Kawasaki Disease Evaluation and Management

Diagnostic / Management

Diagnostic criteria selector

AHA 2024 Figure 1 (corrected 2025), wording as printed. Shows your position in the figure, not a treatment decision. Nothing is saved or sent.

Fever

Fever for at least 4 d (the day of fever onset = day 1 of fever)No temperature threshold is given in either statement.

Principal clinical features

At any point during the illness (does not need to be concurrent).

Alternative disease process

What the statements leave undefined
  • Temperature for the diagnostic fever: none in either statement.
  • “Prolonged” unexplained fever: AHA 2024 gives no number of days. AHA 2017 Figure 3 printed ≥5 days. The selector asks you to judge it.
  • “Infants” and “7 d”: no age range and no ≥ sign in AHA 2024. AHA 2017 Figure 3 printed ≥7 days, and says infants ≤6 months are the most likely to have prolonged fever without other criteria.
  • Laboratory values: no hemoglobin value for “anemia for age”, no threshold for “elevated ALT”, no units on the platelet count. GGT and TBili are ordered (Section A) but are not criteria.
  • “≥3 other suggestive features”: an open list (“including”) in both statements.
  • Z-score system for the ≥2.5 cut-off: not named in either figure. AHA 2024 names the McCrindle 2007 formulas for its risk classification, and asks centers to use one equation over time.
  • How Section D’s three items combine: the stem prints “laboratory or echocardiographic findings” and “Or” stands only before the third. This page reads all three as alternatives (method section).
  • 0 or 1 feature in a child who is not an infant: AHA 2017’s text says “fewer than 4”; its figure and AHA 2024 say 2 or 3.
  • Fewer than 4 days of fever with ≥4 features: only that experienced clinicians may establish the diagnosis earlier, at 3 days of fever (AHA 2024 text, no rule).
  • “Decision to treat per KD team?”: no inputs are stated.
  • “Illness not explained by known alternative disease process”: a judgement. The statements list mimics but give no rule for how many must be excluded or how.
  • The 10-day window is in AHA 2024’s text and in the AHA 2017 recommendations. It is not in AHA 2024’s figure, and only a footnote in AHA 2017’s.
  • KD versus MIS-C: qualitative differences only, no criteria or cut-offs.
  • Section D without the selector: CRP ≥3 mg/dL or ESR ≥40 mm/h, or both, +3 or more of anemia for age, platelets ≥450,000, albumin ≤3 g/dL, elevated ALT, elevated WBCs ≥15,000/mm³, urine WBCs ≥10/hpf; or Z score of LAD CA or RCA ≥2.5; or ≥3 other suggestive features. In the selector these open after fever Yes with 2–3 features and the prolonged tick, or with the infant tick.
Findings that argue against KD, and the differential
  • A clinical diagnosis without a pathognomonic test (AHA 2024).
  • Not characteristic of KD (AHA 2024): oral ulcerations, exudative pharyngitis, exudative or unilateral conjunctivitis, vesicular rash.
  • Also prompt another diagnosis (AHA 2017): splenomegaly; bullous or petechial rash; generalized adenopathy.
  • Differential (AHA 2017 Table 3): measles; other viral infections (eg, adenovirus, enterovirus); staphylococcal and streptococcal toxin-mediated diseases (eg, scarlet fever, toxic shock syndrome); drug hypersensitivity reactions, including Stevens Johnson syndrome; systemic onset juvenile idiopathic arthritis. With epidemiologic risk factors: Rocky Mountain spotted fever or other rickettsial infections; leptospirosis.
  • MIS-C (added by AHA 2024): more prominent gastrointestinal symptoms, headache, thrombocytopenia, lymphopenia, elevated troponin or BNP, decreased left ventricular systolic function, pericardial effusion. More common in KD: rash, conjunctival injection, oral mucosal changes, coronary abnormalities.
  • A positive test does not settle it (AHA 2017): adenovirus with exudative pharyngitis and conjunctivitis makes KD extremely unlikely, but KD stays possible with nonexudative pharyngitis. With group A streptococcus and no improvement after 24 to 48 hours of effective antibiotics, consider KD again.
  • Consider KD in (AHA 2017): infants <6 months with prolonged fever and irritability; infants with prolonged fever and unexplained aseptic meningitis; infants or children with prolonged fever and unexplained or culture-negative shock, cervical lymphadenitis unresponsive to antibiotics, or retropharyngeal or parapharyngeal phlegmon unresponsive to antibiotics.
What each feature looks like (AHA 2017)
  • Rash: usually within 5 days of fever onset; most often diffuse maculopapular, also scarlatiniform erythroderma or erythema multiforme-like; accentuation in the groin with early desquamation is characteristic.
  • Eyes: begins shortly after fever onset and often spares the limbus.
  • Mouth: erythema, dryness, fissuring, peeling, cracking and bleeding of the lips; strawberry tongue with prominent fungiform papillae; diffuse oropharyngeal erythema.
  • Hands and feet: erythema of palms and soles and firm, sometimes painful induration in the acute phase; periungual desquamation usually within 2 to 3 weeks of fever onset; Beau’s lines at 1 to 2 months.
  • Nodes: the least common feature; usually unilateral, ≥1.5 cm, in the anterior cervical triangle. May be the only initial finding and be mistaken for bacterial lymphadenitis.

