Diagnostic Criteria
Complete Kawasaki Disease
The diagnosis of complete KD is clinical and requires fever plus at least 4 of the 5 principal clinical features:
- Bilateral bulbar conjunctival injection without exudate
- Oral mucosal changes (erythema, cracking of lips, strawberry tongue, diffuse oropharyngeal erythema)
- Polymorphous rash (maculopapular, diffuse erythroderma, or erythema multiforme-like)
- Extremity changes (acute: erythema and edema of hands and feet; subacute: periungual desquamation)
- Cervical lymphadenopathy (≥1.5 cm diameter, usually unilateral)
Regarding fever duration: in the presence of ≥4 principal clinical features, the diagnosis can be made with 4 days of fever. Experienced clinicians may establish the diagnosis even earlier, at 3 days of fever, when the clinical picture is classic. The traditional "≥5 days of fever" criterion applies when fewer than 4 principal features are present.
Use the Kawasaki Diagnostic Criteria tool to work through both complete and incomplete cases systematically.
Incomplete Kawasaki Disease
Approximately 25-30% of children with KD do not meet complete diagnostic criteria. Incomplete KD should be considered in any infant or child with prolonged unexplained fever (≥5 days) and 2-3 principal clinical features. Infants under 6 months are at particularly high risk for incomplete presentations and subsequent coronary complications.
The diagnostic algorithm for incomplete KD incorporates:
Laboratory findings: CRP ≥3.0 mg/dL or ESR ≥40 mm/hr, plus ≥3 of the following supplemental criteria: albumin ≤3.0 g/dL, anemia for age, elevated ALT, platelets ≥450,000/μL after day 7, WBC ≥15,000/μL, urine ≥10 WBC/high-power field.
Echocardiographic findings: A coronary artery z-score ≥2.5 for the left anterior descending or right coronary artery has very high specificity for the diagnosis and supports treatment in patients with incomplete presentations.
These laboratory and echocardiographic findings are supportive criteria used in conjunction with clinical assessment. They are not standalone diagnostic thresholds. The finding of coronary abnormalities in a febrile child with compatible clinical features essentially confirms the diagnosis. Use the Coronary Z-Score Calculator for objective coronary assessment.
Clinical Vignettes: Applying the Incomplete KD Algorithm
Vignette 1: Classic Incomplete KD in an Infant
A 4-month-old male presents with 6 days of fever (39.5°C), bilateral conjunctival injection, and erythematous cracked lips. No rash, extremity changes, or lymphadenopathy are noted. Laboratory evaluation reveals CRP 8.2 mg/dL, ESR 65 mm/hr, albumin 2.8 g/dL, hemoglobin 9.1 g/dL, and platelets 520,000/μL.
Analysis: This infant has only 2 principal clinical features but meets laboratory criteria (CRP ≥3.0 mg/dL with ≥3 supplemental criteria: hypoalbuminemia, anemia, thrombocytosis). Echocardiography should be obtained promptly. If the echocardiogram shows coronary z-score ≥2.5, the diagnosis is confirmed. Even with normal coronary arteries, treatment with IVIG is indicated given the high-risk age group and supportive laboratory findings. Infants under 6 months are at highest risk for missed diagnosis and coronary complications.
Vignette 2: Incomplete KD with Coronary Involvement
A 2-year-old female presents with 7 days of fever, polymorphous rash, and cervical lymphadenopathy (2.0 cm). No conjunctival injection, oral changes, or extremity findings are present. CRP is 5.4 mg/dL and ESR is 48 mm/hr. Echocardiography reveals left anterior descending coronary artery z-score of 3.2.
Analysis: This child has only 2 principal clinical features but has coronary artery involvement (z-score ≥2.5). The finding of coronary artery z-scores ≥2.5 in a febrile child with compatible clinical features has very high specificity for KD and essentially confirms the diagnosis. Treatment with IVIG should be initiated immediately. This patient also qualifies as high-risk due to coronary involvement at presentation and may benefit from intensified initial therapy.
Vignette 3: Borderline Presentation Requiring Clinical Judgment
A 3-year-old male presents with 5 days of fever, bilateral conjunctival injection, strawberry tongue, and a faint maculopapular rash. CRP is 4.1 mg/dL, ESR is 35 mm/hr, albumin is 3.4 g/dL, and platelets are 380,000/μL. Echocardiography shows normal coronary arteries (z-scores <2.0).
