Infective Endocarditis

Meet the patient

A 58-year-old man with a known bicuspid aortic valve is admitted with three weeks of low-grade fever, night sweats, and a "flu" that would not go. He has lost six kilograms. On his second morning on the ward he develops sudden weakness of his right arm that resolves over an hour, and the nurse notices tiny linear splinters under two of his fingernails.[11]

His blood pressure is normal, he is in sinus rhythm at 96, and a new soft diastolic murmur has appeared at the left sternal edge that was not documented at his last clinic visit. Three sets of blood cultures are drawn before any antibiotic is given.[1]

Two exam questions are now live and the same answer closes both: is this endocarditis, and is the vegetation going to kill him before the antibiotics work? Everything below exists to apply the Duke criteria without flinching, to draw three sets of cultures before antibiotics every time, and to know the three reasons to call the surgeon before the microbiology phone call.[3][4]

What endocarditis is — and why four weeks, IV, bactericidal

Infective endocarditis (IE) is a microbial infection of the endocardial surface of the heart — most often a heart-valve leaflet, but also the chordae tendineae, the mural endocardium, or the lining of a prosthetic valve or intracardiac device. The pathological hallmark is the vegetation, a mass of platelets and fibrin colonised by microorganisms, usually bacteria and rarely fungi. Once uniformly fatal, IE now kills roughly one in six patients during the index admission despite antibiotics and surgery, so early recognition and disciplined use of the Duke criteria matter.[3]

The central clinical skill in IE is integration, not a single test: the diagnosis combines clinical features, blood cultures and echocardiography through the structured Duke criteria, first proposed in 1994, modified in 2000 and updated again in 2023.[4][5][6] The therapeutic pivot is prolonged, high-dose, bactericidal, intravenous antibiotics — because the vegetation shelters organisms from phagocytes and bacteriostatic drugs — plus, in roughly half of patients, early surgery to deal with the mechanical and uncontrolled-infective complications that antibiotics alone cannot reverse.[1]

Classification — tempo, valve substrate, and portal

IE is classified along three axes that together predict the likely organism and the empiric regimen.[1][3]

A few non-infectious mimics must be separated at the outset because they are culture-negative and demand entirely different management. Non-bacterial thrombotic endocarditis (NBTE, marantic) forms sterile vegetations in malignancy or a hypercoagulable state (Trousseau), and Libman-Sacks endocarditis forms sterile verrucous vegetations in systemic lupus erythematosus. Both produce systemic emboli and murmurs but negative blood cultures and no systemic sepsis — they are mimics of IE, not IE itself.[3]

Epidemiology — a disease that changed its face

The incidence of IE is 3 to 10 cases per 100,000 person-years in industrialised populations, with a clear upward drift over the last two decades driven by ageing, prosthetic material and healthcare exposure. In the United States the hospitalisation rate for IE roughly doubled between 1998 and 2009 in a nationwide analysis, with Staphylococcus aureus the dominant organism.[14] The incidence peaks in two groups: older adults with degenerative or prosthetic valves, and younger people who inject drugs.[1]

The modern face of the disease was drawn by the International Collaboration on Endocarditis–Prospective Cohort Study (ICE-PCS), which prospectively enrolled 2,781 adults with definite IE across 58 hospitals in 25 countries. It documented a median age of about 58 years, three-quarters native-valve IE, and S. aureus as the single commonest organism, with an in-hospital mortality near 18 percent and about half of patients needing surgery.[11]

The rise of prosthetic material, indwelling lines and S. aureus explains why the disease has shifted from young rheumatic-heart patients (where viridans streptococci dominated) to older, healthcare-exposed and prosthetic-valve patients, where S. aureus and coagulase-negative staphylococci now dominate.[1][11]

