Infective Endocarditis in Children

Does this child need surgery, and how soon?

Indication present: surgery need not wait for antibiotics to finish. Delay can be dire; in general illness is not a limit (AHA 2015 Class I, LOE B). No hour or day is established in children.

Stroke or intracranial hemorrhage can change the timing: Stroke timing (Adult; no pediatric data).

A negative echo does not exclude endocarditis (JCS 2026; ESC 2023, CHD). Valve and conduit rows.

What to start · Vegetation size · Repair or replacement

How soon to operate, from JCS 2026 Table 34, Adult: emergency, within 24 h of antibiotics; urgent, within several days; elective, 1–2 weeks after antibiotics. Tap every row that fits. AHA 2015 says its surgical advice mostly extends adult opinion; "listed" is its Table 11, a 1998 adult reprint.
How soon to operateIndicationClass by source
Heart failure
Emergency Adultwithin 24 hCardiogenic shock from acute severe valve dysfunction or shunt, severe pulmonary edema, or new severe heart failure symptomsPediatric AHA 2015 listed, no class · PAC3 2025 Moderate / StrongAdult JCS 2026 · ESC 2023 I B · ACC/AHA 2020 I B-NR
Urgent Adultwithin several daysHeart failure symptoms (PAC3); acute aortic or mitral regurgitation with ventricular failure, or unresponsive heart failure (AHA 2015 Table 11). Adult JCS also: severe regurgitation, worsening hemodynamics on echo, at least moderate pulmonary hypertensionPediatric AHA 2015 listed, no class · PAC3 2025 Moderate / StrongAdult JCS 2026 · ESC 2023 I B · ACC/AHA 2020 I B-NR
Uncontrolled infection
Emergency or urgent Adultwithin 24 h to several daysAbscess, pseudoaneurysm, intracardiac shunt, enlarging vegetation, or new AV block or LBBB Pediatric AHA 2015 named, no class · PAC3 2025 Moderate / StrongAdult JCS 2026 · ESC 2023 I B · ACC/AHA 2020 I B-NR
Emergency or urgent Adultwithin 24 h to several daysSepsis affecting hemodynamics despite appropriate therapy Pediatric no class givenAdult JCS 2026 I B · ESC 2023 IIa B
Urgent or elective Adultwithin several days, or 1–2 wk after antibioticsFungi, S. aureus or other resistant organisms such as MRSA or VRE, or a needed change of antibiotic Pediatric AHA 2015 I C, fungal only · PAC3 2025 Moderate / StrongAdult JCS 2026 · ESC 2023 I C · ACC/AHA 2020 I B-NR
Urgent Adultwithin several daysPersisting infection on appropriate therapy, no other focus. Adult JCS: typically positive cultures or clinical findings at 3–5 days AHA 2015: S. aureus bacteremia may last 3–5 days on a β-lactam, 5–10 on vancomycin; may not clear until a central line is out.Pediatric AHA 2015 named, no classAdult JCS 2026 · ESC 2023 IIa B · ACC/AHA 2020 I B-NR
Urgent or elective Adultwithin several days, or 1–2 wk after antibioticsProsthetic valve: relapse or recurrence Pediatric AHA 2015 IIb BAdult JCS 2026 IIa C · ACC/AHA 2020 I C-LD
Urgent or elective Adultwithin several days, or 1–2 wk after antibioticsProsthetic valve: S. aureus, resistant staphylococci or non-HACEK Gram-negative organisms AHA 2015, S. aureus: valve replacement may be considered, preferable in most if not all.Pediatric AHA 2015 IIb C, S. aureusAdult JCS 2026 · ESC 2023 IIa C
Not timed by name; a shunt or heart failure takes those rowsUnstable prosthesis, ruptured sinus of Valsalva, ventricular septal or mycotic aneurysm Pediatric AHA 2015 named, no class; mycotic aneurysm IIb B · JCS 2026 named, no class
Embolism
Urgent Adultwithin several daysEmbolic event with a vegetation persisting or enlarging on therapy. Adult JCS: >10 mm; ESC: ≥10 mm. Pediatric AHA 2015 listed, no class · PAC3 2025 Low / WeakAdult JCS 2026 · ESC 2023 I B · ACC/AHA 2020 IIa B-NR
Urgent Adultwithin several daysProgressive valve dysfunction. Adult JCS: with mobile >10 mm; ESC: ≥10 mm with another indication.Pediatric JCS 2026 named, no classAdult JCS 2026 I B · ESC 2023 I C
Urgent, Adult only Adultwithin several daysVegetation alone, to prevent a first embolus: mobile >10 mm, no other indicationPediatric: not recommended: no proven benefit, long-term risks of valve replacement (AHA 2015 Class III, LOE C). Same statement, Table 11, a 1998 adult reprint: anterior mitral leaflet, particularly >10 mm, may need surgery. Pediatric AHA 2015 III CAdult JCS 2026 IIb C · ESC 2023 IIb B · ACC/AHA 2020 IIb B-NR

Vegetation alone: the classes differ. Pediatric AHA 2015: surgery to prevent a first embolus is Class III, LOE C. Adult mobile >10 mm: JCS 2026 IIb, C; ACC/AHA 2020 IIb, B-NR; ≥10 mm: ESC 2023 IIb, B. Size alone is hard to interpret in a small child: even 5 mm may matter on a newborn's mitral valve (Backer 5e ch 45 pp 962–963 textbook, opinion). JCS 2026: no clear pediatric data on surgical indications to prevent embolism.

Any row: cardiology, surgery and infectious diseases review every admitted patient, ongoing, to choose medical or surgical care and its timing, including shortening preoperative antibiotics when appropriate (PAC3 High / Strong). Consider early surgery, especially if repair is an option (PAC3 Moderate / Strong). S. aureus: care where there is cardiac surgery and infectious diseases (AHA 2015 Class I, LOE C). Adult AHA 2026: within 48–72 h once indicated, unless major stroke or cerebral hemorrhage.

