123980 971.98

Journal of Antimicrobial Chemotherapy (2004) 54, 971–981DOI: 10.1093/jac/dkh474 Advance Access publication 16 November 2004 Guidelines for the antibiotic treatment of endocarditis in adults: report of the Working Party of the British Society for T. S. J. Elliott1, J. Foweraker2, F. K. Gould3*, J. D. Perry3 and J. A. T. Sandoe4 1Department of Microbiology, Queen Elizabeth Hospital, Birmingham; 2Department of Microbiology, Papworth Hospital, Cambridge; 3Department of Microbiology, Freeman Hospital, Newcastle-upon-Tyne; 4Department of Medical Microbiology, Leeds Teaching Hospitals NHS Trust, Leeds, UK The BSAC Guidelines on Endocarditis were last published in 1998. The Guidelines presented herehave been updated and extended to reflect changes in both the antibiotic resistance characteristics ofcausative organisms and the availability of new antibiotics. Randomized, controlled trials suitable forthe development of evidenced-based guidelines in this area are still lacking, and therefore a consensusapproach has again been adopted. The Guidelines cover diagnosis and laboratory testing, suitableantibiotic regimens and causative organisms. Special emphasis is placed on common causes ofendocarditis, such as streptococci and staphylococci, however, other bacterial causes (such as entero-cocci, HACEK organisms, Coxiella and Bartonella) and fungi are considered. The special circum-stances of prosthetic endocarditis are discussed.
Keywords: antibiotic therapy, guidelines, endocarditis, BSAC 5.7 Prosthetic valve endocarditis (PVE)5.8 Treatment of streptococcal endocarditis for patients 3.4 Alternative antibiotics for patients with penicillin 3.5 Empirical therapy3.6 Duration of therapy In 1998, the Endocarditis Working Party of the British Society for Antimicrobial Chemotherapy published updated guidelines for the treatment of streptococcal, enterococcal and staphylococ- cal endocarditis.1 In the light of the introduction of new antibiotic agents and the emergence of increasingly antibiotic- resistant bacteria causing endocarditis the existing guidelines have been revised. They have been widened to include recom- 5.3 Nutritionally variant streptococci (NVS) mendations for the treatment of other bacterial and fungal causes 5.5 Streptococcus pyogenes (Group A streptococcus) Guidelines such as these have, in the past, received criticism 5.6 Streptococcus agalactiae (Group B streptococcus), for not being evidence based. Whereas we appreciate that the gold standard for all clinical guidelines should ideally be based *Corresponding author and Chair of the Working Party. Tel: +44-191-233-6161; Fax: +44-191-223-1224; E-mail: [email protected] JAC vol.54 no.6 q The British Society for Antimicrobial Chemotherapy 2004; all rights reserved.
on good, prospective, randomized controlled trials, no such trials same reasons, the measurement of serum bactericidal titres is not have ever been performed to assess the benefit of antibiotic regimens in the treatment of endocarditis. Consequently we have Failure to culture a causative microorganism in IE is most not attempted to classify the evidence for our recommendations, commonly due to the administration of antimicrobials prior to which remain consensus based. An extensive review of the lit- sampling, but may also be due to IE caused by fastidious or erature—encompassing a number of different search methods slow-growing microorganisms. Possible microorganisms that and a range of criteria (e.g. endocarditis, staphylococci, etc.)— should be considered include the HACEK group, Bartonella has been carried out, and publications used to support any spp., Brucella spp., Chlamydia spp., Coxiella burnetii (Q-fever), changes we have made to the existing guidelines have been Legionella spp., mycobacteria and various fungi.3,13–15 Diagnos- cited. Publications referring to in vitro or animal models have tic methods should include serological investigations where they only been cited if appropriate clinical data are not available.
are available, and such methods are central to the identificationof Bartonella spp., Brucella spp., Chlamydia spp., C. burnetii,Legionella pneumophila and Mycoplasma pneumoniae. A sys- tematic approach to the serological diagnosis of culture-negativeIE is advised, based on the clinical history of the patient and Since 1994, the Duke criteria have formed the basis for the diag- their exposure to possible risk factors.13–16 Molecular techniques nosis of infective endocarditis (IE).2 Major diagnostic criteria for the detection of microbial nucleic acids show promise as an include positive blood cultures or positive echocardiogram. A adjunct to existing methods for the diagnosis of endocarditis range of minor criteria are included in the Duke strategy and caused by fastidious or slow-growing microorganisms.17–22 supplementary criteria have since been proposed to enhancediagnosis further.3–6 In terms of laboratory testing, blood culture remains the cornerstone of diagnosis for the vast majority of IE cases.
Patients with suspected IE, such as those with pyrexia and a heart murmur, should have blood cultures taken promptly before Gentamicin dosage regimens are based on 8 hourly adminis- antimicrobial chemotherapy is given. Three sets of blood cul- tration of 1 mg/kg body weight, intravenously (iv)/intramuscu- tures should be taken in the first 24 h from separate venepunc- larly (im). Once-daily regimens are now widely used for other tures. It is recommended that at least 30 mL of blood is cultured infections, but data regarding their efficacy in endocarditis are in total.7 The timing of the initiation of therapy after this period limited. Levels should be measured regularly to ensure pre-dose depends upon the condition of the patient. In patients who are (trough) levels remain <1 mg/L. Maintaining post-dose (peak) acutely ill, three sets of blood cultures should be taken within levels between 3 – 5 mg/L has been advocated by other auth- 2 h before starting empirical therapy.
orities, although the evidence to support this is limited. Where a Blood cultures should be incubated for a minimum of 7 days, patient has normal renal function that is stable, twice weekly which should result in a positive culture for 95% of IE cases.7 In monitoring may suffice. In patients with impaired renal function, exceptional circumstances, where blood cultures remain negative dosage should be adjusted according to measured or estimated after 7 days in the absence of prior antibiotic therapy, and IE creatinine clearance and drug monitoring results, and levels remains strongly suspected, extended incubation of cultures should be initiated that may facilitate the isolation of some fasti- Streptomycin dosage is usually 7.5 mg/kg body weight 12 dious microorganisms.3,8–10 In such cases, incubation of blood hourly and should be monitored at least twice weekly, (more cultures should be continued for at least 3 weeks, with weekly often in renal impairment—see above) by maintaining pre-dose subculture onto chocolate agar, which in turn should be incu- levels <3 mg/kg, and adjusted according to renal function as bated for 3 weeks in air plus 5% carbon dioxide.
