Portal de Periódicos da CAPES

Microbiological recommendations for the diagnosis and follow-up of infective endocarditis

1998; Elsevier BV; Volume: 4; Linguagem: Inglês

10.1111/j.1469-0691.1998.tb00862.x

1469-0691

Ernö Gutschik,

Streptococcal Infections and Treatments

These recommendations are based on appreciation of new knowledge of the pathogenesis of the disease gained by endocarditis research, basic principles of collection of clinical specimens for laboratory testing, estimation of the number of cultures needed for diagnosis and control of antimicrobial treatment, handling of clinical specimens in the microbiological laboratory and the principles of antimicrobial susceptibility testing of the microbial isolates. Considering culture-dependent and culture-independent methods, a large number of conventional systems, media, etc. have been described in an effort to increase the speed and sensitivity of microbial detection and have been the subject of numerous, sometimes conflicting, studies. An evaluation of these methods, media, etc. is outside of the scope of this overview. The recommendations presented here describe the most significant and generally accepted procedures which should be considered in every case of infective endocarditis. Early studies demonstrated that the bacteremia of endocarditis is often continuous. There is no significant advantage to obtaining blood cultures at any particular body temperature. Furthermore, there is no significant advantage of arterial over venous blood [1Beeson PB Brannon ES Warren JV Observations on the site of removal of bacteria from the blood of patients with bacterial endocarditis.J Exp Med. 1945; 81: 9-23Crossref PubMed Scopus (112) Google Scholar]. The number of blood cultures and the timing and volume of blood collection are important variables (Table 1). In order to obtain a reliable high detection rate of positive blood cultures, to demonstrate the continuous nature of septicemia and to ensure the possibility of discrimination between 'endocarditis strains' and contaminants, it is recommended to draw 20 mL blood in each of three separate samplings from adults over 1–24 h [2Washington JA Blood cultures: an overview.Eur J Clin Microbiol Infect Dis. 1989; 9: 803-806Crossref Scopus (19) Google Scholar]. Cultures should not be taken through indwelling vascular catheters because of the risk of contaminants.Table 1Blood cultures: collection of blood for diagnostic purposesBlood cultures: three or more separate venepunctures over 1–24 h and according to the clinical presentation of the patient. Perform either two or more blood cultures separated by at least 12 h, or three or more blood cultures with at least 1 h between the first and the last cultureAt least 20 mL of blood should be drawn in each sampling from adultsIt is recommended to utilize a pyridoxal-enriched medium. Two or three additional blood cultures are recommended at 48 h in cases with negative initial blood culture results and a high degree of clinical suspicion of infective endocarditis Open table in a new tab In a recent study of the clinical significance of positive blood cultures [3Weinstein MP Towns ML Quartey SM et al.The clinical significance of positive blood cultures in the 1990s: a prospective comprehensive evaluation of the microbiology, epidemiology, and outcome of bacteremia and fungemia in adults.Clin Infect Dis. 1997; 24: 584-602Crossref PubMed Scopus (1020) Google Scholar], 41.5% of the isolates were judged to represent contamination and another 5.3% were of indeterminate clinical significance. Generally, 81.9% of coagulase-negative staphylococci, and 49.3% of viridans streptococci were indicated as contaminants. This study demonstrates that contamination in blood cultures is more frequent in the 1990s than it was two decades ago. The reason for this, among others, is more frequent use of intravascular and prosthetic devices, and probably newer blood culture systems and media which are more sensitive and capable of detecting viable skin bacteria even after appropriate antiseptic preparation before collection of the blood cultures. This observation underscores the significance of three or more separate venepunctures in order to demonstrate the continuous seeding of blood from the infected heart valve, and to discriminate between true pathogens and contaminants (Table 1). The rate of positive culture may be influenced by the type of microorganism involved. Generally, the bacteremias are low-grade, half of the bacteremic patients having less than one organism per mL blood [4Pierce G Murray PR Current controversies in the detection of septicemia.Eur J Clin Microbiol. 