Transcatheter pulmonary valve or conduit

Tap each row that fits. No source gives a time for these. After a transcatheter pulmonary valve, vegetation seen on echo in 34% of 69 Melody cases abstract.
FindingWhat the source saysClass by source
TPVR, RV-PA conduit or pulmonary bioprosthesis
Mild and stableDefinite or possible right-sided endocarditis after TPVR, mild symptoms, mild prosthetic dysfunction, no hemodynamic compromise: favor medical therapy Pediatric PAC3 2025 Moderate / Strong
S. aureusAfter TPVR: consider early surgical explantAll organisms 27–64% by center; streptococcus 32%, MSSA 53%, MRSA all (Davtyan 2022 abstract). No comparison of explant against medical treatment.Pediatric PAC3 2025 Moderate / Strong
RVOT gradientNew or increased, no structural cause: count as an extra minor Duke criterion Pediatric PAC3 2025 High / Strong
Modified Duke not metRV-PA conduit or pulmonary bioprosthesis: do not exclude endocarditis for that alone. At least possible endocarditis with nothing on TTE or TEE: pursue ICE, CT or PET/CTPediatric PAC3 2025 Very Low / Strong
Bovine jugular valveConsider endocarditis even with negative culturesPediatric PAC3 2025 Moderate / Strong
Prosthetic materialShunt or conduit: bacteriologic cure often impossible without surgery; vegetations may cause dehiscence, obstruction or distal emboliPediatric AHA 2015 no class
Stroke timing

No pediatric data on how long to wait. Every time here is Adult. PAC3 2025: in children with neurologic complications, evidence supports preoperative neuroimaging and pediatric neurology evaluation, left out of its recommendations as not specific to CHD. High index of suspicion for stroke with left-sided lesions (PAC3 Low / Strong).

Adult Side by side, not reconciled. AHA 2026 prints no classes and says it has no prospective data here.
SituationAHA 2026JCS 2026 Table 35ACC/AHA 2020ESC 2023 Table 17AATS 2016 Table 3
No hemorrhage, no severe deficitSmall strokes on imaging, no deficit: no delayEarly surgery indicated, no coma or herniation: no delay I BNo hemorrhage or extensive damage: no delay may be considered IIb B-NRTIA: no delay I B. Stroke with heart failure, uncontrolled infection, abscess or high embolic risk: no delay, if no coma and CT or MRI excludes hemorrhage I BNon-hemorrhagic with a strong cardiac indication: earlier surgery reasonable IIa B
Embolic, no bleeding, unstable or high re-embolism riskUrgent (<72 h) or emergent surgery should be considered. High risk: large or mobile >10 mm, mitral (especially anterior leaflet), staphylococcal——The no-delay row above—
Extensive infarctTake-home point: major ischemic stroke usually at least 2, preferably 4 weeks; its body text ties that delay to hemorrhagePrognosis not hopeless: after neurologic stabilization and antibiotics IIa CIf hemodynamically stable: delay at least 4 weeks may be considered IIb B-NR—Large or multiple strokes, severe symptoms: neurologist first I B
Intracranial hemorrhageUsually at least 2, preferably 4 weeks, by stability and neurologic evaluationWithin 2 weeks once neurologically stable, no hematoma expansion IIa C. With expansion: after at least 4 weeks of antibiotics IIa BIf hemodynamically stable: delay at least 4 weeks may be considered IIb B-NRDelay beyond 1 month if possible, reassessing often IIa C. Unstable: urgent or emergency should be considered, weighing neurologic outcome IIa CRecent: delay 3 or more weeks reasonable IIa B
Embolic stroke with hemorrhageEarly surgery may be considered in select cases: high re-embolism risk, hemodynamic compromise, sepsis or abscess. Otherwise 3–4 weeks————
  • Adult Before surgery, exclude mycotic aneurysm by CT/CTA or MRI/MRA; surgery is not appropriate in a comatose patient (AHA 2026).
  • Adult Antithrombotics, JCS 2026 Table 17: stop antiplatelet and anticoagulant with intracranial hemorrhage except microbleeds I B. ACC/AHA 2020: with embolism or stroke, reasonable to stop anticoagulation for a time IIa B-NR. No pediatric recommendation, and no source on a mechanical valve in a child.
  • Backer 5e ch 45 p 963 textbook: infarction, wait 2–3 weeks where possible; ruptured mycotic aneurysm with subarachnoid hemorrhage is a prohibitive risk for bypass, 2–3 weeks after craniotomy. Older adult citations; newer guidelines above differ.
  • Pediatric Jacobwitz 2023, 68 children: neurologic complications in 23 (34%), an undercount since 38 were imaged; hemorrhage 17, ischemic 17, both in 11. All 6 deaths were among those 23. Median 15 days from neurologic onset to surgery in 10; the authors could not tell whether neurologic concerns set the timing. They suggest neurology review and screening neuroimaging within 24–72 h: their suggestion, untested.
Operation: repair or replacement in a growing child
  • Pediatric PAC3 2025: consider repair when feasible in native valve endocarditis Moderate / Weak; consider early surgery, especially if repair is an option Moderate / Strong; Ross for an unrepairable aortic valve, at experienced centers Moderate / Strong.
  • Pediatric AHA 2015: long-term risks of valve replacement in childhood are its stated reason against prophylactic surgery. Valve-sparing is often possible, by the damage; aortic root complications can take a homograft or Ross.
  • Adult AHA 2026: choice of substitute matters less than eradicating infection and restoring function. Right-sided: repair preferred whenever feasible. Many of those patients inject drugs, so it may not read across to a conduit.
  • Adult ESC 2023: tricuspid repair over replacement when possible IIa B; repair may be needed in children, where replacement options are fewer; Ross may be considered in very selected patients, children in particular. CHD: same surgical indications as acquired disease, in a specialized CHD center.
Pediatric Single-center, surgically selected, and the Melbourne papers overlap (1987–2020). No pediatric trial of repair against replacement. Wu 2024: of the 3 mechanical valves, 2 went to children in extremis with shock, the third before that center began the Ross.
SeriesOperation, nEarly death10-y survival10-y free of reoperation
Wu 2024 aortic, 41Repair, 160/1687.5%30.8%
Ross, 132/13 (15.4%)74.1%63.0%
Homograft root, 93/9 (33.3%)66.7%26.3%
Mechanical, 31/3——
Wu 2023 mitral, 39Repair, 36 (92.3%)2/36 (5.6%)—62.9%; 80.2% free of valve replacement
Replacement, 3———

Wu 2024: survival p>0.05; reoperation, Ross vs repair p=0.15, Ross vs homograft p=0.002.

