Endocarditis for Residents: Cultures, Criteria and Echo

Cast your mind back to medical school. You were taught to examine a patient starting at the hands, looking for splinter haemorrhages, Osler's nodes and Janeway lesions. When did you last actually see any of them? Not because we've stopped looking — because the disease has changed underneath us. In this episode Mike and Calum work through infective endocarditis: why it presents so differently now, how it gets diagnosed, and the traps that make it harder than it needs to be. This is the diagnosis and intensive care half; the theatre half is the companion episode. Please note: antibiotic choices vary enormously between centres and even between regions. Everything here is illustrative — follow your own microbiology advice. Endocarditis used to be a community disease, brewing over weeks, with time to develop immune-complex phenomena in the fingernails. It's now much more of an acute, hospital-associated disorder — roughly a third of cases are healthcare-associated, and those carry a higher mortality. And we generate the risk factors ourselves, with valves, pacemakers, indwelling catheters and dialysis lines. We cover the organisms and how they differ by continent, HACEK (and why the lab needs warning about slow growers), and fungal endocarditis — rare, bulky and grim. Then the pathophysiology, which is more elegant than it first appears: a healthy valve is remarkably resistant to infection, so something has to damage it first. Turbulence injures the endothelium, a sterile vegetation forms from platelets and the clotting cascade, and that bland thrombus becomes the landing pad a later bacteraemia colonises. That two-hit sequence explains where vegetations sit and why any valve, device or structural disease makes a patient high risk. Then the two pillars of the Duke criteria, both of which get routinely mishandled. Blood cultures are frequently done badly — one set, then Tazocin, and by the time the microbiologist rings the antibiotics are already in and everything comes back negative. You've manufactured your own culture-negative endocarditis. Echo access is the second problem, though a new generation of intensivists and anaesthetists confident with bedside scanning is changing the timeline. We give clean rules: transthoracic first, isolated right heart disease can stop there, but a prosthetic valve or any implantable device always needs a TOE. The criteria themselves have grown up — surgical inspection of the valve is now a major criterion, and gated CT and nuclear imaging rescue the uncertain case. We're honest about the antibiotic evidence, which amounts to no high-quality randomised trials at all, and work through exactly why that trial is so hard to design. Then the specifics worth knowing: aminoglycosides are out for staphylococcal native valve disease, daptomycin and fosfomycin are in for MRSA, and rifampicin waits until the bacteraemia has cleared. We finish with what gets mistaken for endocarditis on echo, built around a real diagnostic argument — a lesion that turned out to be a fibroelastoma, because it was on the wrong side of the valve, too smooth and too round, and the patient was far too well. Chapters (00:00) Cold open — what you were taught to look for at the hands (01:10) Why the classical signs have vanished (02:20) The organisms, and why they've changed (03:40) HACEK, and warning the lab (04:50) Fungal endocarditis (05:50) The two-hit pathophysiology (07:10) Risk factors: any valve, any device (08:00) Blood cultures done badly (09:20) Transthoracic, transoesophageal, and who must have one (10:40) The criteria grow up: surgery, CT and nuclear imaging (12:00) The antibiotic evidence — and how you'd design the trial (13:30) Aminoglycosides, daptomycin and the rifampicin timing rule (14:40) What gets mistaken for endocarditis on echo (16:00) Wrap-up Key takeaways The disease has moved from the community into hospital — stop waiting for splinter haemorrhages About a third of cases are healthcare-associated, and we create the risk factors with valves, devices and lines HACEK organisms are slow-growing, so the lab needs warning or you'll get a falsely negative culture Endocarditis rarely affects a normal valve: turbulence damages endothelium, a sterile vegetation forms, and bacteraemia colonises it Vegetations sit on the upstream side of the valve, where the shear stress is Take adequate blood for culture before the antibiotics, or you invent your own culture-negative case Transthoracic first; isolated right heart disease can stop there; a prosthetic valve or device always needs a TOE Surgical inspection of the valve is now a major criterion, and gated CT and nuclear imaging rescue uncertain cases There is no high-quality randomised evidence behind the antibiotic regimens — and there are good reasons why Aminoglycosides are out for staphylococcal native valve disease; daptomycin and fosfomycin are in for MRSA; rifampicin waits until the bacteraemia clears Not everything shaggy is endocarditis — check the side, the shape, and the patient References / further reading Charlesworth M, Williams BG, Ray S. Infective endocarditis. BJA Education 2023 Delgado V et al. 2023 ESC Guidelines for the management of endocarditis. Eur Heart J 2023 Fowler VG et al. The 2023 Duke–ISCVID criteria for infective endocarditis. Clin Infect Dis 2023 Habib G et al. Recommendations for the practice of echocardiography in infective endocarditis. Eur J Echocardiogr Cahill TJ, Prendergast BD. Infective endocarditis. Lancet 2016 National Institute for Health and Care Excellence. Prophylaxis against infective endocarditis (CG64) Follow the podcast Bluesky: @cardiacoutput.bsky.social X: @CardiacOutputMC If you've found this useful, a follow or a share genuinely helps other trainees find it — and do get in touch if there's a topic you'd like covered. This podcast is for medical education for healthcare professionals. It is not clinical advice. Antibiotic regimens and practice described are illustrative and vary by centre — always follow your own microbiology advice, local guidelines and current policy.