Pathophysiology — the six steps from sterile nidus to infected vegetation

Endocarditis begins where turbulent or high-velocity blood flow injures the endocardium. A jet from a stenotic, regurgitant or shunted lesion strikes the downstream endothelium and denudes it; platelets and fibrin deposit on the raw surface to form a sterile vegetation (non-bacterial thrombotic endocarditis, NBTE).[3] During a transient bacteraemia — from tooth-brushing, chewing, a healthcare procedure, an intravenous line or injection drug use — circulating organisms seed this nidus, adhere through surface adhesins (bacterial MSCRAMMs in staphylococci), multiply within the fibrin mesh sheltered from phagocytes, and an infected vegetation is established. Two cardinal rules follow:

1. Vegetations form on the low-pressure, downstream surface where the jet lands. In aortic regurgitation they sit on the atrial face of the mitral valve (the AR jet strikes the anterior mitral leaflet — the "jet lesion"); in mitral regurgitation on the atrial surface of the mitral leaflets; in a VSD on the right (low-pressure) side of the defect and on the tricuspid leaflets opposite the jet; in aortic stenosis on the ventricular side of the aortic valve. Knowing where the jet lands predicts where to look on echocardiography.[3]
2. The right heart is seeded first by bolus IV injection. A bolus of S. aureus injected into a vein passes through the right heart before the lungs filter it, so the tricuspid valve is the IV-drug-user's valve, and right-sided lesions shed septic pulmonary emboli to the lungs rather than systemic emboli to the brain and viscera.[3]

As the vegetation grows it may destroy valve tissue (leaflet perforation, chordal rupture, an abscess extending into the annulus or septum producing fistulae or atrioventricular conduction block) or fragment into the circulation, causing systemic embolisation. Left-sided lesions embolise to the brain (stroke), spleen, kidneys, mesentery and limbs; right-sided lesions seed the lungs as septic pulmonary emboli, infarcts and empyema.[3]

A parallel immune-complex response drives the immunologic phenomena — glomerulonephritis, Osler nodes, Roth spots and a positive rheumatoid factor — and a sustained inflammatory response produces the fever, anaemia and raised inflammatory markers that are the laboratory signature of the disease.[5]

This six-step cascade explains every preventive and therapeutic principle in IE: antibiotic prophylaxis targets the transient bacteraemia of step 3; bactericidal, prolonged, IV therapy overcomes the protected sanctuary of step 5; early surgery reverses the mechanical destruction and uncontrolled infection of step 6. It also explains the organisms' tropism: species with strong dextran-mediated adhesion (Streptococcus mutans, sanguinis) and fibronectin-binding staphylococci are over-represented because they adhere far more efficiently than transient contaminants such as E. coli or Pseudomonas, which rarely cause native-valve IE despite frequent bacteraemia.[3]

Causative organisms — by substrate and context

Knowing the likely organism from the clinical context drives empiric therapy. The table below is the working microbiology of IE every candidate must carry.[11][3]

The single most useful rule: the valve substrate and the clinical portal predict the organism, and the organism predicts the empiric regimen. A young dental-source native-valve case points to viridans streptococci and penicillin; a febrile prosthetic-valve patient three weeks after surgery points to coagulase-negative staph and vancomycin plus rifampicin; a febrile intravenous-drug user with pulmonary infiltrates points to S. aureus and the tricuspid valve.[11]

Clinical presentation — fever plus a murmur, then everything else

Presentation spans a fulminant septic illness to an indolent wasting disease, but the central thread is always fever plus a murmur plus evidence of embolic, immunologic or septic phenomena.[11]

History

The history is built from systemic, septic and embolic threads. Fever (often low-grade with chills, occasionally hectic) is present in about 90 percent of patients; an afebrile patient already on antibiotics does not exclude IE. Constitutional symptoms — malaise, anorexia, weight loss, night sweats, fatigue, myalgia and arthralgia — dominate the subacute presentation and may precede the murmur by weeks. Ask directly about new neurological symptoms (weakness, visual disturbance, headache — embolic stroke or mycotic aneurysm), chest or pleuritic pain and haemoptysis (septic pulmonary emboli in right-sided IE), abdominal or flank pain (splenic or renal infarct), and cold, painful extremities (limb embolus). A focused risk-factor history is essential: dental hygiene and recent procedures, intravenous drug use, indwelling lines, haemodialysis, prosthetic material, recent healthcare exposure, and known predisposing cardiac lesions.[11]