Once the causative microorganism has been isolated, suscep- tibility testing should be performed with suitable antimicro- bials. The choice and duration of treatment depend upon thetype of microorganism and its susceptibility profile, allergy to 3.2.1 Vancomycin. For patients with normal renal function we antimicrobials and whether infection involves a native or pros- recommend 1 g iv 12 hourly. Levels should be monitored to thetic valve. A cidal antimicrobial, or combination of anti- maintain a pre-dose level between 10 – 15 mg/L, although in microbials, is required to eradicate infection. A minimum enterococcal endocarditis some authorities recommend pre-dose inhibitory concentration (MIC) of the chosen antimicrobial levels of 10 – 20 mg/L. Further monitoring and dose adjustment should be established by a standardized laboratory method to in patients with impaired renal function should be performed as ensure susceptibility.11 Etest strips may be useful for establish- ing the susceptibility of fastidious or slow-growing bacteria, 3.2.2 Teicoplanin. Teicoplanin must be administered at a high such as the HACEK group12 (Haemophilus spp., Actinobacillus dosage (10 mg/kg body weight 12 hourly then 10 mg/kg daily).
actinomycetemcomitans, Cardiobacterium hominis, Eikenella Trough levels must be measured to ensure levels of at least 20 mg/L and repeated at least weekly.
As in previous guidelines1 we do not recommend routine measurement of the minimum bactericidal concentration (MBC).
MBC determination is affected by a range of technical factorsthat result in poor intra-laboratory reproducibility and there In our recommendations for sensitive organisms, amoxicillin/ remains a lack of evidence regarding its clinical value. For the ampicillin 2 g iv 4 – 6 hourly may be used instead of benzyl Table 1. Recommendations for empirical therapy Flucloxacillin (8 – 12 g iv daily in 4 – 6 divided doses)plusgentamicin (1 mg/kg body weight iv 8 hourly, modified according to renal function) Penicillin (7.2 g iv daily in 6 divided doses) or ampicillin/amoxicillin (2 g iv 6 hourly)plusgentamicin (1 mg/kg body weight 8 hourly iv, modified according to renal function) Vancomycin (1 g 12 hourly iv, modified according to renal function) rifampicin (300 – 600 mg 12 hourly by mouth)plusgentamicin (1 mg/kg body weight 8 hourly iv, modified according to renal function) penicillin. The time-dependent killing of streptococci by penicil- lin means that the drug should be given at least six times a day Apart from the treatment of certain strains of penicillin-sensitive because of its short serum half-life. There are theoretical streptococci, we recommend a minimum of 4 weeks therapy.
and experimental reasons for using a continuous infusion of There is evidence from patients with enterococcal endocarditis penicillin.23 There are, however, no prospective comparisons of and some data from early studies of streptococcal endocarditis to continuous with intermittent penicillin for streptococcal endocar- suggest that patients who have had symptoms for more than 3 ditis. Dosage modifications for b-lactams may be necessary in months benefit from 6 weeks of penicillin.24–25 Often these indi- patients with impaired renal function and according to the viduals have larger vegetations and mitral valve disease (also indicators of a poorer response). These factors should be takeninto consideration when determining treatment length. Apparent 3.4 Alternative antibiotics for patients with penicillin allergy failure to respond to treatment may indicate the need for surgical It is important to establish the nature of a reported ‘allergy’ to intervention. There is no evidence to support the use of oral ‘fol- penicillin, as there is less experience with alternative antibiotics low-on’ therapy after completion of a course of treatment.
and a higher rate of side effects. For example, a history of a rashwith ampicillin or amoxicillin may not indicate true allergy.
Unless signs of immediate-type hypersensitivity (anaphylaxis, Home therapy for endocarditis has been described. Suitability angio-oedema, bronchospasm, urticaria) were reported, a trial for home therapy will depend on the patient, the availability of with penicillin may be warranted. The American Heart Associ- the infrastructure to support such therapy and the sensitivity of ation (AHA) advises ceftriaxone for the penicillin-allergic the infecting organism to antibiotics, which lend themselves to patient—but this should only be used for allergy other than immediate-type hypersensitivity because of the risk of cross- Home treatment is often considered for streptococcal endocar- ditis, as it can be less destructive, with fewer complications,than infection caused by other organisms. Trials of home therapy have been reviewed.26,27 Antibiotics such as ceftriaxone or See Table 1 for a summary. Bacterial endocarditis (particularly teicoplanin, which can be given once daily iv or im, have been prosthetic or Staphylococcus aureus endocarditis) may progress advocated as the patient may not need a central venous catheter.
rapidly and in such cases antibiotic therapy must be commenced Neutropenia is, however, a well described side effect of ceftriax- as soon as all the appropriate specimens have been collected. If one, occurring in two of 55 patients in one study.28 Teicoplanin the diagnosis of endocarditis is in doubt, the patient is clinically also has side effects, including a high rate of drug fever (see stable and has already received antibiotics, we recommend stop- ping any antibiotics for 2 – 4 days and re-culturing.
If empirical therapy is indicated, we recommend a combi- nation of flucloxacillin (8 – 12 g daily in 4 – 6 divided doses) plusgentamicin (1 mg/kg body weight 8 hourly according to renal See Table 2 for a summary. The choice of treatment for function) if the patient is acutely unwell, or penicillin (or ampi- staphylococcal endocarditis will depend more on the antibiotic cillin/amoxicillin) plus gentamicin if the presentation is more sensitivity of the isolate than whether it is coagulase positive or indolent. If the patient has intra-cardiac prosthetic material, or negative. In the previous guidelines, therapy with benzyl penicil- MRSA is suspected, we recommend vancomycin (1 g 12 hourly lin was recommended for penicillin-sensitive strains. In practice, according to renal function) plus rifampicin (300 – 600 mg 12 hourly, orally) plus gentamicin (1 mg/kg 8 hourly iv). Therapy For methicillin-sensitive strains, we recommend flucloxacillin should be reviewed as soon as the aetiological agent is 2 g 6 hourly29,30 increasing to 2 g 4 hourly in patients weighing >85 kg. There is no evidence that the addition of gentamicin is Table 2. Summary of treatment recommendations for staphylococcal endocarditis Vancomycin (1 g iv 12 hourly), modified according to renal functionplusrifampicin (300 – 600 mg 12 hourly by mouth)*orgentamicin (1 mg/kg body weight 8 hourly, modified according to renal function)*orsodium fusidate (500 mg 8 hourly by mouth)* Flucloxacillin (2 g 4 – 6 hourly iv) or vancomycin (1 g iv 12 hourly, modified plusrifampicin (300 – 600 mg 12 hourly by mouth)*and/orgentamicin (1 mg/kg body weight 8 hourly, modified according to renal function)*and/orsodium fusidate (500 mg 8 hourly by mouth)* more likely to result in a successful outcome, but it is associated with an increased incidence of adverse effects.31,32 Therefore, we Since the publication of the last guidelines, new antibiotics such no longer recommend its routine use in the treatment of staphy- as linezolid51 and quinupristin/dalfopristin52 have become avail- lococcal endocarditis. There is no evidence that the addition able. The use of these agents in the treatment of endocarditis has of sodium fusidate offers any advantage,33 and the benefit of been described in the literature, but experience is still limited.
