• There is little evidence that antibiotics can prevent endocarditis in susceptible individuals, but their administration is routine practice.
• Only a small proportion of cases are due to medical and surgical procedures.
• Antibiotics must be active against the likely pathogens and be administered before the procedure.
• Patients at risk in the intensive care unit will usually require a glycopeptide and an aminoglycoside.
• A cephalosporin or flucloxacillin plus an aminoglycoside are used in cardiac surgery, but courses need not be prolonged more than 24 h.
If a prosthetic valve is seeded during cardiac surgery, the patient will usually recover and leave hospital only to return up to a year later needing life-threatening surgery. Even if the patient recovers, life expectancy is reduced. Multiple antibiotic resistance in organisms causing endocarditis is making treatment more difficult. Therefore use of antibiotics to prevent endocarditis is considered essential, but there is little firm scientific evidence for present advice mainly because of the rarity of the disease. Furthermore, only 10 per cent of cases of endocarditis can be ascribed to bacteremia caused by invasive procedures.
Prevention of endocarditis in patients with abnormal heart valves can be achieved by many general measures, for example regular dental care (covered by antibiotic prophylaxis), reducing the duration of procedures in areas of resident bacterial flora, limiting the number and duration of intravascular catheterizations, and ensuring that bacteriuria has been treated before a urological procedure. Gingival inflammation can be reduced in patients at risk by encouraging brushing, use of chlorhexidine or fluoride mouthwash, and cleaning. Chlorhexidine applied to the gingiva or irrigated in the gingival sulcus prior to dental extraction will reduce bacteremia.
The rationale for the use of antibiotics in the prevention of endocarditis derives from animal models and clinical experience. Animal models depend on the presence of a plastic catheter and simulate prosthetic valve endocarditis rather than native valve disease. Dental extraction results in a bacteremia of about 100 CFU/ml after bacteria have been diluted and cleared by the reticuloendothelial system. However, there is no obvious relationship between the number of circulating bacteria and the likelihood of developing endocarditis. The ability to adhere to platelets and fibrin is more predictive in models.
Many individual cases of endocarditis following dental procedures have been reported, but the risk of developing endocarditis must be very low. Endocarditis following bacteremia associated with urological procedures has been recorded but the risk is similarly low. Underlying cardiac abnormalities (e.g. patent ductus arteriosus, prosthetic valves, aortic valve disease, or previous endocarditis) can greatly increase the risk of endocarditis. Mitral valve prolapse is common but merits antibiotic prophylaxis if it causes a murmur.
In humans, case-control studies suggest a wide range of efficacy of prophylaxis for native valves ( v§n,,,,,d®L,M§®L§L§l 199,2,). Medical and surgical procedures cause only a small proportion of cases of endocarditis. van der Meer and colleagues estimated that, if prophylaxis had been given for all procedures in 275 patients with known lesions in the preceding 180 days, 47 (17 per cent) of the cases might have been prevented. If given only in the last 30 days, 23 (8 per cent) of the cases would potentially be prevented. Only eight of 48 patients having a procedure in the previous 6 months had actually been given antibiotic prophylaxis. In 52 cases of endocarditis preceded by a procedure that could induce bacteremia, mitral valve prolapse was present in a third and prosthetic valves in a fifth. However, most had received an antibiotic during the procedure.
Prophylaxis should be given at the time of the procedure in doses sufficient to ensure therapeutic concentrations. Antibiotics must be active against the likely pathogens (a-hemolytic streptococci and enterococci). Amoxicillin (amoxycillin) is recommended because of its good gastrointestinal absorption and sustained serum concentrations. However, a combination of a penicillin and an aminoglycoside is more likely to be bactericidal in high-risk cases. Antibiotics should be started just before the procedure (no more than 1-2 h before) and not continued for more than 6 to 8 h to avoid the emergence of resistant bacteria.
Procedures causing gingival bleeding should be covered by prophylaxis, but this does not include endotracheal intubation or the adjustment of orthodontic appliances. Tonsillectomy, adenoidectomy, and other surgical operations to the respiratory mucosa should be covered. If the urine is infected, instrumentation must be covered by an antibiotic active against the pathogen isolated. Flexible bronchoscopy with biopsy, cardiac catheterization, endoscopy with biopsy, liver biopsy, and urethral catheterization in the absence of infection do not need prophylaxis. Patients having colonoscopy or sigmoidoscopy probably do not need prophylaxis unless there is a prosthetic valve or previous endocarditis.
Patients in the intensive care unit (ICU) are likely to have a resistant bacterial flora and the standard high-risk regimens may not be adequate. At-risk patients who are having a tracheostomy inserted should be given vancomycin (1 g intravenously) or teicoplanin (12 mg/kg intravenously) with gentamicin before the procedure. Insertion of a chest drain would require an appropriate agent if the effusion is likely to be infected, but the microbiologist should be consulted. Cannulation of peripheral veins and arteries does not need prophylaxis, nor does placement of a central venous catheter unless it is likely to be difficult or is a guidewire insertion into a previously infected site.