1986; 5: 487-491Crossref PubMed Scopus (10) Google Scholar]. In addition, endocarditis may also be caused by microorganisms which are difficult to culture using routine techniques. To maximize the detection of nutritionally deficient streptococci, it is recommended to utilize pyridoxal-enriched medium [5Roberts RB Krieger AG Schiller NL Gross KC Viridans streptococcal endocarditis: the role of various species, including pyridoxal-dependent streptococci.Rev Infect Dis. 1979; 1: 955-966Crossref PubMed Scopus (159) Google Scholar]. Special diagnostic measures should be taken in the case of negative initial blood culture results (Table 1). If the patient received prior antibiotic therapy, it is worth obtaining serial blood cultures off antibiotic therapy. If antibiotics are only given 2–3 days prior to their discontinuation, blood cultures will probably become rapidly positive [6Van Scoy RE Culture-negative endocarditis.Mayo Clin Proc. 1982; 57: 149-154PubMed Google Scholar]. In cases where there is a high degree of clinical suspicion of endocarditis, especially for Staphylococcus aureus, fungemia or anaerobic bacteria, the use of a lysis-centrifugation system should be considered [7Washington JA Ilstrup DM Blood cultures: issues and controversies.Rev Infect Dis. 1986; 8: 792-802Crossref PubMed Scopus (212) Google Scholar,8Heimdahl A Hall G Hedberg M et al.Detection by lysisfiltration of bacteremia after different oral surgical procedures.J Clin Microbiol. 1990; 28: 2205-2209PubMed Google Scholar]. Conventional cultures (vented or unvented) are usually incubated at 35–37°C in 5–10% CO2 and checked daily for visible growth (Table 2). Routine Gram-staining of negative cultures will rarely detect positive cultures because the sensitivity of Gram stains is only slightly better than that of macroscopic examination [4Pierce G Murray PR Current controversies in the detection of septicemia.Eur J Clin Microbiol. 1986; 5: 487-491Crossref PubMed Scopus (10) Google Scholar] and, in addition, an opening (or needle perforation) of the bottles will facilitate possible laboratory contamination. Terminal blind subculture after 7 days is recommended by some centers [4Pierce G Murray PR Current controversies in the detection of septicemia.Eur J Clin Microbiol. 1986; 5: 487-491Crossref PubMed Scopus (10) Google Scholar].Table 2Blood cultures: handling of blood cultures in the clinical microbiology laboratoryBlood cultures should be checked for growth for a minimum of 5 days, with daily examination and a further examination on day 7In cases where there is a strong index of suspicion, the cultures should be incubated until, and retested on, day 14 before being reported as sterileNo special recommendation exists for blood cultures obtained from patients with prosthetic valves. Open table in a new tab Murray [9Murray PR Determination of the optimum incubation period of blood culture broths for the detection of clinically significant septicemia.J Clin Microbiol. 1985; 21: 481-485PubMed Google Scholar] investigated the value of incubating blood culture broths for more than 7 days, using 20 000 blood cultures from a tertiary-care hospital. He found that, although incubation of blood culture bottles for 7 days will not detect some fastidious bacteria and fungi, routine incubation for a longer period is not justified. In cases where there is a strong index of suspicion for infective endocarditis in spite of negative culture results, communication between the microbiology laboratory and the clinical staff is vital for making decisions regarding further incubation and, perhaps, specific measures concerning methods and media for detection of fastidious Gram-negative rods, such as in the HACEK (Haemophilus, Actinobacillus, Cardiobacterium, Eikenella, Kingella) group, nutritionally deficient streptococci, bacterial L-forms (cell wall-deficient bacteria), Brucella spp., Neisseria spp., Legionella spp., mycobacteria, nocardia, rickettsial species, Bartonella spp., Chlamydia spp. and fungi [10Kay KM Kay D Laboratory findings including blood cultures.in: Kaye D Infective endocarditis. 2nd edn. Raven Press Ltd, New York1992: 117-124Google Scholar]. The incubation period of 14 days for such a situation is arbitrarily chosen by the Working Group (Table 2) but will probably represent an optimal time period for detection of the fastidious microorganisms [2Washington JA Blood cultures: an overview.Eur J Clin Microbiol Infect Dis. 1989; 9: 803-806Crossref Scopus (19) Google Scholar,4Pierce G Murray PR Current controversies in the detection of septicemia.Eur J Clin Microbiol. 