We would only recommend the use of such agents as salvage strains, or in patients with penicillin allergy we recommend van- therapy, in patients unable to tolerate conventional therapy, or comycin 1 g 12 hourly. Levels should be monitored and trough from whom particularly resistant stains have been recovered.
levels maintained between 10 – 15 mg/L. Owing to the reported Similarly, the use of co-trimoxazole, quinolones and clindamy- incidence of treatment failure, we do not recommend the routine cin has also been described in the literature, but we cannot advo- use of teicoplanin in the treatment of staphylococcal endo- carditis.35–37 As vancomycin is less active than flucloxacillin,38,39we recommend the addition of a second antibiotic to the treat-ment regimen. The choice of the second antibiotic will depend on the sensitivity of the infecting organism, with rifampicin40,41as the preferred choice, but may include gentamicin or sodium See Table 3 for a summary. Streptococci are common causes of native valve and late prosthetic valve endocarditis. They vary intheir susceptibility to penicillin and degree of resistance toaminoglycosides. Early clinical studies showed that short courses (10 – 14 days) with standard doses of penicillin had a high The complex nature of staphylococcal infection of prostheses relapse rate, even for streptococci that were extremely sensitive.
leads us to recommend combination therapy including rifampicin The combination of penicillin with an aminoglycoside gave bet- (if sensitive) for both methicillin-sensitive and -resistant strains ter results than penicillin alone, especially for streptococci that of staphylococci for at least the first 2 weeks of therapy. If the were relatively penicillin resistant or tolerant. A well considered isolate is resistant to rifampicin, other agents may be considered review of clinical evidence plus animal and in vitro studies led (see above). The use of three antibiotics (if the isolate is sensi- to the development of the AHA guidelines for the treatment of tive) has been advocated by some authorities.43 streptococcal endocarditis.43 We propose that they should beused, but there are areas where we would modify or expand theadvice as follows (see also Table 1).
The majority of patients will require at least 4 weeks therapy, which should be extended to at least 6 weeks in patients withintra-cardiac prostheses, and after removal of infected permanent For native valve endocarditis, there is evidence that 4 weeks of pacing. However, in patients with right-sided endocarditis (often high-dose penicillin can be used for sensitive streptococci iv drug abusers), several trials have demonstrated the efficacy of _ 0.1 mg/L), and that short-course treatment (2 weeks peni- short course iv combination therapy32,44–48 and oral therapy.49 cillin in combination with gentamicin) may be as effective. The Other studies have demonstrated that iv to oral switch therapy to short regimen should not be used if there is an intra-cardiac complete a course of antibiotics is effective,50 and may be con- abscess or infected emboli. The AHA recommends 4 weeks sidered in selected patients, for whom iv access is difficult.
penicillin plus gentamicin for the first 2 weeks for relatively Table 3. Summary of treatment options for streptococcal endocarditis Streptomycin is an alternative to gentamicin for streptomycin-sensitive, gentamicin-resistant isolates. (See sections on susceptibility testing, drug toxicity and monitoring levels.)aPenicillin and gentamicin for 2 weeks should not be used if there is an intracardiac abscess or extra-cardiac focus of infection.
bIf gentamicin or streptomycin is contraindicated (unacceptable risk of toxicity or a resistant bacterium).
cUse only for isolates that are susceptible to ceftriaxone (MIC < 0.5 mg/L).
dSix weeks treatment.
Dosage (NB all need to be adjusted in renal impairment): penicillin, 1.2 – 2.4 g, 4 hourly or by continuous infusion; gentamicin, 1 mg/kg (ideal body weight), 8 – 12 hourly; ampicillin or amoxicillin, 12 g,over 24 h; vancomycin, 1 g, 12 hourly; streptomycin, 7.5 mg/kg body weight, 12 hourly.
resistant strains of viridans streptococci (penicillin MIC > 0.1 – <0.5 mg/L), but there is little evidence for this. We would agree There is limited evidence on the optimum treatment of strepto- that aminoglycosides should ideally be given for the first 2 coccal PVE infections, and sensitive organisms may be difficult weeks, but where the risks of using aminoglycosides are high, 4 to treat. This is partly due to the involvement of the valve ring weeks penicillin alone at the higher dose (2.4 g 4 hourly, and myocardial abscesses. Early studies showed a high relapse adjusted for renal function) could be tried. The treatment of rate with 2 weeks penicillin plus an aminoglycoside. We more resistant viridans streptococci (penicillin MIC > agree with the recommendation of a minimum of 6 weeks of should follow the recommendations for treating enterococci.
penicillin, plus gentamicin for the first 2 weeks, for penicillin-sensitive streptococci (MIC < advises 6 weeks penicillin with 4 weeks gentamicin for more Streptococcus bovis should be treated using the same regimens penicillin-resistant strains, but there are no studies comparing 4 for the viridans-type streptococci, based on the penicillin MIC.
and 6 weeks gentamicin. In view of the difficulty in replacing a There is a strong association between endocarditis due to S. bovis prosthetic valve and the toxicity from a second course of amino- and benign and malignant tumours of the gut and liver disease.57 glycosides should initial treatment fail, it seems prudent to give It is easy in our experience to neglect investigation of the gut in 6 weeks of gentamicin (see Table 1) with penicillin for strains the understandable drive to treat the endocarditis.
with a penicillin MIC > 0.1 mg/L.