Key points

Evaluation and management pathway

Adapted from AHA 2024 Figure 1, corrected.

KD suspected in the pediatrician’s office or emergency department
Perform laboratory testing and consult infectious disease/rheumatology/cardiology or local KD team for evaluation Laboratory testing: CBC with manual differential, ESR, CRP, basic metabolic panel, ALT, GGT, TBili; bagged or clean-catch UA + microscopy.
Complete KD criteria met ECG and echo. Next: the first-line treatment box.

or

Suspected incomplete KD ECG and echo. A normal echo does not exclude KD (AHA 2024 text).
Decision to treat per KD team? No: consider repeat evaluation if fever persists. Exit pathway. Yes: the first-line treatment box.
1. First-line treatment: IVIG and aspirin (doses, time limits). 2. Assess for high-risk features.
Any clinical or echo high-risk features present? ≤6 mo of age; LAD or RCA CA z-score ≥2.5 on baseline echo. AHA 2024’s text prints <6 months, which differs at exactly 6 months.
Yes: intensification. No: reevaluation at 36 h, then discharge

Yes

Intensification of initial therapy (AHA 2024 Table 3). Clinical and laboratory monitoring by KD experts. Repeat echo Q2–3 d or as per cardiology. AHA 2024 Figure 2 for management of evolving coronary arteries. AHA 2024: patients with z ≥2.5 or age <6 months “should be considered for intensification of primary therapy” (Diagnosis Key Point 5). Treatment Key Point 1 and text: intensification “may benefit”; randomized efficacy data are lacking.

No

Reevaluation at 36 h after completion of first dose of IVIG
Fever ≥38°C? Yes: the intensification box. No: discharge.
  • Discharge home
  • Transition to low-dose aspirin
  • No live vaccination for 11 mo post IVIG
  • Follow-up cardiology visit with echo within 7–14 d or sooner for high-risk (confirm date with cardiology or KD team at discharge)
  • Provide KD education (return precaution for fever or return of KD symptoms)
Exit pathway.

Long-term management by risk level: Kawasaki Coronary Z-Scores.

First-line doses and time limits

AHA 2024 Table 3
AgentDose
IVIG2 g/kg IV over 8–12 h. Consider slower infusion in KD shock or myocardial dysfunction, and lean body mass dosing in obesity.
Aspirin30 to 50 mg/kg per d PO divided every 6 h (used in many centers; the dose is under prospective study), then 3–5 mg/kg per d PO once afebrile for 48–72 h. Not with NSAIDs.