Analysis: This child has 3 principal clinical features (conjunctival injection, oral changes, rash). Laboratory findings show elevated CRP but only 1 supplemental criterion. The echocardiogram is normal. This is a borderline case. Given the clinical suspicion and elevated inflammatory markers, treatment with IVIG is reasonable. A normal echocardiogram in the first week does not exclude KD, as coronary dilation may develop later. Serial echocardiography at 2 weeks and 6-8 weeks is essential regardless of initial findings.
Vignette 4: Incomplete KD vs. Viral Illness
A 5-year-old female presents with 4 days of fever, mild conjunctival injection, and a nonspecific maculopapular rash. She has rhinorrhea and cough. CRP is 2.1 mg/dL, ESR is 22 mm/hr, and CBC is unremarkable.
Analysis: This child has only 2 principal clinical features with low inflammatory markers and respiratory symptoms suggesting viral illness. If CRP <3.0 mg/dL AND ESR <40 mm/hr, KD is unlikely. However, if fever persists beyond day 5 without alternative explanation, reassessment with repeat laboratory evaluation and echocardiography is warranted. The presence of exudative conjunctivitis or pharyngitis would argue against KD.
Differential Diagnosis
Before initiating treatment, alternative diagnoses should be considered: scarlet fever (sandpaper rash, positive streptococcal testing), measles (cough, coryza, Koplik spots), drug hypersensitivity reactions, viral infections (adenovirus, enterovirus), and multisystem inflammatory syndrome in children (MIS-C).
Clinical findings that are not characteristic of KD include oral ulcerations, exudative pharyngitis, exudative or unilateral conjunctivitis, and vesicular rash.
Distinguishing Kawasaki Disease from MIS-C
Multisystem inflammatory syndrome in children (MIS-C) emerged during the COVID-19 pandemic and shares overlapping features with KD, including fever, mucocutaneous findings, and potential for coronary involvement. However, only 25-50% of patients with MIS-C meet full diagnostic criteria for KD. Distinguishing between these conditions is clinically important, as they may require different management approaches.
Epidemiologic Differences
Age: KD predominantly affects children under 5 years (median age approximately 3 years), while MIS-C has a broader age distribution with children ages 6-12 years at increased risk.
Ethnicity: KD incidence is highest in East Asian populations (particularly Japan). MIS-C has increased incidence in children of African, Afro-Caribbean, and Hispanic descent.
SARS-CoV-2 association: MIS-C requires evidence of current or recent SARS-CoV-2 infection or exposure. KD has no known infectious trigger.
Clinical Differences
Shock and cardiac dysfunction: MIS-C patients present more frequently in a state of shock with ventricular dysfunction and arrhythmias. Fewer than 10% of KD patients present with KD shock syndrome.
Gastrointestinal symptoms: Prominent abdominal pain, vomiting, and diarrhea are more characteristic of MIS-C.
Neurologic symptoms: Headache and lethargy are more frequently encountered in MIS-C.
Mucocutaneous features: Classic KD features (conjunctivitis, oral changes, lymphadenopathy) are less common in MIS-C, though they can occur.
Coronary involvement: CAAs occur in approximately 25% of untreated KD patients but in approximately 13% of MIS-C patients. Importantly, MIS-C patients without KD features can still develop CAAs.
Laboratory Differences
Key laboratory parameters that help distinguish MIS-C from KD:
Platelet count: KD typically causes thrombocytosis (platelets >450,000/μL after day 7), while MIS-C causes thrombocytopenia (median approximately 151,000/μL).
Lymphocyte count: MIS-C causes profound lymphopenia (median approximately 800/μL), while KD typically shows normal or mildly decreased lymphocyte counts (median approximately 2,800/μL).
CRP: Both conditions cause elevated CRP, but MIS-C typically shows markedly higher levels (median approximately 22 mg/dL vs. approximately 6.7 mg/dL in KD).
Cardiac biomarkers: Troponin and NT-proBNP are markedly elevated in MIS-C (reflecting myocardial involvement), while they are typically normal or mildly elevated in KD.
Ferritin and D-dimer: Both are more markedly elevated in MIS-C.
Albumin: Severe hypoalbuminemia is more characteristic of MIS-C (median approximately 2.4 g/dL vs. approximately 3.8 g/dL in KD).
Age-Dependent Phenotype in MIS-C
Epidemiologic studies suggest that younger children with MIS-C are more likely to present with KD-like features, while older children are more likely to develop myocarditis and shock. This age-dependent phenotype may explain some of the clinical overlap between conditions.