Examination — the peripheral stigmata

Examination looks for the new or changing murmur (the single most important bedside sign; a new regurgitant murmur suggests acute valve destruction) and the peripheral stigmata of embolisation and immune-complex deposition. Fever plus a murmur is the minimum diagnostic cluster, but a constellation of characteristic peripheral signs rewards the thorough examiner:[3]

• Petechiae (the commonest sign) — conjunctival, oral mucosa, palate, extremities.
• Splinter haemorrhages — linear subungual haemorrhages (non-specific; also trauma).
• Osler nodes — tender pea-sized nodules on finger and toe pulps.
• Janeway lesions — painless macules on palms and soles.
• Roth spots — boat-shaped haemorrhages with pale centres on fundoscopy.
• Clubbing — late, in subacute disease.
• Splenomegaly, arthralgia, heart failure, embolic stroke, acute limb ischaemia, renal infarct, mesenteric ischaemia.[5]

The two examinable "lesions on the extremities" are contrasted below — examiners love this distinction.[5]

Atypical presentation is the rule, not the exception, in vulnerable groups. The elderly may be afebrile or present with confusion, heart failure or stroke; the immunocompromised have blunted inflammatory signs; prosthetic-valve IE may present only as low-grade fever with a new regurgitant murmur or valve dehiscence; and right-sided (IVDU) IE may present with septic pulmonary emboli mimicking pneumonia rather than a classic systemic embolic illness.[1]

Tempo and clinical course

The tempo of IE splits the disease into two clinical archetypes, though modern practice blurs them. Recognising the tempo predicts the organism, the urgency and the prognosis.[11]

The differential — IE masquerading, and mimics of IE

The diagnostic challenge runs in two directions: IE masquerading as other febrile or embolic illnesses, and non-infectious vegetations mimicking IE. A structured differential with a discriminator for each is below.[1][3]

Persistently negative cultures with a convincing syndrome raise culture-negative IE (prior antibiotics — the commonest cause — or fastidious or intracellular organisms: Coxiella burnetii, Bartonella spp., Tropheryma whipplei, Brucella) rather than excluding the diagnosis. These need serology and PCR rather than more culture bottles.[1][6]

Bedside assessment — three sets before antibiotics, every time

The single most important investigation is three sets of blood cultures drawn from separate venepuncture sites, spaced in time, before antibiotics wherever possible — because the aetiologic agent and its susceptibility, which dictate weeks of therapy, come from these cultures. In a patient with suspected subacute IE who is haemodynamically stable, it is worth waiting the few hours for three cultures drawn over a day before starting antibiotics. In sepsis or septic shock, however, do not delay empiric antibiotics to wait for cultures: draw them and treat immediately.[3]

Examine for the new or changing murmur (a new regurgitant murmur suggests acute valve destruction and impending heart failure) and the peripheral stigmata above. Examine the fundus for Roth spots, the abdomen for splenomegaly and tenderness (splenic infarct), the urine for microscopic haematuria, and perform a focused neurological examination for an embolic stroke.[1]

A new atrioventricular block — PR prolongation or higher — is a near-pathognomonic bedside clue to a perivalvular aortic-root abscess and mandates transoesophageal echocardiography and urgent surgical referral. A serial ECG showing progressive PR prolongation in a febrile patient with IE is, for practical purposes, an abscess until proven otherwise.[1]

Echocardiography — TTE first, TOE whenever suspicion persists

Echocardiography is the imaging cornerstone. A transthoracic echo (TTE) is first-line to detect vegetations and grade regurgitation and ventricular function, but its sensitivity for vegetations is only about 60 percent, and it is especially limited for prosthetic valves, small vegetations and posterior structures. Transoesophageal echo (TOE) — sensitivity above 90 percent for vegetations and abscess — is mandatory whenever TTE is negative but suspicion persists, for all prosthetic-valve IE, to detect perivalvular abscess or fistula, and to characterise vegetations before surgery.[3][1]