  • Paravalvular abscess, Wu 2023, 30: root replacement 17 (homograft 9, Ross 8), central fibrous body reconstruction 5. Early death 6 (20%), 5 of them on ECMO. Free of reoperation 60.6%, 50.0%, 40.0% at 5, 10, 15 y. Reoperation for recurrence 2.
  • Native valve preserved: 73% (Hickey 2009 abstract); 23/28 (Lee 2021). Carrillo 2023, median 16.7 y: replacement 37 of 47 (78.7%).
  • Growth: Russell 2013 abstract: replacement 22, repair 10, Ross 5; reoperation 28%, nearly all from somatic growth; all 5 operative deaths were infants, mean 2.5 months.
  • Mechanical or biologic in a small child, unsettled: Backer 5e ch 45 p 969 textbook prefers mechanical if Ross is not feasible, weighed against pregnancy, as opinion. Wu 2024's 3 mechanical valves (1 death, 2 of the 3 chosen in shock) say little about the valve. Khoo 2019: aortic valve replacement, HR 3.22 for death, mostly in active disease, confounded. No pediatric comparison of valve types.
  • Principles, Backer 5e ch 45 pp 963–964 textbook: debride all infected material, restore hemodynamics, autologous tissue first; no evidence that biologic tissue lowers reinfection. An infected shunt: a new shunt in a new location, old one removed; no shunt cohort found.
  • Valvectomy: Adult ESC 2023 names it as a less common tricuspid strategy; no pediatric series found.
  • Percutaneous mechanical aspiration: on search, case reports only in children. Adult ESC 2023 Class IIb, LOE C: debulking right intra-atrial septic masses in selected patients at high surgical risk.

Figures, in the book: Backer 5e Figs 45.1–45.4 (tricuspid), 45.5–45.13 (aortic), 45.14–45.21 (mitral).

Substrate and risk
Rates among people at risk. Different denominators: do not compare across rows.
PopulationRateSource
Children with CHD, Quebec, 47,518; cumulative rate from the 34,279 followed since birth6.1 per 1000 children by age 18 (5.0–7.5). Against ASD: cyanotic RR 6.44, endocardial cushion 5.47, left-sided 1.88. Surgery within 6 months RR 5.34. Age <3 y RR 3.53, against ages 6–18Rushani 2013
Children with CHD, Norway2.2 per 10,000 person-years; 8% endocarditis mortalityJortveit 2018 abstract
Tetralogy of Fallot, Denmark, from birth; mixed ages22.4 per 10,000 person-years vs 0.1 in controlsHavers-Borgersen 2021 abstract
After TPVR, 2476, international registry; mixed ages2.2 per 100 patient-years; 9.5% at 5 y, 16.9% at 8 y; median 2.7 y to onset. Younger age, prior endocarditis, higher residual gradient; valve type not a factor on multivariable analysisMcElhinney 2021
Congenital AS, Second Natural History Study27.1 per 10,000 person-years; 15.7 managed medically vs 40.9 surgically. Prior valvotomy: the authors read the excess as severity, not the operation. Before balloon valvuloplasty became the usual first stepGersony 1993 abstract
  • Pediatric PAC3 2025: any CHD, prior endocarditis, male sex High / Strong; indwelling IV catheters Moderate / Strong; first year after surgery, particularly the first 6 months Moderate / Strong; foreign material, particularly the first 30 days Low / Strong; age 0–3 y, especially infants Moderate / Weak.
  • Pediatric PAC3 2025, elevated risk: TOF, particularly after PVR, truncus, transposition, univentricular hearts, AVSD, valve repair or prosthetic valve, shunts, pulmonary-position conduits Moderate / Strong.
  • Pediatric PAC3 2025, RV-PA conduit, high suspicion with any of: at least 2 conduit replacements, 22q11 deletion, prior endocarditis, TPVR at 12 y or younger, post-TPVR peak gradient above 15 mm Hg, recent procedure, skin infection EO / Strong.
  • Pediatric Death: PAC3 2025 names age <1 y, heart failure, S. aureus and vegetation >15 mm (a mortality factor, not a surgical trigger; not fully traced) Moderate / Strong. One of its five sources, Yoshinaga 2008, 137 with CHD, 1 month to 62 y, median 12 y: vegetation ≥20 mm OR 40.6 (2.42–681), age <1 y 19.5 (1.74–219), heart failure 7.16 (1.34–38.4), S. aureus 5.68 (1.16–27.9); ≥15 mm on univariate analysis only. Mixed ages, wide intervals. AHA 2015 quotes 48% with tetralogy and pulmonary atresia: 12 of 25 in the database behind it (Day 2009 abstract).
  • No structural heart disease is not lower risk: mortality 10.6% vs 3.5%, p=0.041 (Chien 2023 abstract). 23 of 170 had no specific congenital condition (JCS 2026).
  • Neonates: no cohort found. Fungal, reported cases only: 71, 92% preterm, mortality 42.2% (Pana 2015 abstract).
  • Melody or SAPIEN, unsettled: 8.5% vs 2.1% unadjusted, in a meta-analysis that excluded pediatric studies, with shorter SAPIEN follow-up (Balakrishna 2023 abstract). PAC3: excess with bovine jugular tissue may reflect small stenotic conduits in the young Moderate / Weak. McElhinney 2021: valve type not a factor on multivariable analysis. No randomized comparison, no pediatric SAPIEN cohort found.
  • Adult Restrictive VSD, ACHD 2025: preventing endocarditis is not a reason to close one without prior endocarditis; after an episode, repair may be contemplated, on limited evidence. Recent VSD endocarditis: closure may be reasonable IIb C-LD.
Organisms

Share of cases, not risk. Mostly coded data, not microbiologic confirmation.