1986; 5: 487-491Crossref PubMed Scopus (10) Google Scholar]. Within several days of initiation of appropriate antimicrobial therapy, bacteremia resolves and there is no need for additional blood cultures. In one large prospective trial of Staphylococcus aureus endocarditis, blood was cultured daily for the first 7 days of treatment and weekly thereafter. The mean duration of bacteremia in non-addicts was 2.8 ±1.4 days for nafcillin plus gentamicin, and 4.1 ±1.4 days for nafcillin. The clinical response was rapid and the patients were febrile for approximately the same amount of time [11Korzeniowski O Sande MA Combination antimicrobial therapy for Staphylococcus aureus endocarditis in patients addicted to parenteral drugs and in nonaddicts.Ann Intern Med. 1982; 97: 496-503Crossref PubMed Scopus (299) Google Scholar]. Thus, unchanged febrile illness beyond the first week of treatment warrants a reassessment of the antibiotic prescription [11Korzeniowski O Sande MA Combination antimicrobial therapy for Staphylococcus aureus endocarditis in patients addicted to parenteral drugs and in nonaddicts.Ann Intern Med. 1982; 97: 496-503Crossref PubMed Scopus (299) Google Scholar]. Additional blood cultures are recommended in these cases (Table 3), with special attention given to metastatic lesions or valve ring abscesses. This is uncommon with pathogens of low virulence such as viridans streptococci, coagulase-negative staphylococci or other highly antibiotic-sensitive organisms, but occurs more often during treatment for Staphylococcus aureus endocarditis. Septicemia and, on occasion, valvular superinfection with bacteria resistant to antibiotic treatment or Candida spp. as a complication of the use of intravascular catheters should also be considered. Negative blood cultures, on the other hand, do not support a claim of bacteriologic cure of infective endocarditis, and thus no monitoring of efficacy is recommended (Table 3).Table 3Blood cultures: cultures during and after therapyNot recommended in cases with a good clinical responseIn cases with no clinical response or delayed response beyond what is expected for the initial isolates, additional blood cultures may be useful and should be obtained at the time corresponding to trough antibiotic levelsNo monitoring of efficacy (therapeutic 'window') is recommendedFollow-up cultures: After standard therapy, and in the absence of symptoms, blood cultures are not recommended. Physicians, however, should be alert to early signs of relapse Open table in a new tab Follow-up cultures just after standard therapy and good clinical response are not recommended. In most relapses, renewed symptoms, signs of infection and bacteremia appear 2–4 weeks after completion of therapy [12Santoro J Ingerman M Response to therapy: relapses and reinfections.in: Kaye D Infective endocarditis. 2nd edn. Raven Press Ltd, New York1992: 423-433Google Scholar]. Thus, follow-up cultures are only recommended when there are signs of relapse. 'Typical microorganisms' comprise one of the major Duke criteria [13Lukes A, Durack DT. Distribution of diagnostic criteria in different groups of patients with endocarditis [abstract 5–6]. In: Abstracts of the 2nd International Symposium on Modern Concepts in Endocarditis, Elsinore, Denmark: Kandrup, 1993: 56.Google Scholar]. Thus, speciation of microorganisms is necessary for diagnostic purposes. Typical organisms are viridans streptococci, Streptococcus bovis, and the HACEK group of fastidious Gram-negative organisms. Speciation of streptococci (and other genera) is useful too, because within the viridans group, and streptococci in general, important differences in the propensity to cause endocarditis have been noted [14Parker MT Ball LC Streptococci and aerococci associated with systemic infection in man.J Med Microbiol. 1975; 9: 275-302Crossref Scopus (255) Google Scholar]. The same consideration seems to be valid for coagulase-negative staphylococci. Indeed, positive blood cultures yielding coagulase-positive staphylococci may present great interpretative difficulties, even when more than one culture set is positive (Table 1). The same biotype and the same antibiogram do not necessarily represent a true pathogen [3Weinstein MP Towns ML Quartey SM et al.The clinical significance of positive blood cultures in the 1990s: a prospective comprehensive evaluation of the microbiology, epidemiology, and outcome of bacteremia and fungemia in adults.Clin Infect Dis. 1997; 24: 584-602Crossref PubMed Scopus (1020) Google Scholar]. In certain circumstances, such as prosthetic valve endocarditis, molecular typing techniques are recommended to confirm the identity of the isolates (Table 4).Table 4Identification of microorganismsSpeciation of the microorganism is necessary for epidemiologic purposes and to adequately interpret the portal of entry and possible recurrencesIt is recommended that blood strains isolated from patients with endocarditis should be stored in order to recognize relapsesIdentical species designations (biotypes and antibiograms) should be demonstrated for coagulase-negative staphylococci, and, in special circumstances, molecular typing techniques should be used to confirm the identity of the isolatesSpeciation