The choice between TTE and TOE follows a simple rule: start with TTE in every suspected case, but if the clinical probability remains moderate-to-high (positive blood cultures, prosthetic valve, new murmur, staphylococcal bacteraemia, persistent fever) proceed to TOE regardless of the TTE result — a negative TTE never excludes IE, and a positive TTE may reveal an abscess or fistula that changes management from medical to surgical.[3]

Blood-culture technique deserves the same rigour as the imaging. Draw three sets (aerobic and anaerobic bottles per set) from three separate venepuncture sites, spaced at least 30 to 60 minutes apart (or, in urgent sepsis, drawn sequentially without undue delay), from different sites — never from existing lines unless a line-draw is specifically being compared. Use rigorous skin antisepsis (chlorhexidine-alcohol, allow to dry) to minimise contamination with coagulase-negative staphylococci, whose single positive bottle is a contaminant, not a diagnosis. The yield falls sharply after antibiotics begin — by 40 to 50 percent within hours — which is why cultures before antibiotics are paramount; if antibiotics have already been given, delay repeat cultures to allow suppressed organisms to recover.[3]

Investigations — the table that frames the work-up

Bloods usually show a normocytic, normochromic anaemia of chronic disease, a leucocytosis (marked in S. aureus, modest or absent in subacute disease), raised CRP and ESR, and often deranged renal function. Urinalysis showing microscopic haematuria supports an immunologic process (glomerulonephritis). The rheumatoid factor may be positive (an immunologic minor criterion).[5]

The Duke criteria — the framework examiners quote

Because no single finding confirms IE, the diagnosis integrates blood-culture and echocardiographic findings (major criteria) with predisposition, fever, vascular phenomena, immunologic phenomena and supporting microbiology (minor criteria) into the categories of definite, possible and rejected endocarditis. The criteria were proposed by the Duke Endocarditis Service in 1994 (two major, six minor; definite or possible or rejected),[4] modified in 2000 (redefining "possible" as one major plus one minor or three minor; promoting S. aureus bacteraemia to a major criterion),[5] and updated again in 2023 by Duke-ISCVID, adding [18F]FDG PET or CT and SPECT or CT abnormal uptake, cardiac CT detecting perivalvular complications, intraoperative evidence, and 16S or 18S rRNA PCR of excised tissue as microbiologic or imaging criteria.[6]

MAJOR criteria (two):[4]

1. A typical positive blood culture for IE:
   • (i) two separate blood cultures positive for typical microorganisms — viridans streptococci, S. aureus, the HACEK group, or community-acquired enterococci — in the absence of a primary focus; OR
   • (ii) persistently positive blood cultures — at least two positive cultures drawn more than 12 hours apart, or a majority of three or more separate cultures with the first and last drawn at least one hour apart; OR
   • (iii) a single positive blood culture for Coxiella burnetii or an anti-phase-I IgG antibody titre greater than 1:800.
2. Endocardial involvement on echo:
   • an oscillating intracardiac mass (vegetation) on a valve or supporting structures, in the path of regurgitant jets, or on implanted material; OR
   • an abscess; OR
   • new partial dehiscence of a prosthetic valve; OR
   • new valvular regurgitation (a worsening or changing of a pre-existing murmur is insufficient).[4][5]

MINOR criteria (five):[5]

1. Predisposition — a predisposing heart condition or intravenous drug use.
2. Fever — temperature 38 degrees C or higher.
3. Vascular phenomena — arterial emboli, septic pulmonary infarcts, mycotic aneurysm, intracranial haemorrhage, conjunctival haemorrhages, Janeway lesions.
4. Immunologic phenomena — glomerulonephritis, Osler nodes, Roth spots, rheumatoid factor.
5. Microbiological evidence — a positive blood culture not meeting a major criterion, or serologic evidence of active infection with an organism consistent with IE.[4][5]

The number rule for definite IE is 2-1-3 / 5: two major, OR one major plus three minor, OR five minor — or pathologic evidence of IE on excised tissue or a positive culture of an embolic specimen. Possible IE meets one major plus one minor, or three minor. IE is rejected if there is a firm alternative diagnosis, resolution with less than four days of antibiotics, or no pathologic evidence at surgery or autopsy.[4][5]