5.3 Nutritionally variant streptococci (NVS) 5.8 Treatment of streptococcal endocarditis for patientsallergic to penicillin Endocarditis with NVS is characterized by an indolent course,frequent embolization and bacteriological relapse. Many NVS Vancomycin tolerance is said to be common, but there is evi- are relatively resistant to penicillin and show tolerance. Anti- dence of synergy with aminoglycosides. The AHA advises 4 biotic susceptibility testing is difficult and may need to be weeks of vancomycin alone for moderately penicillin-resistant carried out by a specialist laboratory. Four to six weeks of peni- streptococci (MIC > 0.1 < 0.5 mg/L). There is no clinical evi- cillin plus an aminoglycoside is recommended.58 Until more evi- dence to justify the omission of an aminoglycoside and the clini- dence is available on alternatives, NVS should also be treated cian has to balance the potential risks of toxicity against concern according to the recommendations for treating enterococci. In over relapse that could lead to a longer course of a potentially view of the high relapse rate, blood should be cultured weekly toxic combination. With close monitoring of vancomycin and during treatment and after completion of therapy.
gentamicin levels, we would advise vancomycin for 4 weekswith gentamicin for the first 2 weeks, for moderately resistant streptococci. Teicoplanin is considered less nephrotoxic bysome, both alone and in combination with gentamicin. There is, Pneumococcal endocarditis is usually associated with pneumonia however, less experience with its use and in early studies many and/or meningitis.59 The high mortality (28% – 60%) seems due patients on long-term teicoplanin developed drug fever.64 A dose to the destructive nature of the disease and patient factors rather of 6 mg/kg is advised for streptococcal endocarditis. Loading than inadequate antibiotic treatment, as no bacteriological doses should be given and a trough level of 15 mg/L estab- relapses were seen in patients with sensitive pneumococci trea- lished.65 We recommend just a single measurement to establish a ted with penicillin alone. There is little experience with high- therapeutic level if the renal function is stable. Following a level penicillin-resistant pneumococci, but a similar approach to single report of S. bovis carrying the vanB gene,66 a vancomycin the treatment of resistant pneumococcal meningitis is proposed, MIC should be measured for streptococci, rather than relying on 5.5 Streptococcus pyogenes (Group A streptococcus) Group A streptococci often behaves like S. aureus, causing acutedestructive endocarditis and right-sided disease in iv drug abu- Enterococci are responsible for  10% of endocarditis episodes, sers. Four weeks penicillin alone for these sensitive organisms an incidence that has remained stable since the introduction of has had a good success rate and is recommended.61 penicillin. Enterococcus faecalis accounts for the majority ofcases, but other species such as Enterococcus faecium, Entero- 5.6 Streptococcus agalactiae (Group B streptococcus), coccus durans and Enterococcus gallinarum are occasionallyimplicated. Older men (mean age, >60 years) are most com- monly affected, but younger women may also be affected in Group B streptococci can cause acute endocarditis with a high association with pregnancy. Over 40% of the patients have no mortality, often in patients with diabetes.62 Some strains are underlying heart disease. Although one study has shown worse tolerant and relatively penicillin resistant; there is very little outcomes in patients with more than 3 months of symptoms published on treatment. We therefore endorse the AHA cau- prior to presentation,67 a more recent analysis has failed to corro- tionary approach of recommending 2 weeks of aminoglycosides in addition to 4 weeks penicillin for even sensitive strains. As Treatment of native valve infection requires a minimum of information on endocarditis caused by group C and G strep- 4 weeks of iv antibiotics, whereas prosthetic valve endocarditis tococci is so limited, 4 weeks penicillin with 2 weeks gentamicin should be treated for a minimum of 6 weeks (see earlier com- ments on duration of therapy). Thereafter, the clinical response Table 4. Recommended regimens for treatment of enterococcal endocarditis caused by an ampicillin-susceptible(MIC < patient provided isolate isvancomycin susceptible (MIC < patient provided isolate isteicoplanin susceptible (MIC < aProvided isolate is high-level gentamicin-susceptible (MIC <_ 128 mg/L).
Table 5. Regimens for treatment of enterococcal endocarditis caused by high-level gentamicin-resistant (MIC > 128 mg/L)isolate patient or ampicillin-resistant(MIC > 8 mg/L) isolate patient or ampicillin-resistant(MIC > 8 mg/L) isolate aStreptomycin can be added if the isolate is not high-level resistant. If streptomycin is considered inappropriate or the isolate is streptomycinresistant, the cell-wall acting agent should be continued for a minimum of 8 weeks.
Table 6. Regimens for treatment of enterococcal endocarditis VanB phenotype (vancomycin resistant, teicoplanin susceptible) caused by an ampicillin-resistant isolate (MIC > 8 mg/L) have been treated with teicoplanin, but resistance may developduring monotherapy with this agent and treatment failures have The HACEK group of microorganisms includes Haemophilus parainfluenzae, Haemophilus influenzae, Haemophilus aphrophi-lus, Haemophilus paraphrophilus, Actinobacillus actinomycetem- If high-level gentamicin-susceptible (MIC < and Kingella spp. These fastidious extracellular Gram-negativebacteria cause an estimated 3% of all cases of IE.16,58,70–74 to therapy, inflammatory markers, repeat cultures and echocar- Previously, treatment was based on ampicillin and gentamicin.
diographic findings should guide the need for further anti- As a result of the emergence of b-lactamase-producing strains, a biotics. Recommended regimens for the more frequently b-lactamase-stable cephalosporin should be selected for empiri- encountered resistance patterns are given in Tables 4 – 6. Endo- cal treatment.12,43 Native valve IE should receive 4 weeks of carditis caused by penicillin-resistant, glycopeptide-resistant, treatment and those with prosthetic valve IE, 6 weeks.
infrequently. In the event of such a case, antimicrobial options include linezolid, or quinupristin/dalfopristin or combinations ofantibiotics according to in vitro susceptibility and new agents If amoxicillin sensitive, 2 g amoxicillin/ampicillin should be in development. Endocarditis caused by isolates displaying a administered iv 4 – 6 hourly, plus gentamicin 1 mg/kg body weight according to renal function (for the first 2 weeks only).
bacterial endocarditis, antifungal regimens that are fungicidal Monitoring should be regular. If amoxicillin resistant: ceftriax- may be preferable, although considerable work in terms of vali- one 2 g should be administered (to a maximum of 4 g) iv (once dation remains. Specific regimens must be given for a minimum of 6 weeks, but usually for much longer and in some circum-stances (e.g. prosthetic valves) therapy may be life long.