  • CHD overall: staphylococci and streptococci lead, either order (PAC3 2025 Moderate / Strong). S. aureus brings more heart failure, abscess, septic emboli and CNS events Low / Weak.
  • US administrative data: S. aureus 57%, viridans streptococci 20% of coded cases (Day 2009 abstract). CHD, 2108 patients: staphylococci 24.1%, streptococci 22.3%, no organism coded 41.3% (Elsisy 2026).
  • Paravalvular abscess: S. aureus 33.3%; S. pneumoniae HR 9.2 for death (1.6–51.7) (Wu 2023).
  • S. aureus bacteremia in children, definite endocarditis (risk, not share): 8.8%, 15 of 171 (Pakotiprapha 2026 abstract); 11.8% (Valente 2005 abstract).
  • Kingella: previously healthy under 36 months, boys, oral aphthae, with or without CHD (Lowenthal 2021 abstract).
  • Culture-negative: recent antibiotics are the commonest cause; PCR on valve tissue when valve surgery is done (PAC3 2025 Moderate / Strong).
Duke and imaging
  • Criteria. PAC3 2025: use modified Duke in CHD; it may underperform after TPVR Low / Strong. Sensitivity in 41 episodes: von Reyn 63%, Duke 81%, modified Duke 88% (Tissières 2003 abstract). A central line may prolong bacteremia and keep criteria unmet until removed (AHA 2015).
  • Duke-ISCVID 2023 (Fowler 2023 Tables 1 and 2). Definite: pathologic, or 2 major, or 1 major and 3 minor, or 5 minor. Possible: 1 major and 1 minor, or 3 minor. New majors: findings at direct inspection during surgery, cardiac CT, FDG PET/CT, PCR for Coxiella, Bartonella or T. whipplei. New predispositions: previous endocarditis, endovascular device. No pediatric validation found; JCS 2026 says studies are awaited.
  • Cultures. 1–3 mL in infants and young children, 5–7 mL in older children, by bottle system (AHA 2015). If not severely ill and cultures are negative or may be contaminants, withhold antibiotics at least 48 h for more cultures (Class IIa, LOE C; its diagnosis section words this more narrowly, Class IIb, LOE C). Severely ill and unstable: 3 cultures over 1–2 h, then start empiric treatment (Class I, LOE C).
  • TTE first (PAC3 2025 Low / Strong). Sensitivity 97% under 60 kg, 70% at 60 kg or more (Penk 2011). For the pulmonary valve TTE may be better than TEE (Low / Strong).
  • TEE: poor TTE windows, or negative TTE with high suspicion (PAC3 Low / Strong). AHA 2015: chest wall disruption from surgery or trauma, or thoracic cage anomalies, Class I, LOE B; high risk of aortic root abscess, Class IIa, LOE C. Even TEE can miss vegetations, as in a Blalock-Taussig shunt (Wong 2e ch 19 p 614 textbook). Adult ESC 2023, CHD: TEE favored in complex anatomy, particularly adults and larger children, but may do no better than TTE for the RVOT or distal stents.
  • Repeat echo, suspicion still high: no pediatric interval. Adult within 1 week (AHA 2026); 7–10 days (Wong 2e ch 19 p 614; Lai 3e ch 41 textbook).
  • After TPVR detection by TTE and TEE is low; cross-sectional imaging may be warranted (PAC3 Moderate / Weak). Vegetation on echo in 34% (Abdelghani 2018 abstract), 10 patients across 23 episodes (Bos 2021 abstract). ICE when TTE and TEE are both negative and suspicion is high (Low / Weak).
  • PET/CT, high risk, high suspicion, nondiagnostic echo (PAC3 Low / Strong). Studies differ, and a negative scan does not exclude: sensitivity 95.9%, specificity 88.5% (Cleveland 2026, median 21 y); 79.1%, 72.7%, negative predictive value 47.1% in suspected prosthetic pulmonary valves (Venet 2022 abstract). Adult preparation: high-fat, low-carbohydrate ketogenic diet for 24 h (AHA 2026).
  • Cardiac CT adds when criteria are already met (PAC3 Very Low / Weak). Adult TEE vs CT: vegetation 94% vs 64%, perforation 81% vs 41%, paravalvular leak 69% vs 44%; pseudoaneurysm and abscess 69% vs 78% (AHA 2026).
  • Vegetation measurement: no pediatric standard found.

Figures, in the book: Lai 3e Figs 41.16 (vegetation on a homograft in an RV-PA conduit), 41.19 (ICE into a Melody valve), 41.11 and 41.14 (ring abscess); Moss 10e Fig 62.2 (neonatal tricuspid vegetations).

Antibiotics and durations

Before any dose. Boxed warnings and key contraindications: next box. Other label lines: closed box after it. A drug in neither had no label read. AHA 2015 gives no neonate or infant doses: consult infectious diseases and pediatric clinical pharmacists.

FDA labels. Each drug is contraindicated with known hypersensitivity to it; penicillins and cephalosporins, to any of their class.