IVIG time limits, graded, AHA 2017 (p. e945)

ClassRecommendationLOE
IComplete KD criteria, or meeting the algorithm criteria for incomplete KD: high-dose IVIG (2 g/kg as a single intravenous infusion) within 10 days of illness onset, as soon as possible after diagnosis.A
IIaReasonable after the 10th day of illness with persistent fever without other explanation, or coronary artery abnormalities together with ongoing systemic inflammation (ESR or CRP elevated; CRP >3.0 mg/dL).B
IIIIVIG generally should not be given beyond the tenth day of illness in the absence of fever, significant elevation of inflammatory markers, or coronary artery abnormalities.C
More on IVIG and aspirin
  • IVIG resistance: persistent or recrudescent fever ≥36 hours after the completion of the initial infusion (AHA 2024).
  • IVIG: hemolytic anemia is dose dependent and more frequent with A, B or AB blood groups. ESR is not a reliable marker of inflammation after IVIG (AHA 2024).
  • When aspirin cannot be used: for example glucose-6 phosphate dehydrogenase deficiency or acute influenza infection, an alternative antiplatelet class can be used (Table 3).
  • How long low-dose aspirin continues: AHA 2024 prints 6 to 8 weeks after onset of illness in its text and 6 wk at risk level 1 in its Table 2. AHA 2017 prints both 6 to 8 weeks (p. e945) and 4 to 6 weeks (its Class I row and Tables 7 and 11). By risk level: Kawasaki Coronary Z-Scores.
Intensification and refractory disease (AHA 2024 Table 3)

Doses as printed in AHA 2024 Table 3; check with your KD team or pharmacy. AHA 2024 Diagnosis Key Point 5: patients with z ≥2.5 at diagnosis or age <6 months “should be considered for intensification of primary therapy”. Its Treatment Key Point 1 and text say intensification “may benefit” high-risk patients (baseline z ≥2.5, infants <6 months, or high risk on the Son score, on Kawasaki Coronary Z-Scores), and that efficacy data from randomized controlled trials are lacking.

AgentDose, as printed
Intensification, IVIG-resistant or high-risk patients
Prednisoloneas per RAISEIV, 2 mg/kg per d divided every 8 h for 5 d (maximum 60 mg/d) while hospitalized; then PO prednisolone 2 mg/kg per d divided every 8 h; slow tapering over 15 d (maximum 30 mg/dose) once CRP normalized. Famotidine is given to prevent gastric ulcer.Unresolved: 60 mg/d divided every 8 h is 20 mg per dose, yet the taper prints a maximum of 30 mg/dose and no taper frequency. The RAISE abstract gives only 2 mg/kg per day given over 15 days after CRP normalised abstract. AHA 2024: the RAISE results are “compelling, but have not been validated in a non-Japanese population”.
Methylprednisoloneas per North American studiesIV, 2 mg/kg per d divided every 12 h for 5 d (maximum 60 mg/d) while hospitalized; then PO prednisone 2 mg/kg per d divided every 12 h; once CRP <1 mg/dL, corticosteroids are tapered over 2–4 wk with the dose cut in half every 5 d. Famotidine is given to prevent gastric ulcer.
InfliximabIV, 10 mg/kg given over 2 h.The dose in KIDCARE, a randomized comparison with second IVIG abstract. AHA 2017 Table 6 printed 5 mg/kg, the dose of an earlier phase I randomized trial.
EtanerceptSC, 0.8 mg/kg weekly ×3 doses.
Additional therapy for refractory disease
CyclosporinePO, 5 mg/kg per d divided every 12 h; check 2 h level after 3rd dose (goal of 300–600 ng/mL); start to taper (by 10% every 3 d) once patient afebrile, clinically improving, and CRP ≤1.0 mg/dL or 10 d of therapy, whichever is longer. Use Neoral; Mg protein complex (133 mg) once daily; not with statins.
AnakinraIV/SC, 10 mg/kg per d (IV divided q12 h preferred to SC) while hospitalized; wean once ready for discharge (5 mg/kg per d for 1 d, then stop).AHA 2024 cites 2–11 mg/kg per day, IV or SC, in 22 patients with coronary aneurysms (Yang 2022, phase I/IIa abstract), and 2 to 6 mg/kg per day SC in 16 IVIG-resistant patients (Koné-Paut 2021, phase II abstract). AHA 2017 Table 6 printed 2–6. AHA 2024: no data regarding efficacy in patients with coronary abnormalities.
Second dose of IVIGIV, 2 g/kg given over 8–12 h. Increased risk of hemolytic anemia with A, B, and AB blood types.
CyclophosphamideConsult rheumatology. Table 3 prints: IV, 10 mg/kg per d in 1 or 2 doses; adequate hydration needed; consider rheumatology consult. It prints no duration.How the series AHA 2024 cites gave it: Halyabar 2021. AHA 2024: in giant aneurysms continuing to progress despite IVIG and other adjunctive anti-inflammatory therapies, cyclophosphamide “has been used in patients with KD who are refractory to multiple other therapies”, usually in consultation with rheumatologists.