Resuscitation — sepsis and heart failure in the first hours

Resuscitation addresses the sepsis and the heart failure that dominate the first hours, while securing the microbiological diagnosis.[1][3]

• Septic or septic-shock IE: give oxygen and insert two large-bore IV cannulae, draw three sets of blood cultures from separate sites before antibiotics, then start empiric IV antibiotics immediately — do not delay therapy in sepsis. Give crystalloid fluids cautiously and use vasopressors (noradrenaline first-line) for shock; correct electrolytes; monitor urine output, lactate and renal function. Follow the sepsis-six bundle and escalate to critical care early.
• Acute severe regurgitation and pulmonary oedema: this is the commonest surgical emergency in IE. Sit the patient up, give high-flow oxygen, IV furosemide, a vasodilator (glyceryl trinitrate infusion if blood pressure permits) and non-invasive ventilation for respiratory failure — but these measures only buy time. Definitive treatment is urgent valve surgery, and surgical referral must be made at the moment the diagnosis of acute severe regurgitation is made, without waiting for culture completion.[1]
• Embolic stroke: maintain perfusion and oxygenation, avoid hypotension, and involve neurology and cardiothoracic surgery; the question of early surgery after stroke is complex and time-dependent.[1]

The non-negotiable rule: three sets of blood cultures before antibiotics, every time — the whole of definitive therapy hangs on what these cultures grow.[3]

Definitive therapy — antibiotics, surgery, prevention

Definitive therapy rests on three pillars: bactericidal antibiotic therapy directed at the organism and its MIC, early surgery for mechanical or uncontrolled complications, and prevention of recurrence through prophylaxis.[3][10]

Empiric antibiotic therapy (before culture results)

Empiric therapy is dictated by the valve substrate and the clinical context. Every regimen below combines a cell-wall-active bactericidal agent (a beta-lactam or glycopeptide) with cover for the most likely organisms; doses assume normal renal function and must be adjusted to MIC and renal function once cultures return.[1][3]

Definitive (organism-directed) therapy

Once the organism and MIC are known, therapy is narrowed and prolonged. Total duration is counted from the first day of effective antibiotics (the first negative culture day for some guidelines).[3][9]

• Penicillin-susceptible viridans streptococci (MIC at or below 0.12 mg/L): benzylpenicillin (penicillin G) 1.2 g (2 MU) IV 4-hourly (total 12 to 18 MU per day) — or ceftriaxone 2 g IV once daily for 4 weeks. A 2-week course adding gentamicin 1 mg/kg IV 8-hourly for the first 2 weeks is acceptable in selected uncomplicated cases (young, native valve, MIC at or below 0.12, no emboli, prompt response).[3]
• Penicillin-relatively-resistant streptococci (MIC over 0.12 to at or below 0.5 mg/L): benzylpenicillin or ceftriaxone for 4 weeks plus gentamicin 1 mg/kg IV 8-hourly for the first 2 weeks.[3]
• MSSA native-valve IE: flucloxacillin (or nafcillin or oxacillin) 2 g IV 4- to 6-hourly for 6 weeks; gentamicin 1 mg/kg IV 8-hourly for the first 3 to 5 days is optional and no longer routine. For uncomplicated right-sided MSSA IE in IVDU, a 2-week course of flucloxacillin plus gentamicin may suffice.[3]
• MRSA IE: vancomycin 15 mg/kg IV 12-hourly (or 25 to 30 mg/kg per day continuous or intermittent, trough 15 to 20) for 6 weeks; daptomycin 6 to 10 mg/kg IV daily is an alternative. Gentamicin is not routinely added (nephrotoxicity, no benefit).[9]
• Enterococcal IE (ampicillin-sensitive): ampicillin 2 g IV 4-hourly plus gentamicin 1 mg/kg IV 8-hourly for 4 to 6 weeks; for aminoglycoside-resistant or HLAR strains use ampicillin plus ceftriaxone (a synergistic double-beta-lactam) for 6 weeks. Vancomycin replaces ampicillin for ampicillin-resistant strains.[3]
• HACEK IE: ceftriaxone 2 g IV daily for 4 weeks (ampicillin-sulbactam is an alternative).[13]
• Culture-negative IE: treat the underlying organism once identified — doxycycline plus hydroxychloroquine for Coxiella (at least 18 months), gentamicin (or ceftriaxone) plus doxycycline for Bartonella, trimethoprim-sulfamethoxazole for Tropheryma whipplei.[6]