Risk Stratification for IVIG Resistance
Approximately 10-15% of patients do not respond to initial IVIG therapy, defined as persistent or recrudescent fever ≥36 hours after completion of the IVIG infusion. These patients are at increased risk for coronary complications.
Kobayashi Score Limitations
The Kobayashi score was developed in Japanese populations to predict IVIG resistance using seven weighted variables (sodium ≤133 mmol/L, illness day ≤4 at treatment, AST ≥100 IU/L, neutrophils ≥80%, CRP ≥10 mg/dL, age ≤12 months, platelets ≤300,000/μL). However, this score performs poorly in non-Japanese populations. A 2023 meta-analysis found the Kobayashi score had a summary C-statistic of only 0.65 in external validation studies, indicating poor discriminative ability. Current AHA guidelines explicitly note that Japanese risk scores "have not performed well in North American cohorts."
The Kobayashi Score Calculator is available on Centilo for reference, but its results should be interpreted with these limitations in mind.
High-Risk Features
Current guidelines identify the following as high-risk features warranting consideration of intensified initial therapy: coronary artery z-score ≥2.5 at diagnosis, age <6 months, and high-risk category using validated risk scores. The decision to intensify therapy requires integration of multiple clinical factors and should involve consultation with a clinician experienced in KD management.
Initial Treatment
IVIG
IVIG (2 g/kg as a single infusion over 8-12 hours) remains the standard of care and should be administered within the first 10 days of illness, ideally by day 5-7 of fever. IVIG can still be administered after day 10 in patients with ongoing fever, elevated inflammatory markers, or coronary abnormalities.
Aspirin
Aspirin is administered during the acute phase, though optimal dosing remains unclear and recent evidence suggests no difference in coronary outcomes between high, moderate, and low-dose aspirin:
High-dose: 80-100 mg/kg/day divided every 6 hours (United States practice). Moderate-dose: 30-50 mg/kg/day (Japan, Western Europe). Low-dose: 3-5 mg/kg/day.
Multiple retrospective studies and meta-analyses have found no significant differences in CAA rates, IVIG resistance, or hospital length of stay between different aspirin doses. The 2024 AHA update notes increasing evidence that medium- or high-dose aspirin in the acute phase is likely not associated with improved coronary outcomes.
After defervescence, transition to low-dose aspirin (3-5 mg/kg/day) for antiplatelet effect, continued for at least 6-8 weeks or until coronary arteries are confirmed normal.
Intensification of Initial Therapy
For patients with high-risk features (z-score ≥2.5 AND age <6 months, or other high-risk criteria), intensification of initial therapy may be considered:
Corticosteroids: The 2021 ACR/VF guideline conditionally recommends adding glucocorticoids to IVIG for patients with both coronary artery z-score ≥2.5 AND age <6 months. Typical dosing is prednisone 2 mg/kg/day (maximum 60 mg/day) tapered over 15 days.
Infliximab: May be considered as an alternative to corticosteroids in high-risk patients, particularly when corticosteroids are contraindicated.
Importantly, a z-score ≥2.5 alone does not automatically mandate intensified therapy. The decision requires integration of multiple risk factors and clinical context.
IVIG-Resistant Disease
Persistent fever ≥36 hours after completion of IVIG infusion defines IVIG resistance. Management options include:
Second IVIG dose (2 g/kg): Conditionally recommended as first-line therapy for IVIG-resistant disease.
Corticosteroids: A reasonable alternative to second IVIG, particularly in patients at risk for hemolytic anemia (non-type O blood groups).
Infliximab: May be considered, though head-to-head comparisons with corticosteroids are lacking.
A meta-analysis of 372 patients with IVIG-resistant disease found no significant differences in coronary outcomes between second IVIG, corticosteroids, or infliximab.
Coronary Artery Assessment and Classification
Coronary artery dimensions should be assessed using z-scores normalized to body surface area. Use the Coronary Z-Score Calculator for objective measurement and the Coronary Classification tool to categorize aneurysm severity per AHA guidelines.
The AHA classification system:
- No involvement: z-score always <2.0
- Dilation only: z-score 2.0 to <2.5, or if initially <2.0, a decrease in z-score during follow-up ≥1.0
- Small aneurysm: z-score ≥2.5 to <5.0
- Medium aneurysm: z-score ≥5.0 to <10.0, or absolute dimension 5-8 mm
- Large/Giant aneurysm: z-score ≥10.0, or absolute dimension ≥8 mm
The threshold of z-score ≥2.5 defines a true aneurysm. Values of 2.0-2.5 represent dilation without aneurysm formation.