Graded rows on these options, from AHA 2017 (pp. e947, e949). They postdate RAISE and predate KIDCARE and the anakinra studies. Class and level as published.

ClassRecommendationLOE
IIbHigh-risk patients, when identifiable before treatment: a longer course of corticosteroids (eg, tapering over 2–3 weeks) with IVIG 2 g/kg and aspirin may be considered.B
IIISingle-dose pulse methylprednisolone should not be given with IVIG as routine primary therapy.B
IIaIVIG-resistant: a second dose of IVIG (2 g/kg) is reasonable with persistent or recrudescent fever at least 36 hours after the end of the first infusion.B
IIbIVIG-resistant: high-dose pulse steroids (usually methylprednisolone 20–30 mg/kg IV for 3 days, with or without an oral prednisone taper) may be considered as an alternative to a second infusion, or for retreatment.AHA 2017 contradicts itself here: its Table 6 footnote says pulsed high-dose steroid treatment is not recommended.B
IIbRetreatment: a longer (eg, 2–3 weeks) tapering course of prednisolone or prednisone with IVIG 2 g/kg and aspirin may be considered.B
IIbIVIG-resistant: infliximab (5 mg/kg) may be considered as an alternative to a second infusion or corticosteroids.C
IIbRefractory, after a second infusion, infliximab or a course of steroids has failed: cyclosporine may be considered.C
IIbHighly refractory: immunomodulatory monoclonal antibody therapy (except TNF-α blockers), cytotoxic agents, or (rarely) plasma exchange may be considered.C
Other graded rows, first-line (AHA 2017; AHA 2024 has no classes)

AHA 2024 carries no class of recommendation or level of evidence, so these rows are from AHA 2017 (pp. e943, e945). Paraphrased; class and level as published.

ClassRecommendationLOE
IIaModerate- (30–50 mg/kg/d) to high-dose (80–100 mg/kg/d) aspirin is reasonable until afebrile, although there is no evidence that it reduces coronary artery aneurysms.C
IEchocardiography when the diagnosis is considered; unavailability or technical limitations should not delay treatment.B
IImage the coronary arteries, with luminal dimensions normalized as z scores adjusted for body surface.B

Prognostic / Outcome

Medium-term outcomes by maximum coronary z (IKDR)

International Kawasaki Disease Registry, 1651 patients with an aneurysm (maximum z ≥2.5). The paper reports complications almost entirely confined to maximum z ≥10, where risk kept rising over time. The Results report one exception: a patient with a medium aneurysm (maximum z 8.5, all 4 branches) who died of chronic ischemia 6 months after the acute phase.

A prognosis table, not a risk calculator. The bands use z alone and drop the AHA 8 mm rule, so do not map onto AHA risk levels.

