Intended for healthcare professionals


Kawasaki disease

BMJ 1997; 315 doi: (Published 09 August 1997) Cite this as: BMJ 1997;315:322

Early recognition is vital to prevent cardiac complications

  1. Nigel Curtis, MRC clinician scientist (n.curtis{at}
  1. a Paediatric Infectious Diseases Unit, Department of Paediatrics, Imperial College School of Medicine at St Mary's, London W2 1PG

    Dr Tomisaku Kawasaki described the disease that bears his name nearly 30 years ago.1 Kawasaki disease is now the commonest cause of acquired heart disease in children in developed countries. Its cause remains unknown, and it presents doctors with many difficulties in diagnosis and management. Kawasaki disease is a systemic febrile vasculitis predominantly affecting children aged under 5 years. The incidence in Britain is 3.4 per 100 000 children aged under 5 years2—about a third of the incidence reported in the United States and a 30th of that in Japan.3 The most important complication, coronary arteritis leading to formation of aneurysms, occurs in 20-30% of untreated patients.3 Thrombosis within an aneurysm, myocardial infarction, and dysrhythmias may occur in the acute phase of the illness. The case fatality rate in Britain in 1990 was 3.7%,2 which compares unfavourably with the United States and Japan, where in some centres it is as low as 0.1%.2 Patients also suffer long term morbidity as a result of scarring of coronary arteries, intimal thickening, and accelerated atherosclerosis.3

    Diagnostic criteria for Kawasaki disease

    Presence of at least five of six conditions:

    • Fever for five days or more

    • Bilateral (non-purulent) conjunctivitis

    • Polymorphous rash

    • Changes in lips and mouth:

    Reddened, dry, or cracked lips

    Strawberry tongue

    Diffuse redness of oral or pharyngeal mucosa

    • Changes in extremities:

    Reddening of palms or soles

    Indurative oedema of hands or feet

    Desquamation of skin of hands, feet, and groin (in convalescence)

    • Cervical lymphadenopathy:

    More than 15 mm in diameter, usually unilateral, single, non-purulent, and painful

    Exclusion of diseases with similar presentation:

    • Staphylococcal infection (such as scalded skin syndrome, toxic shock syndrome)

    • Streptococcal infection (such as scarlet fever, toxic shock-like syndrome). Throat carriage of group A streptococcus does not exclude the possibility of Kawasaki disease

    • Measles and other viral exanthems

    • Leptospirosis

    • Rickettsial disease

    • Stevens-Johnson syndrome

    • Drug reaction

    • Juvenile rheumatoid arthritis


    There is no diagnostic test for the disease, and many cases are missed. Six deaths were recorded in Britain in 1990, but only one case was diagnosed during life.2 The diagnosis is based on fulfilling clinical criteria (see box).4 However, many common childhood infections have similar clinical features. Furthermore, the diagnostic features of Kawasaki disease may appear sequentially rather than simultaneously. The two features that doctors most often remember are desquamation of the rash and thrombocytosis. Unfortunately, these features are the least useful in reaching an early diagnosis because they usually occur later in the disease. Moreover, the clinical diagnostic criteria do not identify every case; “incomplete” or “atypical” cases have come to light because coronary artery aneurysms have been found on echocardiography or at necropsy.5

    How can a doctor distinguish Kawasaki disease from more common causes of fever with a rash? Remaining alert to the possibility of the diagnosis is critical. Many “textbook” cases of the disease are missed simply because the diagnosis is not considered. Kawasaki disease can be mistaken for measles, but measles has become less common in Britain since the recent vaccination campaign. Kawasaki disease is a systemic disease and typically affects many systems; doctors should not be diverted from the diagnosis because of less characteristic features such as rhinorrhoea, cough, abdominal pain, vomiting, diarrhoea, pain and swelling of joints, involvement of the central nervous system, abnormal liver function tests, and sterile pyuria, which can all occur in Kawasaki disease.

    There are some useful clues to the diagnosis. Most of the diseases for which Kawasaki disease is mistaken do not cause fever for more than five days. In addition, the fever in Kawasaki disease is often unresponsive to antipyretics, and the child is inconsolably miserable (1). An additional sign sometimes seen in the acute phase is redness and induration at the site of a BCG scar. Kawasaki disease is an inflammatory process, and an acute phase response is characteristic—its absence suggests an alternative diagnosis. Presumptive treatment should be started in any child with a persistent fever, some of the clinical features of Kawasaki disease, and an acute phase response even if symptoms do not meet the full diagnostic criteria.

    Fig 1
    Fig 1

    The child with Kawasaki disease is characteristically inconsolable (Reproduced with parents' consent)

    Intravenous immunoglobulin is the most effective treatment (see 1). Given within 10 days of onset, this significantly decreases both the incidence and severity of aneurysm formation,6 7 as well as providing dramatic resolution of inflammation and relief of symptoms. A single high dose of 2 g/kg is more effective than the previously recommended regimen of 400 mg/kg for four days.7 However, despite its proved efficacy, many patients receive delayed or inadequate amounts of immunoglobulin, or even none at all.8 In Britain in 1990 only 7% of patients were given the recommended optimal treatment, and 39% were not given any immunoglobulin at all.3

    Table 1

    Treatment of Kawasaki disease

    View this table:

    There is often uncertainty about the use of immunoglobulin in patients diagnosed more than 10 days after the start of the disease. This has not been tested in a prospective controlled trial. However, many paediatricians recommend the use of immunoglobulin in such patients if there is evidence of ongoing inflammation, as suggested by continuing fever, malaise, and raised acute phase reactants. Retreatment with immunoglobulin is recommended for persistent or recrudescent disease. Aspirin remains an integral part of treatment, although its use in Kawasaki disease has not been subjected to prospective controlled trials.9

    Many of the difficulties in diagnosing and managing patients with Kawasaki disease would be solved if the cause of the disease was known. The epidemiology strongly suggests an infectious aetiology.10 At the recent international symposium on Kawasaki disease there was much interest in the hypothesis that the disease is caused by a bacterial superantigen toxin, similar to those responsible for the staphylococcal and streptococcal toxic shock syndromes.11 The recent report of a patient who fulfilled the clinical criteria for staphylococcal toxic shock syndrome but who also had coronary artery lesions typical of Kawasaki disease12 supports the suggestion that Kawasaki disease and toxic shock syndrome share a common aetiology. However, there is still conflicting evidence for the superantigen theory. Future advances in diagnosis and treatment are likely to depend on the definitive identification of the cause of the disease.



    1. 1.
    2. 2.
    3. 3.
    4. 4.
    5. 5.
    6. 6.
    7. 7.
    8. 8.
    9. 9.
    10. 10.
    11. 11.
    12. 12.
    View Abstract