should also be performed in order to establish the antibiotic susceptibility pattern of the pathogen and for specific additional investigations of associated diseasesThe viridans streptococci should be differentiated from Streptococcus bovisAmong the enterococci, differentiation of Enterococcus faecalis and Enterococcus faecium is necessary Open table in a new tab Furthermore, speciation is necessary for epidemiologic purposes, and it is mandatory to diagnose recurrences (Table 4). Recent efforts to provide a simple diagnostic testing system have allowed most laboratories to speciate these organisms, but speciation or further taxonomic positioning can be done by reference laboratories. It is important to store all endocarditis isolates in order to recognize relapses. Isolates should be kept refrigerated at –80°C for 1 year or more. Speciation of the isolates early in the diagnostic procedure can be necessary to establish the validity of the antibiotic susceptibility pattern and to decide the type of antibiotic for antimicrobial monotherapy or combination therapy. Among enterococci, the differentiation between Enterococcus faecalis and Enterococcus faecium is mandatory for selection of a synergistic β-lactam–aminoglycoside regimen. It is important to notice that Enterococcus faecium strains resist synergistic killing by penicillin in combination with kanamycin, tobramycin and netilmicin even in the absence of high-level resistance to these agents [15Eliopoulos GM Enterococcal endocarditis.in: Kaye D Infective endocarditis. 2nd edn. Raven Press Ltd, New York1992: 209-223Google Scholar]. Streptococcus bovis should be differentiated from other streptococci (Table 4) because adults with these bacteremic isolates have a significant incidence of underlying gastrointestinal neoplasm [16Klein RS Catalano MT Edberg SC et al.Streptococcus bovis septicemia and carcinoma of the colon.Ann Intern Med. 1979; 91: 560-562Crossref PubMed Scopus (183) Google Scholar,17Hønberg PZ Gutschik E Streptococcus bovis bacteremia and its association with alimentary-tract neoplasm.Lancet. 1987; 1: 163-164Abstract PubMed Scopus (23) Google Scholar]. The consequence of this diagnosis should be that all patients with Streptococcus bovis bacteremia with or without endocarditis must undergo diagnostic studies, including colonoscopy, to rule out an occult colonic malignancy, and, if the initial study is negative, tests should be repeated 6 months later. Optimal therapy of infective endocarditis requires bactericidal antimicrobial agents. The most reliable bactericidal agents include: penicillins, cephalosporins, aminoglycosides, fluoroquinolones and rifampicin. In some cases, combination therapy is needed for enhancement of bacterial activity. For fungal endocarditis there is no routine, standardized sensitivity test available. Minimal inhibitory concentration (MIC) determination is preferred for susceptibility testing. The MIC is a prerequisite for demonstrating full sensitivity of viridans streptococci to penicillin or testing enterococci for high-level resistance to aminoglycosides (Table 5). Disk testing will suffice in most cases as an initial testing procedure in uncomplicated cases, when full sensitivity is demonstrated for the drug usually preferred for treatment. The method of disk sensitivity testing is not uniform between different laboratories. However, the sensitivity (breakpoint) is usually set at one-sixteenth to one-fourth of the peak levels of drug available in serum [18Neu HC General concepts on the chemotherapy of infectious diseases.Med Clin North Am. 1987; 71: 1065-1078PubMed Google Scholar]. Since MICs of the organisms are usually far below this point, disk susceptibility testing is acceptable as the (initial) key to choosing an appropriate antimicrobial dosing regimen. MIC susceptibility testing should be carried out as soon as possible for relevant antibiotics in order to re-evaluate the therapeutic regimen.Table 5Antimicrobial susceptibility testing: recommendation of standardized sensitivity testsDetermination of MIC is recommended when available, but disk testing will suffice in most cases as an initial test procedureMIC determinations for penicillin and in some cases vancomycin, teicoplanin, aminoglycoside antibiotics and quinolones are necessary for the choice of therapeutic regimenStaphylococci: test for methicillin resistanceViridans streptococci: demonstrate full sensitivity to penicillinHACEK: test for β-lactamase productionEnterococci: test for high-level resistance to streptomycin (≥2000 mg/L) and gentamicin (≥500 mg/L). Routine susceptibility testing should include glycopeptide antibiotics for Gram-positive organisms Open table in a new tab A simple diagnostic test system for β-lactamase production is available in most laboratories, e.g. the nitrocefin test [19O'Callaghan CH Morris HA Kirkby SM et al.Novel method for detection of beta-lactamases by using a chromogenic cephalosporin substrate.Antimicrob Agents Chemother. 