening of the anterior mitral valve leaflet
will eventually develop IE.
In the critical care unit, various monitoring and therapeutic devices (i.e., vascular lines, endotracheal tubes
and hyperalimentation catheters) regularly damage the normal barriers to bloodstream invasion. Intravascular catheters,
for instance, are associated with bacteremia rates of 4 to 14%,(6) accounting for at least 120,000 cases of nosocomial
bacteremia a year. Central catheters cause 30 to 90% of bloodstream infections found in critical care units. The
risk of infection increases significantly after these devices have been in place for more than 4 days. Forty-five
percent of catheter-associated bacteremia occurs in patients with prosthetic valves, and 15% of these become infected
during the time of bacterial invasion of the bloodstream.
Because clindamycin is associated with a high prevalence of pseudomembranous colitis (up to 10% of patients), it
would be prudent to limit its use as a prophylactic agent. The newer macrolides, azithromycin and clarithromycin
supersede erythromycin; however, clarithromycin and azithromycin are more costly and, for many people, no better
tolerated than erythromycin. An advantage of the macrolides is their lower potential for severe allergic reactions.
This is especially important in cases where the exact degree of risk is not certain. One study estimated that for
MVP patients who were given penicillin intravenously, one case of fatal anaphylaxis occurred for every two to three
cases of IE that were prevented.(7) Thus, because of its probable lower allergic potential, Bor and Himmelstein
proposed that erythromycin be used as the standard prophylactic antibiotic for MVP.(8)
In instances of prosthetic valve endocarditis (PVE), oral administration of an antimicrobial appears to be an adequate
prophylaxis. However, because of the serious consequences of an infection of an artificial valve, strong consideration
should be given to an intravenous route, with its more assured serum levels. Alternatives are the newer macrolides
and the third and fourth generation quinolones (i.e., levofloxacin and trovofloxacin). While these antibiotics
provide reliable gastrointestinal absorption, their role in prophylaxis is not yet proven.
In general, invasive vascular procedures (i.e., replacement of stents) do not require antibiotics. Additionally,
procedures or surgeries that cut through properly prepared skin do not require antimicrobial prophylaxis. Moreover,
there is little clinical evidence to support the efficacy of antibiotic prophylaxis before oral surgery. With respect
to this issue, three case studies of varying design have been performed. One study showed that prophylaxis was
91% effective against infection;(9) while two other studies showed minimal (10) or no effect (11). It is important
to emphasize that an extremely large number of patients would be needed to definitively prove the efficacy of antibiotic
prophylaxis in most situations. It may be easier to demonstrate an effect among patients at highest risks.(12)
It is important to note that S. epidermidis is not covered by any of the recommended regimens. Although it is a major
cause of PVE,(13) it is almost inevitably acquired during implantation of the valve. Even cases that present themselves
one year after surgery are due to the retarded clinical expression of an infection that originated in the operating
room. However, special consideration must be paid to patients receiving ongoing antibiotic therapy for prevention
of an infection, such as penicillin for rheumatic fever. Fifteen percent of S.
viridans inhabiting the oral cavities of these individuals
are moderately to highly resistant to this drug. Increasing its dose for prophylaxis during procedures is not recommended.
Oral flora will return to its normal state 9 to 14 days after penicillin is withheld, although withholding prophylaxis
for this length of time may not be acceptable to individuals who are strongly committed to such a program. In addition,
while clindamycin, erythromycin or clarithromycin may be employed, resistant
S. viridans are no more sensitive to cephalosporins than
they are to penicillins. Furthermore, one must consider that the
risk of pseudomembranous colitis due to Clostridium difficile
makes clindamycin an inappropriate prophylactic agent.(14)
Patient education and compliance are important factors in the success rate of antibiotic prophylaxis. Failure rates
have been associated with a patient's lack of knowledge.(15) In one study, 78% of patients had received adequate
instruction, but only 20% of them retained any useful information concerning the specifics of prophylaxis.(16)
These numbers suggest the need for expansion of non-pharmacological approaches to prophylaxis. A major way of reducing
nosocomial bacteremias, for instance, is to adhere to strict sterilization techniques during the insertion and
daily care of intravascular catheters. If such techniques are fully followed, they may have a significant effect
in reducing nosocomial bloodstream invasion and subsequent nosocomial IE. Certain interventions, such as incorporating
iodinated alcohol in the hub of the vascular catheter, or the impregnating of chlorhexidine and silversulphadiazine
into the wall of the catheter, have shown variable success.(17) The simple use of an antiseptic mouthwash before
a dental extraction can also markedly limit the risk of bacteremia.
With the continual development of new prophylactic procedures, it is difficult for guidelines to remain up-to-date.
Thus, the experience often forms the basis on which decisions are made. The choice of a prophylactic agent depends
on the specifics of the patient, including allergic history, renal function and age. Each patient must be evaluated
as a whole and not simply as a cardiac valvular problem. And while newer classes of antibiotics should be considered
for prophylaxis, especially in light of the increasing resistance of cardiac pathogens, if we are to learn anything
from the past, we must limit the use of these promising drugs in order to retain their efficacy.
- American College of Cardiology/American Heart Association
guidelines for the clinical application of echocardiography: a report of the ACA/AHA Task Force on Practice Guidelines
(Committee on Clinical Application of Echocardiography). 1997. Circulation March 18; 95(6): 1686-744.
- Wagner O, Raz R. 1997. Infectious Diseases in Clinical
- Bayer AS. 1989. Chest 96:893; Francioli P, Moreillon
P, Glauser MP. 1983. Medicine 62:83.
- Lerner PI, Weinstein L. 1966. N Engl J Med 274:199.
- Weinstein L, Brusch JL. 1996. In Infective Endocarditis.
New York, NY: Oxford University Press: (15) 322.
- Brusch JL. 1998. In Infectious Diseases in Critical
Care Medicine edited by BA Cunha. New York, NY: Marcel Dekker: (20): 387.
- Hickey AJ, MacMahon SW, Wilcken DE. 1985. Am Heart
- Bor DH, Himmelstein DU. 1984. Am J Med 76:716.
- Imperiale TF, Horwitz RI. 1990. Am J Med 88:131.
- LaCassin F, Hoen B, Leport C et al. 1995. Eur Heart
- Vandermeer JT, Van Wijk W, Thompson J et al. 1992.
- Durack D. 1995. In Principles and Practice of Infectious
Disease edited by GL Mandell, J Bennett, R Dolin. New York, NY: Churchill Livingston: 793.
- Horstkotte D. 1991. In Infective Endocarditis edited
by D Horstkotte, E Bodnar. London, England: ICR Publishers: 233.
- Sprunt K. 1977. In Infective Endocarditis edited by
EL Kaplan, AV Taranta. Dallas, TX: American Heart Association (AHA Monograph No. 52: 17).
- Sholler GF, Celermajer JM. 1984. Med J Aust 140:650.
- Caldwell RI, Hurwitz CA, Girod DA. 1971. Am J Dis Child
- Gramsden WR, Breathnack AS. 1998. Current Opinion in
Infectious Dis 11:461.