Thromboprophylaxis by Coronary Status
No involvement or dilation only (z <2.5): Low-dose aspirin (3-5 mg/kg/day) for 4-6 weeks until normalization documented. No anticoagulation indicated.
Small aneurysm (z 2.5 to <5.0): Low-dose aspirin indefinitely. No anticoagulation indicated. Serial echocardiography every 6-12 months.
Medium aneurysm (z 5.0 to <10.0): Low-dose aspirin. Dual antiplatelet therapy (aspirin + clopidogrel) may be considered. Anticoagulation is not routinely indicated at this level unless there are additional risk factors (rapidly expanding aneurysms, thrombus). Serial echocardiography every 4-6 months.
Large/Giant aneurysm (z ≥10.0 or ≥8 mm): Antiplatelet therapy (aspirin, or dual antiplatelet with clopidogrel). Anticoagulation (warfarin with INR target 2.0-3.0, or LMWH) is reasonable. "Triple therapy" (aspirin + clopidogrel + anticoagulation) may be considered in patients with large/giant aneurysms and recent coronary thrombosis. Serial echocardiography, cardiac catheterization for stenosis/thrombus assessment. Lifelong cardiology follow-up.
Long-Term Prognosis
Recent evidence has substantially refined our understanding of long-term outcomes.
Small and medium CAAs (z <10): Contemporary studies demonstrate near-universal regression and near-zero risk of adverse cardiac events. A large registry study of 1,651 patients showed normalization of coronary diameter in 99% of small CAAs and 92% of medium CAAs over 10 years. No adverse cardiac events occurred in patients with maximal z-score <10. A Japanese cohort study found 10-year coronary event-free survival rates of 100% for small CAAs in both sexes.
Large/Giant CAAs (z ≥10): These patients remain at substantial risk for thrombosis, stenosis, myocardial infarction, and need for coronary intervention. Coronary events occurred in 29-48% of patients with z-score ≥10 in long-term follow-up.
These data suggest that decreased frequency of surveillance may be warranted in patients with small CAAs, while intensive long-term monitoring remains essential for those with large or giant aneurysms.
Echocardiographic Surveillance Schedule
- At diagnosis: Baseline echocardiogram
- Week 2: Repeat echocardiogram
- Week 6-8: Repeat echocardiogram to assess for resolution or progression
For patients with IVIG resistance or coronary changes, more frequent imaging is indicated. Long-term surveillance frequency is determined by maximal coronary involvement.
Key Clinical Priorities
- Diagnose KD (including incomplete presentations) before day 10 of fever when possible
- Recognize that diagnosis can be made with 4 days of fever when ≥4 principal features are present
- Identify high-risk features (z-score ≥2.5, age <6 months) that may warrant intensified therapy
- Initiate IVIG promptly; aspirin dosing in the acute phase does not appear to affect coronary outcomes
- Use coronary z-scores to guide long-term thromboprophylaxis and surveillance
- Recognize that small and medium CAAs have excellent long-term prognosis with near-universal regression
- Ensure lifelong cardiology follow-up for patients with large or giant aneurysms
- Consider MIS-C in the differential diagnosis, particularly in older children with shock, prominent GI symptoms, lymphopenia, and thrombocytopenia
Missing the diagnosis of KD results in significant risk of coronary complications that may manifest as myocardial infarction in adolescence or adulthood.
References
- Jone PN, Tremoulet A, Choueiter N, et al. Update on Diagnosis and Management of Kawasaki Disease: A Scientific Statement From the American Heart Association. Circulation. 2024;150(23):e481-e500.
- McCrindle BW, Rowley AH, Newburger JW, et al. Diagnosis, Treatment, and Long-Term Management of Kawasaki Disease: A Scientific Statement for Health Professionals From the American Heart Association. Circulation. 2017;135(17):e927-e999.
- Kuniyoshi Y, Tsujimoto Y, Banno M, et al. Prediction Models for Intravenous Immunoglobulin Resistance in Kawasaki Disease: A Meta-Analysis. Pediatrics. 2023;151(5):e2022059175.
- Gorelik M, Chung SA, Ardalan K, et al. 2021 American College of Rheumatology/Vasculitis Foundation Guideline for the Management of Kawasaki Disease. Arthritis & Rheumatology. 2022;74(4):586-596.