Surgery — the three ESC indications

Early surgery (during the initial antibiotic course) is needed in roughly half of patients, driven by three ESC-defined indications.[1][2]

Surgery is also indicated for prosthetic-valve IE with dehiscence or obstruction, large vegetations from organisms difficult to eradicate (fungi), and CIED infection (requiring complete device and lead extraction). The operation is usually valve replacement with a mechanical or bioprosthetic prosthesis, but valve repair is increasingly used for isolated mitral or tricuspid lesions where feasible, avoiding long-term anticoagulation.[1]

The clearest evidence for surgery as embolic prevention is Kang and colleagues' randomised trial of early surgery (within 48 hours) versus conventional treatment in left-sided IE with large vegetations: early surgery reduced the composite of in-hospital death and embolic events (3 percent versus 23 percent, hazard ratio 0.10) — a benefit driven entirely by fewer embolic events, since all-cause mortality alone was not significantly different (3 percent versus 5 percent).[7]

The decision to operate, and how urgently, follows a structured assessment of the indication and the operative risk.[1]

Prophylaxis — the great international divergence

Special situations you will actually meet

Culture-negative IE follows prior antibiotics (the commonest reason — organisms suppressed, not killed) or fastidious or intracellular organisms (Coxiella, Bartonella, Tropheryma, Brucella, Legionella) and fungi (Candida, Aspergillus), which need serology or PCR rather than more culture bottles. The 2023 Duke-ISCVID update explicitly adds 16S or 18S rRNA PCR and metagenomic sequencing of excised tissue as microbiologic criteria, reflecting this reality.[6]

Streptococcus gallolyticus (formerly S. bovis) bacteraemia or IE mandates colonoscopy, because of its strong (about 30 percent) association with advanced colonic neoplasia — finding the organism in a blood culture is itself an indication to screen the colon.[1] Fungal IE (Candida, Aspergillus) occurs with immunosuppression, long-term indwelling lines, total parenteral nutrition and IV drug use, produces large friable vegetations with major embolic risk, and has a poor prognosis — it almost always needs surgery combined with prolonged antifungal therapy (liposomal amphotericin B or an echinocandin, step-down to fluconazole).[1]

Cardiac-implantable electronic device (CIED) infections of the pocket or lead may seed the tricuspid valve; they usually require complete percutaneous or surgical device and lead extraction in addition to targeted antibiotics, as infection of the insulated lead is essentially impossible to sterilise medically.[1]

The HACEK group are slow-growing, fastidious Gram-negative oropharyngeal commensals that need prolonged incubation (now mostly recovered in modern automated blood-culture systems within 5 days). They cause about 1 to 3 percent of IE, typically subacute on native valves, and are uniformly susceptible to ceftriaxone 2 g IV daily for 4 weeks.[13]

How patients with endocarditis come to harm (the preventable list)

• Treating S. aureus bacteraemia as a simple line infection without echocardiographic exclusion of IE — significant SAB carries a high probability of endocarditis; suboptimal evaluation (omitting echo, too-short antibiotics) is associated with worse outcome.[12]
• Delaying surgery for heart failure in the hope of medical stabilisation — heart failure is the commonest indication for surgery and the commonest cause of death; the threshold to operate is appropriately low.[1]
• Missing a perivalvular abscess — a new AV block or persistent fever on correct antibiotics is an abscess until proven otherwise; only TOE and surgery will resolve it.[1]
• Over-diagnosing IE from a single positive culture — one bottle growing a coagulase-negative staphylococcus is a skin contaminant, not PVE; the Duke criteria require persistence.[5]
• Drawing cultures after antibiotics — the yield falls by 40 to 50 percent within hours; three sets before antibiotics, every time.[3]
• Forgetting rifampicin's timing in PVE — add it only after bacteraemia clears (3 to 5 days), or you breed resistance.[1]
• Missing the S. bovis colonoscopy — finding gallolyticus in a blood culture is itself an indication to screen the colon for neoplasia.[1]