Amphotericin B does not penetrate well into vegetations, but has been used successfully in Candida endocarditis. It is, how- C. burnetii is an obligate intracellular pathogen. Endocarditis is ever, toxic, particularly if given for prolonged periods, and there the prime manifestation of chronic Q-fever16,75,76 and C. burnetii are few data concerning the efficacy of lipid-associated formu- causes up to 3% of all cases of IE in England and Wales.77 The estimated incidence of IE in those who contract Q-fever ranges fungistatic and is only active against some Candida spp., from 7%78 to 67%.79 Patients likely to develop Q-fever IE are Trichosporon spp. and some other yeasts. Flucytosine is also those with predisposing valvular damage or prosthetic heart fungistatic, although the combination of amphotericin plus flu- valves.80 Antibody titres should be determined every 6 months cytosine is more likely to be fungicidal. Flucytosine, however, is while on treatment and then every 3 months once treatment has associated with bone marrow toxicity and trough levels should been discontinued, for a minimum of 2 years. Long-term treat- not exceed 50 mg/L. Caspofungin is usually fungicidal for Can- ment (3 years) with tetracycline and a quinolone has been rec- dida spp. (although some isolates of Candida parapsilosis and Candida guilliermondii are less susceptible). However, there isas yet no experience of the use of caspofungin in endocarditis, and the penetration of caspofungin and other echinocandins intovegetations is unknown.
Doxycycline 100 mg should be administered once daily orally, Susceptibility testing must be undertaken for any fungus caus- plus ciprofloxacin 500 mg 8 hourly orally for at least 3 years.
ing endocarditis, including the determination of minimal fungici-dal concentrations. For drugs with variable bioavailability, therapeutic drug monitoring is important.
Bartonella spp. are facultative intracellular Gram-negativeaerobic bacteria that cause up to 3% of all cases of IE.14,16,83–89 100 mg/kg should be administered in four divided doses, Amoxicillin/ampicillin 2 g iv should be administered 4 – 6 according to renal function (first choice).
hourly, plus gentamicin 1 mg/kg 8 hourly. If penicillin allergic, use a tetracycline or a macrolide. Treatment should be continued Fluconazole 400 mg 12 hourly orally (second choice) Caspofungin 70 mg as a loading dose followed by 50 mg once daily, or 70 mg per day if weight >80 kg (first choice if intoler- ance or resistance precludes other options).
A wide range of other Gram-negative species with different sus- Voriconazole 6 mg/kg 12 hourly for two doses ceptibility patterns has been described and it is therefore difficult (loading) should be administered, then 4 mg/kg 12 hourly to recommend general treatment guidelines.3,16,58,90–108 Clinical intravenously or 400 mg 12 hourly for 24 h, followed by 200 mg experience has been well reviewed,16,91 and the contribution of animal studies to combination therapy has been described.96 The general themes are that a combination of antibiotics may offer Amphotericin B 1 mg/kg daily according to renal function.
synergy and prevent the emergence of resistance in long-termtreatment. The prognosis is generally poor, but best outcomesresult from using high-dose antibiotics (chosen after careful sen- sitivity testing) and early surgery.
The Working Party would like to thank Dr David Denningfor his input on the section relating to fungal endocarditis.
We would also like to thank Dr Ruth Rae and Dr Richard Fungal endocarditis is an unusual form of endocarditis compris- ing 2% – 4% of all cases. It is most common in iv drug abuseand prosthetic valve endocarditis, but has also been described in patients with neutropenia or haematological malignancy, follow- 1. Working Party of the British Society for Antimicrobial Chemo- ing cardiac surgery, solid organ transplantation and chronic gran- therapy (1998). Antibiotic treatment of streptococcal, enterococcal and ulomatous disease. The most common fungi causing endocarditis staphylococcal endocarditis. Heart 79, 207 – 10.
are Candida spp. and Aspergillus spp., with rarer examples 2. Durack, D. T., Lukes, A. S. & Bright, D. K. (1994). New criteria including Trichosporon spp. and Mucorales.
for diagnosis of infective endocarditis: utilization of specific echocardio- The treatment of fungal endocarditis is currently unsatisfac- graphic findings. Duke Endocarditis Service. American Journal of tory and usually requires surgical intervention. Analogous to 3. Bayer, A. S., Bolger, A. F., Taubert, K. A. et al. (1998).
23. Cremieux, A-C. & Carbon, C. (1992). Pharmacokinetic Diagnosis and management of infective endocarditis and its compli- and pharmacodynamic requirements for antibiotic therapy of experi- cations. Circulation 98, 2936 – 48.
mental endocarditis. Antimicrobial Agents and Chemotherapy 36, 4. Fournier, P. E., Casalta, J. P., Habib, G. et al. (1996).
Modification of the diagnostic criteria proposed by the Duke Endocardi- 24. Malacoff, R. F., Elliot, F. & Andriole, V. T. (1979). Streptococcal tis Service to permit improved diagnosis of Q fever endocarditis.
endocarditis (Nonenterococcal Non group A). Journal of the American American Journal of Medicine 100, 629 – 33.
Medical Association 241, 1807– 10.
5. Lamas, C. C. & Eykyn, S. J. (1997). Suggested modifications to 25. Wilson, W. R. & Geraci, J. E. (1985). Treatment of streptococ- the Duke criteria for the clinical diagnosis of native valve and prosthetic cal infective endocarditis. American Journal of Medicine 78, Suppl.
valve endocarditis: analysis of 118 pathologically proven cases. Clinical 26. Francioli, P. B. & Stamboulian, D. (1998). Outpatient treatment 6. Li, J. S., Sexton, D. J., Mick, N. et al. (2000). Proposed of infective endocarditis. Clinical Microbiology and Infection 4 Suppl. 3, modifications to the Duke criteria for the diagnosis of infective endocarditis. Clinical Infectious Diseases 30, 633 – 8.
27. Stamboulian, D., Bonvehi, P., Arevalo, C. et al. (1991).
7. Washington, J. A. (1987). The microbiological diagnosis of Antibiotic management of outpatients with endocarditis due to infective endocarditis. Journal of Antimicrobial Chemotherapy 20, penicillin-susceptible streptococci. Reviews of Infectious Diseases 13, 8. Chen, Y. C., Chang, S. C., Luh, K. T. et al. (1991). Actinobacillus 28. Francioli, P., Etienne, J., Hoigne´, R. et al. (1992). Treatment of actinomycetemcomitans endocarditis: a report of four cases and review streptococcal endocarditis with a single daily dose of ceftriaxone of the literature. Quarterly Journal of Medicine 81, 871 – 8.
sodium for 4 weeks. Efficacy and outpatient treatment feasibility.
9. Guyot, A., Bakhai, A., Fry, N. et al. (1999). Culture-positive Journal of the American Medical Association 267, 264 – 7.
Bartonella quintana endocarditis. European Journal of Clinical Micro- 29. Gengo, F. M., Mannion, T. W., Nightingale, C. H. et al. (1984).
biology and Infectious Diseases 18, 145 – 7.