  • Gentamicin, tobramycin, amikacin, boxed warning: nephrotoxicity; ototoxicity, usually irreversible; avoid with potent diuretics such as furosemide, and with other nephrotoxic drugs, vancomycin named; regimens below pair them, so weekly levels and creatinine (AHA 2015). Amikacin: safety beyond 14 days not established.
  • Amphotericin B, boxed warning: not above 1.5 mg/kg; overdose can cause fatal cardiac or cardiopulmonary arrest; verify the product name and dose.
  • Flucytosine, boxed warning: extreme caution with impaired renal function; contraindicated in complete DPD deficiency.
  • Ceftriaxone: contraindicated in premature neonates to 41 weeks postmenstrual age and at 28 days or less with calcium-containing IV solutions; not in hyperbilirubinemic neonates.
  • Rifampin: contraindicated with several antivirals, praziquantel and lurasidone; lowers warfarin exposure, follow prothrombin time.
  • Cefepime: contraindicated with immediate hypersensitivity to cephalosporins, penicillins or other β-lactams.
  • Piperacillin-tazobactam: contraindicated with allergy to any penicillin, cephalosporin or β-lactamase inhibitor.
  • Ampicillin-sulbactam: contraindicated with serious β-lactam hypersensitivity, or prior cholestatic jaundice or hepatic dysfunction with it.
Other FDA label lines, including daptomycin, vancomycin and age limits
  • All: one manufacturer's label per drug, read on DailyMed 6 November 2026.
  • Daptomycin: avoid under 1 y; not indicated in children with renal impairment, or for left-sided S. aureus endocarditis; weekly CPK. Not studied in prosthetic valve endocarditis.
  • Rifampin: with high-dose cefazolin, postmarketing, prolonged prothrombin time, can be severe or fatal: avoid at bleeding risk; if no alternative, monitor prothrombin time and coagulation tests, vitamin K as indicated.
  • Vancomycin: hypersensitivity only; infuse over at least 60 min; kidney injury rises with exposure; confirming serum concentrations may be appropriate in children; flushing reported with anesthetic agents in children.
  • Gentamicin, tobramycin, amikacin: with anesthetics or neuromuscular blockers, consider neuromuscular blockade, respiratory paralysis or prolonged apnea. Hypersensitivity to another aminoglycoside contraindicates tobramycin and may contraindicate gentamicin or amikacin. Tobramycin, amikacin: caution in premature and neonatal infants. Tobramycin: auditory changes are irreversible. Gentamicin: more nephrotoxicity reported with cephalosporins; for children the label read points to the pediatric product's information.
  • Ceftriaxone: the calcium line includes parenteral nutrition.
  • Flucytosine, boxed warning: close hematologic, renal and hepatic monitoring. Not systematically studied in children.
  • Amphotericin B: pediatric safety and effectiveness not established in controlled studies; smallest effective dose.
  • AmBisome: hypersensitivity only; treated at 1 month to 16 y; under 1 month not established.
  • Amphotericin lipid complex: no label found.
  • Penicillin G, oxacillin, nafcillin, ampicillin: contraindicated with penicillin hypersensitivity.
  • Penicillin G: slower clearance in newborns.
  • Oxacillin, nafcillin: slower clearance in children; oxacillin, frequent levels advisable. Their labels say pediatric safety and effectiveness are not established (nafcillin: intravenous), while AHA 2015 Table 7 gives both by weight.
  • Cefazolin, ceftazidime, cefotaxime, ceftriaxone: contraindicated with cephalosporin allergy.
  • Cefazolin: not established in premature infants and neonates. Cephalosporins can lower prothrombin activity: monitor those at risk, including on anticoagulants; vitamin K as indicated.
  • Cefepime: neurotoxicity, especially with renal impairment and unadjusted doses. Established at 2 months to 16 y for listed infections, endocarditis not among them.
  • Cefotaxime: rapid central-line bolus, life-threatening arrhythmia reported.
  • Ampicillin-sulbactam: established at 1 y and older for skin infections.
  • Piperacillin-tazobactam: established at 2 months and older for listed infections.
AHA 2015 Tables 7 and 6, Pediatric, 2015: expert consensus, not comparative studies (Class IIa, LOE C). Treatment generally 4–6 weeks; longer for recurrent or prosthetic valve endocarditis and uncommon species. General recommendation: a prolonged course, at least 4, often 6–8 weeks (Class I, LOE B). Do not exceed the maximum daily or adult dose on a per kg basis. All IV unless stated.
OrganismFirst choice, daily dose; alternative; duration
Unknown organism: native (community), or prosthetic >1 y after surgeryAmpicillin-sulbactam 200–300 mg/kg/d divided every 4–6 h, up to 12 g daily, plus gentamicin 3–6 mg/kg/d divided every 8 h (Adult 3–5), with or without vancomycin 60 mg/kg/d every 6 h, up to 2 g. Prosthetic: add rifampin 15–20 mg/kg/d divided every 12 h, up to 600 mg.Cultures before the first dose. Severely ill and unstable: 3 cultures over 1–2 h, then start (AHA 2015 Class I, LOE C). Not severely ill, cultures still negative: AHA 2015 finds holding antibiotics at least 48 h for more cultures reasonable (Class IIa, LOE C; Class IIb, LOE C in its diagnosis section). ESC 2023 starts promptly after 3 sets at 30-minute intervals. When to start rifampin: sources differ, prosthetic staphylococci row. ESC 2023 Recommendation Table 10, native or prosthetic ≥12 months after surgery (IIa, C): ampicillin 200–300 mg/kg/d in 4–6 doses, up to 12 g/d, with ceftriaxone 100 mg/kg in 1 dose, or with (flu)cloxacillin 200–300 mg/kg/d in 4–6 doses and gentamicin 3 mg/kg/d in 3 doses. Allergic to penicillin (IIb, C): cefazolin 100 mg/kg/d in 3–4 doses, up to 6 g/d, or vancomycin 40 mg/kg/d in 2–3 doses, with gentamicin 3 mg/kg/d in 3 doses.Alternative: vancomycin plus gentamicin. Duration, culture-negative: native 4–6 wk, prosthetic 6 wk.
Unknown organism: nosocomial with vascular cannulae, or prosthetic ≤1 y after surgeryVancomycin 60 mg/kg/d divided every 6 h, up to 2 g, plus gentamicin 3–6 mg/kg/d divided every 8 h (Adult 3–5), with or without rifampin 20 mg/kg/d divided every 8 h, up to 900 mg/d, if prosthetic material; plus cefepime 100–150 mg/kg/d divided every 8–12 h, up to 6 g/d, or ceftazidime 100–150 mg/kg/d divided every 8 h, up to 2–4 g daily.Cultures as in the row above. When to start rifampin: sources differ, prosthetic staphylococci row. ESC 2023 Recommendation Table 10, prosthetic <12 months after surgery or healthcare-associated (IIb, C): vancomycin 40 mg/kg/d in 2–3 doses, gentamicin 3 mg/kg/d in 3 doses, rifampin 20 mg/kg/d in 3 doses, up to 900 mg/d; its text says ideally also cover non-HACEK Gram-negative organisms.Alternative: none given (Table 7 prints a question mark). Duration, culture-negative prosthetic: 6 wk.