IKDR 2020 Results, Table 2, Limitations. Band: maximum z of 4 branches.
BandSmall2.5≤Z<5Medium5≤Z<10LargeZ≥10
Patients854357440
Median follow-up, y (IQR)1.2 (0.4–3.4)2.8 (1.0–6.3)5.2 (1.9–8.9)
Had angiography or MRI1%5% to 8%≈50%
10-year cumulative incidence per patient (± standard error)
Luminal narrowing (>50%, on angiography)none reportednone reported20±3%
Coronary artery thrombosisnone reportednone reported18±2%
Composite major adverse cardiac complicationsnone reported1 patient14±2%
Normalization of luminal diameter to Z <2.5, 10 years (± as printed)99±4% as printed92±1%57±3%
Other 10-year outcomes
Per patient (± standard error)SmallMediumLarge
Coronary artery interventionsnone reportednone reported10±2%
Myocardial infarctionnone reportednone reported5±2%
Death from cardiac causes or heart transplantationnone reported1 patient (the same one)2±1%
Cohort, measurement and other limits
  • Everyone had z ≥2.5, so nothing here speaks to dilation only or normal coronaries.
  • Cohort: 34 pediatric referral centers in Canada, the United States and Taiwan; age <18 years; complete or incomplete KD diagnosed 1999 to 2017. Median follow-up 2.1 years overall (IQR 0.9–6.4).
  • Z scores: from echocardiography reports with no core laboratory; Boston equations with Haycock BSA; the LAD equation was used for the LCX. A patient’s band rests on one maximum z, which measurement error can move across a boundary.
  • Horizon: the title says medium-term. Estimates stop at 10 years because too few patients remained at risk, and numbers at risk are not printed in the text.
  • Per-branch results in the paper are a different denominator, restricted to branches with z ≥10. They are not shown here, so nothing in the table mixes the two.
  • Design: retrospective, with no auditing of data at centers. Outcomes were often found at the time of imaging rather than when they occurred.
  • Within z ≥10: the authors propose z ≥20 as a separate, higher-risk category. That is their proposal, not an AHA band.
  • Not answered here: treatment effect. Aspirin and long-term antithrombotics, beta-blockers and statins were left out of the risk-factor analysis.

Criteria selector, pathway, doses, outcomes by coronary z band

Related on HeartCalc

Sources and how this page was checked

Full text read

Abstract only

How the selector reads Figure 1

Fever duration across the statements

Where AHA 2024 Figure 1 prints the criteria differently from AHA 2017

AHA 2024 says the diagnostic criteria have not changed. As printed, these differ. Feature numbering differs too, so this page never cites a feature by number.

ItemAHA 2017 (Table 3, Figure 3)AHA 2024 (Figure 1, corrected)
RashMaculopapular, diffuse erythroderma, or erythema multiforme-likePolymorphous rash
Hands and feetErythema and edema in the acute phase and/or periungual desquamation in the subacute phasePalmar and plantar erythema, usually with swelling; desquamation described as what follows
Nodes≥1.5 cm diameter, usually unilateralUsually unilateral, cluster of nodes ≥1.5 cm
ExclusionIn prose and a figure footnoteA bullet in the list
Incomplete, childrenFever ≥5 days and 2 or 3 criteriaProlonged unexplained fever and 2–3/5
Incomplete, infantsFever for ≥7 days without other explanationUnexplained fevers 7 d
Platelets≥450,000 after the 7th day of fever≥450,000
Echo routesZ ≥2.5; aneurysm observed; or ≥3 other suggestive featuresZ ≥2.5; or ≥3 other suggestive features
Laboratory-negativeCRP <3.0 mg/dL and ESR <40 mm/hr: serial re-evaluation if fevers persist, echo if typical peeling developsNo such branch; the arm goes to the KD team decision
Incomplete, outcomeTreat box (AHA 2017 Figure 3)Decision to treat per KD team?
High-risk featuresNot in the figure≤6 mo of age; LAD or RCA z-score ≥2.5 on baseline echo

Doses

How the outcome table was built

Not covered