Integration of pharmacokinetics and pharmacodynamics of methicillin 10. Lane, T., MacGregor, R. R., Wright, D. et al. (1983).
in curative treatment of experimental endocarditis. Journal of Anti- Cardiobacterium hominis: an elusive cause of endocarditis. Journal of microbial Chemotherapy 14, 619 – 31.
30. Sutherland, R., Croydon, E. A. P. & Rolinson, G. N. (1970).
11. Andrews, J. M. (2001). Determination of minimum inhibitory Flucloxacillin, a new isoxazolyl penicillin, compared with oxacillin, concentrations. Journal of Antimicrobial Chemotherapy 48, Suppl 1, cloxacillin and dicloxacillin. British Medical Journal 4, 455 – 60.
31. Frimodt-Moller, N., Espersen, F. & Rosdahl, V. T. (1987).
12. Kugler, K. C., Biedenbach, D. J. & Jones, R. N. (1999).
Antibiotic treatment of Staphylococcus aureus endocarditis. A review of Determination of the antimicrobial activity of 29 clinically important 119 cases. Acta Medica Scandinavica 222, 175 – 82.
compounds tested against fastidious HACEK group organisms.
32. Korzeniowski, O. & Sande, M. A. & The National Collaborative Diagnostic Microbiology and Infectious Disease 34, 73 – 6.
Endocarditis Study Group. (1982). Combination antimicrobial therapy 13. Houpikian, P. & Raoult, D. (2003). Diagnostic methods.
for Staphylococcus aureus endocarditis in patients addicted to Current best practices and guidelines for identification of difficult-to- parenteral drugs and in nonaddicts. Annals of Internal Medicine 97, culture pathogens in infective endocarditis. Cardiology Clinics 21, 33. Whitby, M. (1999). Fusidic acid in septicaemia and endocardi- 14. Raoult, D., Fournier, P. E., Drancourt, M. et al. (1996).
tis. International Journal of Antimicrobial Agents 12, 17 – 22.
Diagnosis of 22 new cases of Bartonella endocarditis. Annals of 34. Watanakunakorn, C. & Tisone, J. C. (1982). Antagonism between nafcillin or oxacillin and rifampin against Staphylococcus 15. Watkin, R. W., Lang, S., Lambert, P. A. et al. (2003). The aureus. Antimicrobial Agents and Chemotherapy 22, 920 – 2.
microbial diagnosis of infective endocarditis. Journal of Infection 47, 35. Calain, P., Krause, K. H., Vandaux, P. et al. (1987). Early termination of a prospective, randomised trial comparing teicoplanin 16. Brouqui, P. & Raoult, D. (2001). Endocarditis due to rare and and flucloxacillin for treating severe staphylococcal infections. Journal fastidious bacteria. Clinical Microbiology Reviews 14, 177 – 207.
of Infectious Diseases 155, 187 – 91.
17. Bosshard, P. P., Kronenberg, A., Zbinden, R. et al. (2003).
36. Davey, P. G. & Williams, A. H. (1991). Teicoplanin mono- Etiologic diagnosis of infective endocarditis by broad-range polymerase therapy of serious infections caused by Gram-positive bacteria: a chain reaction: a 3-year experience. Clinical Infectious Diseases 37, re-evaluation of patients with endocarditis or Staphylococcus aureus bacteraemia from a European open trial. Journal of Antimicrobial 18. Fenollar, F., Gauduchon, V., Casalta, J. P. et al. (2004).
Chemotherapy 27, Suppl. B, 43 – 50.
Mycoplasma endocarditis: two case reports and a review. Clinical 37. Gilbert, D. N., Wood, C. A., Kimvrough, R. C. et al. (1991).
Failure of treatment with teicoplanin at 6 milligrams/kilogram/day in 19. Gauduchon, V., Chalabreysse, L., Etienne, J. et al. (2003).
patients with Staphylococcus aureus intravascular infection. Antimicro- Molecular diagnosis of infective endocarditis by PCR amplification and bial Agents and Chemotherapy 35, 79 – 87.
direct sequencing of DNA from valve tissue. Journal of Clinical 38. Gentry, C. A., Rodvold, K. A., Novak, R. M. et al. (1997).
Retrospective evaluation of therapies for Staphylococcus aureus 20. Grijalva, M., Horvath, R., Dendis, M. et al. (2003). Molecular endocarditis. Pharmacotherapy 17, 990 – 7.
diagnosis of culture negative infective endocarditis: clinical validation in 39. Gonzalez, C., Rubio, M., Romero-Vivas, J. et al. (1999).
a group of surgically treated patients. Heart 89, 263 – 8.
Bacteremic pneumonia due to Staphylococcus aureus: a comparison 21. McCracken, D., Barnes, R., Poynton, C. et al. (2003).
of disease caused by methicillin-resistant and methicillin-susceptible Polymerase chain reaction aids in the diagnosis of an unusual case of organisms. Clinical Infectious Diseases 29, 1171– 7.
Aspergillus niger endocarditis in a patient with acute myeloid 40. Bayer, A. S. & Lam, K. (1985). Efficacy of vancomycin leukaemia. Journal of Infection 47, 344 – 7.
plus rifampin in experimental aortic-valve endocarditis due to 22. Voldstedlund, M., Pedersen, L. N., Fuursted, K. et al. (2003).
methicillin-resistant Staphylococcus aureus: in vitro-in vivo correlations.
Different polymerase chain reaction-based analyses for culture- Journal of Infectious Diseases 151, 157 – 65.
negative endocarditis caused by Streptococcus pneumoniae. Scanda- 41. Levine, D. P., Fromm, B. S. & Reddy, B. R. (1991).
navian Journal of Infectious Diseases 35, 757 – 9.
Slow response to vancomycin or vancomycin plus rifampin in methicillin-resistant Staphylococcus aureus endocarditis. Annals of 59. Aronin, S., Mukherje, S., West, J. et al. (1998). Review of pneumococcal endocarditis in adults in the penicillin era. Clinical 42. Fantin, B., Leclercq, R., Duval, J. et al. (1993). Fusidic acid alone or in combination with vancomycin for therapy of experimental 60. Siegel, M. & Timpone, J. (2001). Penicillin-resistant Streptococ- endocarditis due to methicillin-resistant Staphylococcus aureus. Anti- cus pneumoniae endocarditis: a case report and review. Clinical microbial Agents and Chemotherapy 37, 2466 – 9.