Streptococci, highly susceptible (MBC ≤0.1 µg/mL)Penicillin G 200,000–300,000 U/kg/d divided every 4 h, up to 12–24 million U daily, or ceftriaxone 100 mg/kg/d divided every 12 h or 80 mg/kg/d every 24 h, up to 4 g daily (over 2 g, divide twice daily).Alternative: vancomycin 40 mg/kg/d divided every 8–12 h, up to 2 g daily; cefazolin 100 mg/kg/d divided every 8 h, up to 12 g daily; ceftriaxone. Cannot take a β-lactam: vancomycin, with gentamicin as current practice, 4 wk. Prosthetic material and cannot take a β-lactam: vancomycin 6 wk, gentamicin first 2 wk. Duration: native 4 wk; a 2-week course is not recommended for children, for lack of data; prosthetic material with viridans or S. bovis 6 wk, with gentamicin for the first 2 wk (Class I, LOE C).
Streptococci, relatively resistant to penicillin (MBC ≥0.2 µg/mL), includes enterococci. MBC in Table 7, MIC in its text (0.1–0.5 relatively resistant)Penicillin G 200,000–300,000 U/kg/d divided every 4 h, up to 12–24 million U daily, or ampicillin 200–300 mg/kg/d divided every 4–6 h, up to 12 g daily; plus gentamicin 3–6 mg/kg/d divided every 8 h (Adult 3–5) for the first 2 wk, or the whole course for enterococci.Alternative: vancomycin plus gentamicin for enterococci; ampicillin plus ceftriaxone for aminoglycoside-resistant enterococci or an aminoglycoside-intolerant patient; ceftriaxone plus gentamicin, not for enterococci. Abiotrophia, Granulicatella or MIC >0.5: the enterococcal regimen (Class I, LOE C). Prosthetic material, MIC >0.1: gentamicin for all 6 wk (Class I, LOE C). Enterococci: always with infectious diseases (Class I, LOE C). Duration: native 4 wk; enterococcus 4–6 wk, 6 wk with vancomycin. Adult AHA 2026 consensus: ceftriaxone alone for viridans or S. gallolyticus with intermediate penicillin resistance; favors ampicillin plus ceftriaxone for E. faecalis.
Staphylococci, susceptible to ≤1 µg/mL penicillin G (rare), as printedPenicillin G 200,000–300,000 U/kg/d divided every 4 h, up to 12–24 million U daily.This threshold and the next row's overlap in Table 7 and the statement does not reconcile them. Almost all staphylococci are highly resistant to penicillin (p 1502): penicillin G only on a qualified laboratory's susceptible result. Alternative: oxacillin or nafcillin 200 mg/kg/d divided every 4–6 h, up to 12 g/d; cefazolin 100 mg/kg/d divided every 8 h, up to 12 g daily, no type 1 penicillin allergy; vancomycin 40 mg/kg/d divided every 8–12 h, up to 2 g daily, at least 6 wk. Duration: native, oxacillin-susceptible 4–6 wk.
Staphylococci, resistant to 0.1 µg/mL penicillin G, as printedOxacillin or nafcillin 200 mg/kg/d divided every 4–6 h, up to 12 g/d, with or without gentamicin 3–6 mg/kg/d divided every 8 h (Adult 3–5 mg/kg) for 3–5 d.Overlaps the row above: see there. Gentamicin here may be considered (Class IIb, LOE C). ESC 2023 no longer recommends an aminoglycoside in native valve staphylococcal disease, for renal toxicity. Alternative: cefazolin 100 mg/kg/d divided every 8 h, up to 12 g daily, no type 1 penicillin allergy; vancomycin 40 mg/kg/d divided every 8–12 h, up to 2 g daily, if highly allergic to β-lactams, at least 6 wk. Duration: native, oxacillin-susceptible at least 4–6 wk (Class I, LOE A).
MRSA, resistant to 4 µg/mL oxacillinVancomycin 40 mg/kg/d divided every 8–12 h, up to 2 g daily.Duration: native at least 6 wk, with or without gentamicin for the first 3–5 d (Class I, LOE B). Methicillin-resistant staphylococci, including coagulase-negative: no nafcillin, oxacillin or cephalosporin, even if reported susceptible (AHA 2015). ESC 2023 native: vancomycin alone.Daptomycin: the sources conflict. AHA 2015 Table 7 offers it as the alternative, for right-sided endocarditis, maybe left-sided, and if vancomycin resistant or intolerant: 6 mg/kg every 24 h; under 6 y, 10 mg/kg. JCS 2026 Table 46: 6 mg/kg, for MRSA 6–10 mg/kg once daily, off-label for pediatric endocarditis in Japan. FDA label: not indicated for left-sided S. aureus endocarditis, poor outcomes in limited adult data; not studied in prosthetic valve endocarditis; avoid under 1 y; not indicated in children with renal impairment; pediatric indication is S. aureus bacteremia at 1–17 y, known endocarditis excluded from its trial, at 12 mg/kg 1–6 y, 9 at 7–11 y, 7 at 12–17 y, more than AHA 2015 at each age. Adult ESC 2023: 10 mg/kg with cloxacillin, ceftaroline or fosfomycin (IIb, C). AHA 2026: 8–12 mg/kg.What to do: left-sided or prosthetic, vancomycin stays first unless the organism is not susceptible (AHA 2015 Class I, LOE B). Daptomycin there is outside the label. This page's suggestion, not a label statement: decide it and its dose with infectious diseases.
All staphylococci, prosthetic material presentAdd rifampin, doses in the unknown-organism rows, plus gentamicin for the first 2 wk.Duration: at least 6 wk. In a child now: AHA 2015 Table 7 and ESC 2023 (ESC Class I, LOE B) keep gentamicin for 2 wk; JCS 2026 says consider adding it; no pediatric source read drops it. Decide with infectious diseases; if used, weekly levels and creatinine. Adult AHA 2026 consensus: gentamicin risk outweighs benefit; no consensus on rifampin. Rifampin start: ESC 2023 after 3–5 days of effective therapy, once bacteremia has cleared; Adult AHA 2026 after 5–7 days, once cleared; AHA 2015 gives no timing. FDA rifampin label, postmarketing: with high-dose cefazolin, prolonged prothrombin time, can be severe or fatal; avoid at bleeding risk; if no alternative, monitor prothrombin time and coagulation tests, vitamin K as indicated.
Gram-negative enteric bacilliCeftazidime 100–150 mg/kg/d divided every 8 h, up to 2–4 g daily; cefepime, dose in the nosocomial row; cefotaxime 200 mg/kg/d divided every 6 h, up to 12 g daily; or ceftriaxone 100 mg/kg/d divided every 12 h or 80 mg/kg/d every 24 h, up to 4 g daily. Plus gentamicin or tobramycin 3–6 mg/kg/d divided every 8 h (Adult 3–5), or amikacin 15 mg/kg/d divided every 8–12 h, by susceptibility.Alternative: a broad-spectrum penicillin, such as piperacillin/tazobactam 240 mg/kg/d divided every 8 h, up to 18 g daily, plus an aminoglycoside. Duration: at least 6 wk (Class I, LOE C). Choose with infectious diseases, by susceptibility.
HACEKCefotaxime 200 mg/kg/d divided every 6 h, up to 12 g daily; or ceftriaxone 100 mg/kg/d divided every 12 h or 80 mg/kg/d every 24 h, up to 4 g daily; or ampicillin-sulbactam 200–300 mg/kg/d divided every 4–6 h, up to 12 g daily.Alternative, susceptible organisms: ampicillin 200–300 mg/kg/d divided every 4–6 h, up to 12 g daily, plus an aminoglycoside. Duration: 4 wk (Class I, LOE C).
Fungi: Candida, AspergillusSurgical resection plus amphotericin B 1 mg/kg/d over 3–4 h, or liposomal or lipid-associated 3–5 mg/kg/d as a single daily dose, with or without flucytosine 150 mg/kg/d orally divided every 6 h (100–150 in the text).FDA label: conventional amphotericin B not above 1.5 mg/kg; verify the product name and dose. Our reading: the 3–5 mg/kg here is liposomal or lipid only. Surgery with antifungals for most, Class I, LOE C. Flucytosine: Candida with a susceptible strain, Class IIb, LOE C, with blood levels (p 1504). Table 7 gives no duration: JCS 2026 Table 46 8 wk; Adult AHA 2026 6 wk. Alternative, if surgery cannot be done: amphotericin B then imidazole suppression, possibly lifelong.