Recommendations for prophylactic regimens differ from country to country (Table..! and Ta.b.l.e.,,2). Most countries agree that dental procedures (causing gingival bleeding) and oropharyngeal, gastrointestinal, and urological procedures present a risk, but recommendations vary with respect to bronchoscopy, colonoscopy, and vaginal hysterectomy (LeP9LL§í.§L 1995). A single large oral dose of amoxicillin (3 g), as used in Europe, results in a high ratio of serum level to likely minimum inhibitory concentration of the pathogens and in animals was effective against bacteremia that would produce infection in 90 per cent of controls. An additional dose can be used if the likely inoculum is high, and in the United States all patients are required to take a second dose. Clindamycin or erythromycin is the usual oral alternative. Ampicillin or amoxicillin can be given intravenously 1 h before a procedure, with an oral dose 6 h later. High-risk patients are given aminoglycosides in addition, and vancomycin or teicoplanin is used for patients allergic to penicillins. Parenteral prophylaxis is used for gastrointestinal and urological procedures, and aminoglycosides are included in the regimen to provide synergistic activity against enterococci.
Table 1 European consensus for the prevention of endocarditis in at-risk patients
Table 2 Recommendations of the American Heart Association for the prevention of endocarditis
Only a minority of patients receive adequate prophylaxis or are properly advised about its use. In a German study, 75 per cent of cardiologists issued written and oral instructions to patients but 60 per cent of procedures were not covered ( Gohlke-Barwolf e.LaL 1995). Only 57 per cent of dentists administered antibiotics according to published recommendations. The need for prophylaxis was recognized by 53 per cent of patients at risk but, of patients recently having dental procedures, only 66 per cent (71/107) with prosthetic valves and 46 per cent (10/22) with abnormal native valves received antibiotic prophylaxis. It is a common error to administer antibiotic prophylaxis more than 30 to 60 min before the procedure, but this is not justified and risks the emergence of a resistant bacterial flora. Prolongation of prophylaxis presents a similar risk and increases the chance of adverse effects. The high oral dose of amoxicillin provides a sufficient serum concentration, but the parenteral route is preferable in most patients in the ICU.
Many a-hemolytic streptococci are tolerant to the action of amoxicillin, but prophylaxis is successful even below the minimum bactericidal concentration. Some benefit is gained even if antibiotics are administered after the bacteremia has occurred. Prevention of adherence to fibrin has been demonstrated at low antibiotic concentrations, rendering the bacteria susceptible to host defense. Low numbers of bacteria adhering to vegetations are eventually eradicated in the absence of antibiotics, and inhibition of adherence of greater numbers by antibiotics may allow this process to continue ( B.ialt®I.§Dd..Fl§Dco.!i..1995). It is possible that bactericidal agents are produced by the platelets in the vegetation.
Patients having open-heart surgery are at high risk of bacteremia. Coagulase-negative staphylococci and coryneforms are commonly present in the blood during surgery and can be disseminated from central venous and peripheral catheters in the early postoperative period. During surgery, host defense is compromised by complement activation after contact with plastics in the pump, antibiotics are affected by protein binding and hemodilution, a large operative area is at risk, many staff are involved, and surgery is often prolonged. Blood accumulating in the chest cavity is returned to the circulation by suction, together with contaminating organisms. If prosthetic valve endocarditis develops soon after surgery, it is often fatal. Patients with dental disease should have treatment or extractions, under antibiotic cover, before surgery.
The use of antibiotic prophylaxis is routine during cardiac surgery. Flucloxacillin plus an aminoglycoside or a cephalosporin is a common choice in the United Kingdom, while cephalosporins are often used alone in the United States. In most Scandinavian countries, single-drug antistaphylococcal agents are used for less than 48 h. Animal models suggest that antibiotic prophylaxis is effective only up to 4 h after surgery, and in humans 2-day or even single-dose regimens are effective in preventing wound infection. Although the efficacy of prophylaxis has been demonstrated in the prevention of wound infection, the effect on endocarditis has not been unequivocally shown in a controlled trial. Improvements in surgical technique are probably responsible for most of the improvement in the rate of postoperative endocarditis, but a risk of early endocarditis remains in 0.5 per cent of patients. The coagulase-negative staphylococci associated with early prosthetic valve endocarditis are usually resistant to prophylactic antibiotics, suggesting that colonization or bacteremia has occurred in the operating room or the ICU. Therefore use of antibiotics beyond 24 h is to be discouraged.
Regimens including aminoglycosides would be expected to be more active than cephalosporins against the majority of coagulase-negative staphylococci, but both regimens work well in practice. The rarity of endocarditis makes direct comparisons difficult. If aminoglycosides are used, only 24-h dosing should be given. Some surgeons ask for 48-h administration to cover intravascular catheterization, in which case a single trough level should be taken the morning after surgery. If it is higher than 2 mg/l, the gentamicin should be stopped. Even a 2-day course of aminoglycoside is associated with a significantly higher serum creatinine after surgery than other antibiotics. Vancomycin has been used, particularly in units where methicillin-resistant Staphylococcus aureus (MrSA) is prevalent, but it is associated with an increased need for inotropes in the early postoperative period.
Was this article helpful?