1972; 1: 283-288Crossref PubMed Scopus (1478) Google Scholar]. Routine testing for β-lactamase production for all of the HACEK group of isolates is recommended (Table 5). Disk testing and MIC determination measure only the inhibitory activity of the agents. It seems logical to have additional laboratory methods that assess the bactericidal activity of the antimicrobial agents. Minimal bactericidal concentrations (MBCs), which are the concentrations needed to kill the majority of viable microorganisms under a given set of conditions, can be used to define tolerance. Tolerance is the situation in which a particular organism shows a substantial difference (32-fold or more) between the MIC and MBC for an antibiotic which is normally bactericidal. However, a number of technical factors can have a significant effect upon the outcome and the reproducibility of bactericidal testing [20Stratton CW The role of the microbiology laboratory in the treatment of infective endocarditis.J Antimicrob Chemother. 1987; 20 (suppl A): 41-49Crossref PubMed Scopus (13) Google Scholar]. Although tolerance is described for staphylococcal and for streptococcal strains and the significance of this phenomenon has been shown in the animal models of endocarditis [21Voorn GP Thompson J Goessens WHF Schmall-Bauer W Broeders PHM Michel ME Role of tolerance in cloxacillin prophylaxis of experimental Staphylococcus aureus endocarditis.J Infect Dis. 1992; 166: 169-173Crossref PubMed Scopus (9) Google Scholar, 22Brennan RO Durack DT Therapeutic significance of penicillin tolerance in experimental streptococcal endocarditis.Antimicrob Agents Chemother. 1983; 23: 273-277Crossref PubMed Scopus (48) Google Scholar, 23Meeson J McColm AA Acred P Greenwood D Differential response to benzylpenicillin in vitro of tolerant and non-tolerant variants of Streptococcus sanguis II.J Antimicrob Chemother. 1990; 25: 103-109Crossref PubMed Scopus (13) Google Scholar], tolerance has not been shown to result in decreased cure rates in humans. Thus, MBC is not routinely recommended by the Working Group (Table 6). Killing curves constitute an excellent method to test the efficacy of combinations of drugs, and for evaluation of synergy and antagonism. This method is, however, time-consuming and not recommended.Table 6Antimicrobial susceptibility testing: additional sensitivity testsMBC: not routinely recommendedKilling curves: not routinely recommended. However, in relapses, the use of unusual antibiotic combinations, new antibiotics for research, etc., is justifiedSerum bactericidal test: not routinely recommendedSerum inhibitory test: not routinely recommended. Its use in therapy failures, relapses and unusual therapeutic regimens is still being debated Open table in a new tab The serum inhibitory test (SIT) and serum bactericidal test (SBT) measure the inhibitory capacity or the bactericidal activity of blood during antimicrobial treatment of endocarditis. These tests are subject to the same problems as MBC, and their usefulness is hampered by lack of standardization. SIT and SBT titers cannot be expected to predict anything that serum levels and/or MBCs or MICs cannot predict [20Stratton CW The role of the microbiology laboratory in the treatment of infective endocarditis.J Antimicrob Chemother. 1987; 20 (suppl A): 41-49Crossref PubMed Scopus (13) Google Scholar]. The definitive proof of infective endocarditis is demonstration of bacteria in the endocardial vegetations. It is therefore mandatory that valves and vegetations removed at cardiac surgery are transported immediately to the clinical microbiological laboratory for culturing and for direct microscopy. Histologic investigation is also recommended. Other relevant materials (Table 7) should be processed in the same way.Table 7Diagnosis on vegetations, valves, emboli, foreign material, etc.Specimens should be submitted for Gram stain, culture and histopathologic investigationsThe presence of Gram-positive cocci on smear only does not indicate failure of therapyMolecular biological techniques are recommended when unusual organisms are suspected (see Table 8)For foreign material, including valves, electrodes, patches, sutures, teflon plates, etc., culture should be performed in addition to Gram stain Open table in a new tab During antibiotic treatment, the culture positivity of vegetations will fall to 64% during the first week of 'standard antimicrobial therapy' and to 21% after two weeks' of therapy. No positive culture was found after 3 weeks' of therapy [24Morris A, Strickett A, Macculloch D. Gram stain, culture and histology results of heart valves removed during active bacterial endocarditis [abstract 1174]. In: Thirty-first ICCAC meeting, Chicago, USA, 1991.Google Scholar]. The Gram stain of vegetations will show bacteria in 64–75% of all cases during 4–6 weeks of antimicrobial treatment and in 27% of cases 1–6 months after the end of standard-duration antimicrobial therapy [24Morris A, Strickett A, Macculloch D. Gram stain, culture and histology results of heart valves removed during active bacterial endocarditis [abstract 1174]. In: Thirty-first ICCAC meeting, Chicago, USA, 1991.Google Scholar]. Thus, a positive Gram stain or acridine orange stain do not necessarily indicate antibiotic failure. Unusual causes of infective endocarditis may present difficulties in diagnosis; thus the group responsible for 'culture-negative' endocarditis probably consists of many of these organisms (Table 8). Most of these will grow poorly in culture, or culture may take weeks to show evidence of growth [25Siller KA Johnson WA Unusual bacterial causes of endocarditis.in: Kaye D Infective endocarditis. 2nd edn. Raven Press Ltd, New York1992: 265-297Google Scholar]. An example is endocarditis caused by Bartonella quintana. The organism was cultured 28 and 42 days after incubation in BACTEC 26 bottles [26Spach DH Callis KP Paauw DS et al.Endocarditis caused by Rochalimaea quintana in a patient infected with human immunodeficiency virus.J Clin Microbiol. 1993; 31: 691-692Google Scholar]. Methods for the microbiological diagnosis have been recently discussed [27Raoult D Fournia PE Drancourt M et al.Diagnosis of 22 new cases of Bartonella endocarditis.Ann Intern Med. 1996; 125: 646-652Crossref PubMed Scopus (318) Google Scholar].Table 8Microbiological diagnosis: serology, biotechnologyIn cases of initial negative blood cultures, and according to the clinical setting, specific laboratory investigations are recommended for the following microorganisms:Legionella spp. (in special media, and appropriate serologic tests)Coxiella burnetii (in special cell cultures, use of PCR and serologic tests)Chlamydia spp. (in McCoy cell or other appropriate cell cultures, use of PCR and serologic tests)Mycoplasma spp. (in special media, and serologic tests)Mycobacterium spp. (in special media, or PCR)Bartonella spp. and Afipia spp. (in AIDS patient) (PCR and immunofluorescence). Be aware of cross-reactions with Chlamydia pneumoniaeFungi (routine serology is only available in some reference laboratories) (diagnosis is usually made on the basis of clinical findings, continuous positive blood cultures, history of emboli and characteristic echocardiographic patterns)No routine serology exists for streptococci and staphylococci Open table in a new tab Diagnostic procedures for infective endocarditis due to these unusual microorganisms include specific substrates for culture, cell lines for incubation and culture, different serologic methods, immunofluorescent techniques and polymerase chain reaction (PCR) (Table 8). Continuous bacteremia is a characteristic feature of endocarditis bacteremia. The antibody response of the offending bacteria is pronounced early in the course of endocarditis and could be used for diagnostic purposes. However, a wide range of antibody specificities in normal sera have forced most investigators to use an increase in antibody titer to discriminate between different groups of patients with and without endocarditis. Direct antigen detection, as in the case of Staphylococcus aureus endocarditis, was unsuccessful due to low sensitivity, probably because of the binding of the antigen in immune complexes [28Espersen F. Serology in diagnosis of staphylococcal endocarditis [abstract s:39]. In: Abstracts of the 2nd International Symposium on Modern Concepts in Endocarditis, Elsinore, Denmark: Kandrup, 1993.Google Scholar]. Crossed Immunoelectrophoresis has been successful in finding cases of bacterial endocarditis caused by viridans streptococci [29Kjærulf A Tvede M Hoiby N Crossed Immunoelectrophoresis used for bacteriological diagnosis in patients with endocarditis.APMIS. 1993; 101: 746-752Crossref PubMed Scopus (15) Google Scholar] but these techniques are still not reliable enough for routine use. Today, infective endocarditis is a curable disease, but appropriate treatment requires the cooperation of a team of specialists. The role of the clinical microbiological laboratory is to demonstrate and characterize the infective organism, and to recommend antibiotic treatment based on laboratory susceptibility testing. The microbiologist should be medically qualified, because relevant guidance and cooperation require attention to the clinical situation. This is especially important when the performance and the interpretation of the test results are controversial. Ideally, the use of all microbiological methods should be tempered by clinical judgment in each individual case.