Levels and kidneys.

  • Gentamicin for synergy: peak 3–4 µg/mL, trough <1 µg/mL (AHA 2015 Class IIa, LOE B). Gram-negative rods: up to 7.5 mg/kg/d, peak 5–10, trough <1 to 1.5.
  • Gentamicin dosing: Doses here are divided every 8 h. Once daily: insufficient experience in children with endocarditis (AHA 2015 Class III, LOE C). ESC 2023, at 3 mg/kg/d: 1 dose preferred for staphylococci, 1 or 3 for streptococci, 3 divided for enterococci and empiric treatment; once daily, trough <1, peak about 10–12 mg/L.
  • Gentamicin toxicity: Gentamicin adds renal and otic toxicity (Class IIb, LOE C). The 4-week penicillin route for highly susceptible streptococci avoids aminoglycosides with renal impairment, other nephrotoxic drugs or eighth nerve impairment. Vancomycin with gentamicin: levels and creatinine weekly (Class IIb, LOE C).
  • Renal failure: doses need adjusting, with pharmacy, infectious diseases or renal guidance (AHA 2015 Class I, LOE B); no renal doses are shown.
  • Adult AHA 2026: gentamicin risk outweighs benefit in staphylococcal prosthetic valve endocarditis; ESC 2023 keeps it for 2 weeks.
  • Vancomycin differs by source: AHA 2015 40 mg/kg/d in organism rows, 60 in unknown-organism rows, 40–60 starting in its text; trough usually 10–15 µg/mL, 15–20 µg/mL for MRSA with MIC >1 or poor response. JCS 2026 Table 46: 15–20 mg/kg 4 times daily under 13 y (60–80 mg/kg/d). ESC 2023: 30 mg/kg/d in 2–3 doses targeted, 40 mg/kg/d in 2–3 doses empiric, trough 10–15 mg/L; no more than 2 g/d unless levels are monitored. Adult ESC 2023 text prefers an AUC/MIC target of 400–600 mg·h/L, MIC 1 assumed, within 48 h. No pediatric source read sets an AUC target.
  • Pediatric doses in ESC 2023 Recommendation Tables 7–10, where they differ from Table 7. Gentamicin 3 mg/kg/d: streptococci 1 dose or 3 divided; staphylococci 1 (preferred) or 2; enterococci 3 divided; maximum 240 mg/day, weekly renal function and levels. Cefazolin 100 mg/kg/d in 3–4 doses, up to 6 g/d. Penicillin G 200,000 U/kg/d in 4–6 doses. Ceftriaxone 100 mg/kg/d in 1 dose for streptococci, 2 for enterococci. (Flu)cloxacillin 200–300 mg/kg/d in 4–6 doses. Rifampin 20 mg/kg/d in 3 doses, up to 900 mg/d. MRSA native: vancomycin 4–6 wk, Class I, LOE B. Enterococci: ampicillin or amoxicillin with ceftriaxone for 6 wk or with gentamicin for 2 wk, Class I, LOE B.
  • IV, not intramuscular, because of small muscle mass (AHA 2015 Class I, LOE C). Culture-negative regimens: AHA 2015 Table 8.
  • Oral step-down, children: no regimen to show. PAC3 2025: insufficient data to recommend. The trial behind it enrolled adults (AHA 2026).
  • Adult durations, AHA 2026 Tables 1–4, native and prosthetic: staphylococcal 6 and 6 wk; viridans or S. gallolyticus 4 and 6; enterococcal 6 and 6; HACEK 4 and 6.
Antibiotics after valve surgery: unsettled, not studied in children.
Adult AHA 2026, consensusJCS 2026, adult-framedPediatric
Planned 4–6 wk less the days given before surgery, at least 14 days after it; do not restart the clock. Culture or molecular positivity does not change the duration. Course finished, no active infection, negative tissue cultures: no more antibiotics.Tissue culture positive: full course from the day of surgery. Negative: full course counting days before surgery.AHA 2015 gives none. Russell 2013 gave 6 wk after surgery abstract.
Follow-up

No pediatric schedule exists in anything read: echo interval, duration and culture timing are gaps.

  • Adult AHA 2026: TTE at end of therapy as the new baseline; TTE 6–12 weeks after valve surgery; TEE at 3 months may be considered after complex operations; then by residual lesion, function and substitute. JCS 2026: most careful in the first year, no standard interval.
  • Blood cultures after treatment may be considered, but may isolate a contaminant (AHA 2015).
  • Recurrence, surgical series: free of recurrence 94.7% at 25 y (Khoo 2019); 5 of 47, 10.6% (Carrillo 2023). Adult 2–9% (JCS 2026). After medical-only treatment of TPV endocarditis, 33% of those readmitted had relapsed (Fox 2023 abstract).
  • Reoperation is frequent and late, usually not for recurrence (Wu 2023 abscess series).
  • PAC3 2025: dental clearance before and after TPVR, close monitoring, education, rapid cardiac assessment for unexplained fever Very Low / Strong.
Prophylaxis

Before any dose. FDA labels: each drug is contraindicated with known hypersensitivity to it; further lines only are listed.

  • Amoxicillin: also serious hypersensitivity to other β-lactams; the label modifies dosing at 12 weeks or younger.
  • Ampicillin: any penicillin.
  • Cefazolin, ceftriaxone, cephalexin: any cephalosporin. Cephalosporins not after anaphylaxis, angioedema or urticaria with penicillin or ampicillin (AHA 2021 Table 5). Cephalexin capsules: under 1 y not established.
  • Ceftriaxone: contraindicated in premature neonates to 41 weeks postmenstrual age and at 28 days or less with calcium-containing IV solutions; not in hyperbilirubinemic neonates.
  • Azithromycin: also any macrolide or ketolide; contraindicated after cholestatic jaundice or hepatic dysfunction with it; QT prolongation and torsades reported, can be fatal; under 6 months not established for its listed infections.
  • Clarithromycin: also any macrolide; contraindicated with cisapride, pimozide, lomitapide, lovastatin, simvastatin, ergot alkaloids, lurasidone, with colchicine in renal or hepatic impairment, and after cholestatic jaundice or hepatic dysfunction with it; avoid with known QT prolongation; under 6 months not established.
  • Azithromycin or clarithromycin: caution with a known prolonged QTc; doxycycline is the alternative (AHA 2021).
  • Doxycycline: tooth discoloration, more common with long-term use, also after repeated short courses; the label limits use at 8 y or younger to severe or life-threatening conditions, while AHA 2021 Table 5 gives it by weight, one dose, no age limit printed.
  • Clindamycin: no longer recommended (AHA 2021 Table 5).
AHA 2021 Table 5. Dental procedures that manipulate gingiva or the periapical region or perforate oral mucosa. Single dose, 30–60 min before; if missed, up to 2 h after.
SituationAgentChildrenAdult
OralAmoxicillin50 mg/kg2 g
Unable to take oralAmpicillin50 mg/kg IM or IV2 g IM or IV
Cefazolin or ceftriaxone50 mg/kg IM or IV1 g IM or IV
Allergic to penicillin or ampicillin, oralCephalexin50 mg/kg2 g
Azithromycin or clarithromycin15 mg/kg500 mg
Doxycycline<45 kg, 2.2 mg/kg; >45 kg, 100 mg100 mg
Allergic, unable to take oralCefazolin or ceftriaxone50 mg/kg IM or IV1 g IM or IV

ESC 2023 Table 6 prints the same child doses with maxima: amoxicillin and ampicillin 2 g; cefazolin or ceftriaxone 1 g; cephalexin 2 g; azithromycin or clarithromycin 500 mg.