43. Wilson, W. R., Karchmer, A. W., Dajani, A. S. et al. (1995).
61. Burkert, T. & Watanakunakorn, C. (1991). Group A streptococ- Antibiotic treatment of adults with infective endocarditis due to cus endocarditis: a report of five cases and review of the literature.
streptococci, enterococci, staphylococci, and HACEK microor- Journal of Infection 23, 307 – 16.
ganisms. Journal of the American Medical Association 274, 62. Gallagher, P. G. & Watanakunakorn, C. (1986). Group B streptococcus endocarditis: report of seven cases and review of the 44. Torres-Tortosa, M., de Cueto, M., Vergara, A. et al. (1994).
literature, 1962-1985. Reviews of Infectious Diseasess 8, 175 – 88.
Prospective evaluation of a two-week course of intravenous 63. Karchmer, A. W. & Longworth, D. L. (2003). Infections of antibiotics in intravenous drug addicts with infective endocarditis.
intracardiac devices. Cardiology Clinics 21, 253 – 71.
European Journal of Clinical Microbiology and Infectious Diseases 64. Venditti, M., Gelfusa, V., Serra, P. et al. (1992). 4-week 45. Abrams, B., Sklaver, A., Hoffman, T. et al. (1979). Single or high-dose teicoplanin. Antimicrobial Agents and Chemotherapy 36, combination therapy of staphylococcal endocarditis in intravenous drug abusers. Annals of Internal Medicine 90, 789 – 91.
65. Wilson, A. P. R. & Gaya, H. (1996). Treatment of endocarditis 46. Chambers, H. F., Miller, T. F. & Newman, M. D. (1988). Right- with teicoplanin: a retrospective analysis of 104 cases. Journal of sided Staphylococcus aureus endocarditis in intravenous drug abusers: Antimicrobial Chemotherapy 38, 507 – 21.
two-week combination therapy. Annals of Internal Medicine 109, 66. Poyart, C., Pierre, C., Quesne, G. et al. (1997). Emergence of vancomycin resistance in the genus streptococcus: characterization of 47. Fortun, J., Navas, E., Martinez-Beltran, J. et al. (2001). Short a vanB transferable determinant in Streptococcus bovis. Antimicrobial course therapy for right-side endocarditis due to Staphylococcus Agents and Chemotherapy 41, 24– 9.
aureus in drug abusers: cloxacillin versus glycopeptides in combination 67. Wilson, W. R., Wilkowske, C. J., Wright, A. et al. (1984).
with gentamicin. Clinical Infectious Diseases 33, 120 – 5.
Treatment of streptomycin-susceptible and streptomycin-resistant 48. Suter, F., Maserati, R., Carnevale, G. et al. (1984). Manage- enterococcal endocarditis. Annals of Internal Medicine 100, 816 – 23.
ment of staphylococcal endocarditis in drug addicts. Combined therapy 68. Olaison, L. & Schadewitz, K. (2002). Enterococcal endocarditis with oral rifampicin and aminoglycosides. Journal of Antimicrobial in Sweden, 1995-1999: can shorter therapy with aminogylcosides be used? Clinical Infectious Diseases 34, 159 – 66.
49. Heldman, A. W., Hartert, T. V., Ray, S. C. et al. (1996). Oral 69. Furlong, W. B. & Rakowski, T. A. (1997). Therapy with RP antibiotic treatment of right-sided staphylococcal endocarditis in 59500 (quinupristin/dalfopristin) for prosthetic valve endocarditis due to injection drug users: prospective randomised comparison with parent- enterococci with VanA/VanB resistance patterns. Clinical Infectious eral therapy. American Journal of Medicine 101, 68 – 76.
50. Parker, R. H. & Fossieck, B. E. (1980). Intravenous followed by 70. Lepori, M., Bochud, P. Y., Owlya, R. et al. (2001). Endocarditis oral antimicrobial therapy for staphylococcal endocarditis. Annals of due to HACEK bacteria. A case report of endocarditis due to Kingella kingae. Revue Medicale de la Suisse Romande 121, 47 – 50.
51. Birmingham, M. C., Rayner, C. R., Meagher, A. K. et al.
71. Castillo, J. C., Anguita, M. P., Ramirez, A. et al. (2000). Long (2003). Linezolid for the treatment of multidrug-resistant. Gram-positive term outcome of infective endocarditis in patients who were not drug infections: experience from a compassionate-use program. Clinical addicts: a 10 year study. Heart 83, 525 – 30.
72. Das, M., Badley, A. D., Cockerill, F. R. et al. (1997). Infective 52. Drew, R. H., Perfect, J. R., Srinath, L. et al. (2000). Treatment endocarditis caused by HACEK microorganisms. Annual Review of of methicillin-resistant Staphylococcus aureus infections with quinupris- tin-dalfopristin in patients intolerant of or failing prior therapy. For the 73. el Khizzi, N., Kasab, S. A. & Osoba, A. O. (1997). HACEK Synercid Emergency-Use Study Group. Journal of Antimicrobial group endocarditis at the Riyadh Armed Forces Hospital. Journal of 53. Bengtsson, E., Svanbom, M. A. J. & Tunevall, G. (1974).
74. Steelman, R., Einzig, S., Balian, A. et al. (2000). Increased Trimethroprim-sulphamethoxazole treatment in staphylococcal endo- susceptibility to gingival colonization by specific HACEK microbes in carditis and Gram-negative septicaemia. Scandanavian Journal of children with congenital heart disease. Journal of Clinical Pediatric 54. Chayakul, P. & Yipintsoi, T. (1993). Intravenous followed by 75. Raoult, D., Tissot-Dupont, H., Foucault, C. et al. (2000). Q fever oral antimicrobial therapy for staphylococcal endocarditis. Journal of 1985-1998. Clinical and epidemiologic features of 1,383 infections.
the Medical Association of Thailand 76, 559 – 63.
55. Rouse, M. S., Wilcox, R. M., Henry, N. K. et al. (1990).
76. Maurin, M. & Raoult, D. (1999). Q fever. Clinical Microbiology Ciprofloxacin therapy of experimental endocarditis caused by methicillin-resistant Staphylococcus epidermidis. Antimicrobial Agents 77. Palmer, S. R. & Young, S. E. (1982). Q-fever endocarditis in England and Wales, 1975-81. Lancet 2, (8313), 1448 – 9.
56. Tamer, M. A. & Bray, J. D. (1982). Trimethroprim-sulpha- 78. Connolly, J. H., Coyle, P. V., Adgey, A. A. et al. (1990). Clinical methoxazole treatment of multiantibiotic-resistant staphylococcal endo- Q fever in Northern Ireland 1962-1989. Ulster Medical Journal 59, carditis and meningitis. Clinical Pediatrics 21, 125 – 6.