Prognosis and disposition

Despite modern antibiotics and surgery, IE remains lethal, with in-hospital mortality of about 15 to 20 percent for native-valve IE, 20 to 40 percent for prosthetic-valve IE, and 40 to 60 percent when S. aureus is the organism. One-year mortality approaches 30 percent.[11][3] Adverse prognostic features are the same factors that predict death and drive surgery: S. aureus infection, prosthetic-valve involvement, heart failure, older age, perivalvular complications (abscess, conduction block), embolic events, large vegetations, and healthcare-associated acquisition.[11]

Conversely, viridans-streptococcal infection is comparatively favourable, and indicated surgery is itself protective — a reminder that the threshold to operate in IE is appropriately low, and that delaying surgery for heart failure in the hope of medical stabilisation is a recurring and lethal error.[11]

Disposition requires 4 to 6 weeks of intravenous therapy (often via a peripherally inserted central catheter, with outpatient parenteral antibiotic therapy once the patient is stable, afebrile, culture-negative and surgically cleared), serial inflammatory markers and repeat echocardiography to document shrinking vegetations and stable valve function, dental clearance (eliminate the oral source of bacteraemia), infectious-diseases, cardiology and (where relevant) cardiothoracic-surgery follow-up, and patient education on the recurring risk and the need for prophylaxis. In S. aureus bacteraemia, adherence to infectious-diseases specialist recommendations (including echocardiographic evaluation) measurably improves outcome.[12]

Pregnancy and special populations

Evidence, guidelines and regional differences

The evidence base for IE rests on a small number of landmark studies and a set of converging guidelines that now agree on most points but diverge sharply on prophylaxis.[1][3]

• ESC 2023 (Delgado) versus AHA or IDSA 2015 (Baddour): both endorse the Duke framework and the three surgical indications (heart failure, uncontrolled infection, prevention of embolism). The ESC 2023 guideline is the most current and adds [18F]FDG PET or CT and cardiac CT as major diagnostic criteria, mirroring the 2023 Duke-ISCVID update; the AHA statement is older but its antibiotic recommendations remain standard.[1][3][6]
• Kang 2012 RCT is routinely (and wrongly) summarised as "early surgery saves lives" — it did not show a significant mortality reduction. Its benefit was a drop in embolic events within a composite endpoint (death plus emboli), which is exactly how it should be cited. Subsequent observational data and the 2023 ESC guidance accept early surgery for large (over 10 mm) mobile vegetations with high embolic risk, especially after a first embolic event.[7]
• Prophylaxis — the great divergence: the AHA (US) restricts prophylaxis to high-risk cardiac lesions for dental procedures only; NICE (UK) went further and recommends no routine antibiotic prophylaxis at all, emphasising oral hygiene instead; ESC (Europe) takes a middle position, recommending prophylaxis for the highest-risk group for dental procedures. India (ICMR or NCDC) does not mandate routine dental prophylaxis but recognises the high-risk groups; empiric antibiotic choices in India must account for local resistance (high community MRSA, ESBL Gram-negatives).[8]
• IDSA MRSA guidance (Liu 2011) remains the reference for vancomycin dosing (trough 15 to 20 mg/L), duration (6 weeks for native, at least 6 weeks for prosthetic), and the option of daptomycin for vancomycin failures or intolerance.[9]
• Controversy: routine surgery for large vegetations over 10 mm without embolism remains debated; observational data suggest benefit is concentrated in the first week after presentation, before the embolic risk falls.[7]

The mantra, and the viva honesty line

The mantra: three sets before antibiotics, Duke to diagnose, four weeks of a bactericidal IV drug — and call the surgeon for failure, fever, or a flying vegetation.[1][3]

Ward-round test — three stems, thirty seconds each