  • AHA 2021 suggests it for: prosthetic valve or material, including transcatheter valves and repair devices; ventricular assist device or implantable heart; previous, relapsing or recurrent endocarditis; unrepaired cyanotic CHD, including palliative shunts and conduits; CHD repaired with prosthetic material or device, first 6 months; repaired CHD with a residual defect at or next to it; surgical or transcatheter pulmonary valve or conduit such as Melody and Contegra; heart transplant with valvulopathy. Not suggested: pacemakers and similar devices, septal closure devices with complete closure, peripheral grafts and patches, stents, ventriculoatrial shunts, vena cava filters, pledgets.
  • AHA 2026 does not change it: it points to the 2021 statement and lists no groups of its own. PAC3 2025 leaves prophylaxis out of scope.
  • Pediatric AHA 2015: highest risk, Class IIb, LOE C.
  • Adult ACC/AHA 2020 IIa C-LD, before dental procedures that manipulate gingiva or the periapical region or perforate mucosa: prosthetic valve including transcatheter and homograft, prosthetic repair material, previous endocarditis, unrepaired cyanotic CHD, repaired CHD with residual shunt or regurgitation at or near a patch or device, transplant with valve regurgitation from a structurally abnormal valve. Not for nondental procedures without active infection III: No benefit B-NR.
  • Adult ESC 2023, Class I: previous endocarditis (LOE B); surgical prosthetic valve or repair material, transcatheter aortic and pulmonary valves, untreated cyanotic CHD, palliative shunts or conduits, ventricular assist devices (LOE C); after repair with no residual defect, the first 6 months only. Heart transplant IIb, C.
  • JCS 2026: high risk, Strength 1, Certainty B; moderate risk, Strength 2, Certainty D. Its pediatric groups add aortic coarctation to high risk; its list includes a systemic-pulmonary shunt and LVAD.
  • England, all ages: repaired CHD with prosthetic material carried lower risk than moderate-risk groups, even in the first 6 months, against the AHA 6-month rule (Thornhill 2018 abstract).
Where the sources disagree
Side by side, not resolved here.
QuestionOne sourceAnotherWhy it stays open
Pediatric against adult surgical classes: operate to prevent a first embolus?Pediatric AHA 2015 Class III, LOE C: lack of proven benefit, long-term risks of valve replacement. JCS 2026 pediatric section: no clear data.Adult JCS 2026 IIb, C (mobile >10 mm, low surgical risk). ESC 2023 IIb, B (≥10 mm, low surgical risk). ACC/AHA 2020 IIb, B-NR. AHA 2026 lists large vegetations (>1 cm), no class.No pediatric study tests a size. Series report it: larger vegetation and embolic events, mitral OR 1.15 (1.02–1.31), aortic OR 1.08 (1.00–1.16), no unit printed (Wu 2023, Wu 2024); mobile >10 mm 30% vs 11% with and without neurologic complications, p=0.048 (Jacobwitz 2023); vegetation indexed to BSA HR 1.06 per mm/m² for death, left-sided (Khoo 2019).
Does surgery lower death in CHD?Pediatric PAC3 2025: surgery is an independent predictor of lower in-hospital mortality Moderate / Strong. It rests on Yoshinaga 2008, 137 aged 1 month to 62 y: surgery OR 0.045 (0.003–0.70) in its stepwise model, univariate OR 0.38 (0.12–1.20), p=0.10. 14 deaths, 5 factors in the model; crude death 5 of 78 operated, 9 of 59 not.Pediatric Elsisy 2026, 2108 CHD patients, after PAC3's search: in-hospital death 64 (10.1%) surgical vs 142 (9.6%) medical, p=0.72.Elsisy is unmatched, from codes, and the surgical group was sicker, so equal rates there do not show surgery is neutral. PAC3 itself notes emerging adult CHD data with equal mortality.
Antibiotics after valve surgeryAdult AHA 2026: planned course less days given, at least 14 days after surgery; positivity does not change it.JCS 2026: culture positive, full course from surgery; negative, full course counting days before surgery.Not studied in children; table in Antibiotics and durations.
Pediatric regimen against the 2026 adult shift: staphylococcal prosthetic valvePediatric AHA 2015: add rifampin, plus gentamicin first 2 wk. JCS 2026 Table 46: consider adding; methicillin-susceptible at least 2 wk, methicillin-resistant 2 wk. ESC 2023: gentamicin 2 wk, Class I, LOE B.Adult AHA 2026: gentamicin risk outweighs benefit; no consensus on rifampin.ESC 2023 keeps it while citing a study in S. aureus prosthetic valve disease with no better outcome, and its nephrotoxicity. No pediatric update read. E. faecalis differs the same way: AHA 2015 penicillin or ampicillin plus gentamicin, ceftriaxone if intolerant; AHA 2026 favors ampicillin plus ceftriaxone.
Once-daily gentamicin in childrenPediatric AHA 2015 Class III, LOE C: insufficient experience in endocarditis.ESC 2023 pediatric doses: 1 dose preferred for staphylococci, 1 or 3 for streptococci, 3 divided for enterococci and empiric treatment.AHA 2015: no published outcomes in children with endocarditis.
TPV endocarditis: surgery or notPAC3 2025: favor medical therapy if mild and stable Moderate / Strong.Fox 2023 favors surgery abstract, unadjusted; sicker patients were operated on.No controlled comparison.
Routine TEE before surgeryWu 2023 abscess series screens routinely.PAC3 2025 and Humpl 2003 keep it selective.Neither is a controlled comparison.
RVOT gradient in DukePAC3 2025: minor criterion High / Strong.Gafner 2025 proposes major: 11 cases vs 15 febrile controls abstract.Small.

Delay after stroke: the table in Stroke timing.

Tags: abstract authors' abstract only. textbook not the original study. PAC3 prints quality / strength; AHA 2026 prints no classes.

Key points

Sources and how this page was checked

Full text read

Abstracts only

Tagged abstract where used. Results not read, so these may change.

Textbooks

Tagged textbook with edition, chapter and page: a textbook, not the original study. Paraphrased. No figure, table image or page render from any book is on this page; figure numbers are pointers to the book.

Not read, named only

How it was checked