57. Ballet, M., Gevigney, G., Gare, J. P. et al. (1995). Infective 79. Brouqui, P., Dupont, H. T., Drancourt, M. et al. (1993). Chronic endocarditis due to Streptococcus bovis. European Heart Journal 16, Q fever. Ninety-two cases from France, including 27 cases without endocarditis. Archives of Internal Medicine 153, 642 – 8.
58. Berbari, E. F., Cockerill, F. R. & Steckelberg, J. M. (1997).
80. Siegman-Igra, Y., Kaufman, O., Keysary, A. et al. (1997).
Infective endocarditis due to unusual or fastidious microorganisms.
Q fever endocarditis in Israel and a worldwide review. Scandinavian Mayo Clinic Proceedings 72, 532 – 42.
Journal of Infectious Diseases 29, 41 – 9.
81. Levy, P. Y., Drancourt, M., Etienne, J. et al. (1991).
95. Crum, N. F., Utz, G. C. & Wallace, M. R. (2002). Stenotropho- Comparison of different antibiotic regimens for therapy of 32 cases of monas maltophilia endocarditis. Scandanavian Journal of Infectious Q fever endocarditis. Antimicrobial Agents and Chemotherapy 35, 96. Daikos, G. L., Kathpalia, S. B., Lolans, V. T. et al. (1988). Long- 82. Raoult, D., Houpikian, P., Tissot, D. H. et al. (1999). Treatment term oral ciprofloxacin: experience in the treatment of incurable infective of Q fever endocarditis: comparison of 2 regimens containing endocarditis. American Journal of Medicine 84, 786 – 90.
doxycycline and ofloxacin or hydroxychloroquine. Archives of Internal 97. Fantin, B. & Carbon, C. (1992). In vivo antibiotic synergism: contribution of animal models. Antimicrobial Agents and Chemotherapy 83. Breathnach, A. S., Hoare, J. M. & Eykyn, S. J. (1997). Culture- negative endocarditis: contribution of Bartonella infections. Heart 77, 98. Felner, J. M. & Dowell, V. R. (1970). Anaerobic bacterial endocarditis. New England Journal of Medicine 283, 1188 – 970.
84. Doern, G. V. (2000). Detection of selected fastidious bacteria.
99. Farrugia, D. C., Eykyn, S. J. & Smyth, E. (1994). Campylobac- Clinical Infectious Diseases 30, 166 – 73.
ter fetus endocarditis: two case reports and review. Clinical Infectious 85. Drancourt, M., Mainardi, J. L., Brouqui, P. et al. (1995).
Bartonella (Rochalimaea) quintana endocarditis in three homeless 100. Gilleece, A. & Fenelon, L. (2000). Nosocomial infective men. New England Journal of Medicine 332, 419 – 23.
endocarditis. Journal of Hospital Infection 46, 83 – 8.
86. La Scola, B. & Raoult, D. (1999). Culture of Bartonella 101. Gradon, J. D., Chapnick, E. K. & Luwick, L. I. (1992). Infective quintana and Bartonella henselae from human samples: a 5-year endocarditis of a native valve due to Acinetobacter: case report and experience (1993 to 1998). Journal of Clinical Microbiology 37, review. Clinical Infectious Diseases 14, 1145– 8.
102. Joly, V., Pangon, B., Vallois, J. M. et al. (1987). Value of 87. Maurin, M., Birtles, R. & Raoult, D. (1997). Current knowledge antibiotic levels in serum and cardiac vegetations for predicting of Bartonella species. European Journal of Clinical Microbiology and antibacterial effect of ceftriaxone in experimental Escherichia Infectious Diseases 16, 487 – 506.
coli endocarditis. Antimicrobial Agents and Chemotherapy 31, 1632– 9.
88. Pesanti, E. L. & Smith, I. M. (1979). Infective endocarditis with 103. Morrison, D. J., Sperling, L., Schwartz, D. A. et al. (1997).
negative blood cultures. An analysis of 52 cases. American Journal of Escherichia coli endocarditis of a native aortic valve. Archives in Pathology and Laboratory Medicine 121, 1292– 5.
89. Raoult, D., Fournier, P. E., Vandenesch, F. et al. (2003).
104. Mylonakis, E. & Calderwood, S. (2001). Infective endocarditis Outcome and treatment of Bartonella endocarditis. Archives of Internal in adults. New England Journal of Medicine 18, 1318 – 29.
105. Nastro, L. J. & Finegold, S. M. (1973). Endocarditis due to 90. Black, J. R. & Brint, J. M. (1988). Successful treatment anaerobic Gram-negative bacilli. American Journal of Medicine 54, of gonococcal endocarditis with ceftriaxone. Journal of Infectious 106. Rodriguea, C., Olcoz, M. T., Izquierdo, G. et al. (1990).
91. Carbon, F., Gutmann, L., Bure, A. et al. (1990). Ceftriaxone- Endocarditis due to ampicillin-resistant nontyphoidal Salmonella: cure sulbactam combination in rabbit endocarditis caused by a strain with a third generation cephalosporin. Reviews of Infectious Diseases TEM-3 b-lactamase. Antimicrobial Agents and Chemotherapy 34, 107. Synder, N., Atterbury, C. E., Correia, J. P. et al. (1977).
Increased concurrence of cirrhosis and bacterial endocarditis. Gastro- 92. Cohen, P. S., Maguire, J. H. & Weinstein, L. (1980). Infective endocarditis caused by Gram-negative bacteria: a review of the 108. Tellez, I., Chrysant, G. S., Omer, I. et al. (2000). Citrobacter literature, 1945-1977. Progress in Cardiovascular Diseases 22, diversus endocarditis. American Journal of the Medical Sciences 320, 93. Cooper, R. & Mills, J. (1980). Serratia endocarditis. Archives of 109. Pappas, P. G., Rex, J. H., Sobel, J. D. et al. (2004). Guidelines for the treatment of candidiasis. Clinical Infectious Diseases 38, 94. Crane, L. R., Levine, D. P., Zervos, M. J. et al. (1986).
Bacteremia in narcotic addicts at the Detroit Medical Center.
110. Stevens, D., Kan, V., Judson, M. et al. (2000). Practice Microbiology, epidemiology, risk factors and empiric therapy. Reviews guidelines for the diseases caused by Aspergillus. Clinical Infectious of Infectious Diseases 8, 364 – 73.

Source: http://www.webdoki.hu/dokumentumok/Endocarditis2004JAC.pdf