While reports of the occurrence of KD have come from every continent (not including Antarctica), most epidemiological data comes from Japan and the United States and Canada with increasing reports coming from Taiwan, China and Korea in recent years (Tab. 2).


Since 1970, a total of 17 retrospective incidence surveys have been conducted in Japan (i.e., every 2 years) under the auspices of the Ministry of

Table 1. Clinical and laboratory features of Kawasaki disease

Epidemiological case definition (classic clinical criteria)* Fever persisting at least 5 days Presence of at least 4 principal features:t Changes in extremities

Acute: Erythema of palms, soles; edema of hands, feet Subacute: Periungual peeling of fingers, toes in weeks 2 and 3 Polymorphous exanthem

Bilateral bulbar conjunctival injection without exudates

Changes in lips and oral cavity: Erythema, lips cracking, strawberry tongue, diffuse injection of oral and pharyngeal mucosae Cervical lymphadenopathy (> 1.5 cm diameter), usually unilateral Exclusion of other diseases with similar findings

* Patients with fever at least 5 days and < 4 principal criteria can be diagnosed with Kawasaki disease (KD) when coronary artery abnormalities detected by 2-D echocardiography or angiography are present. t In presence of a 4 principal criteria, KD diagnosis can be made on day 4 of illness. Experienced clinicians who have treated many KD patients may establish diagnosis before day 4.

Table 2. Global distribution of KD beyond Japan and North America


Table 2. Global distribution of KD beyond Japan and North America











Hong Kong










South America


South Africa









Health, Labor and Welfare. Questionnaires were sent to hospitals with pediatric departments and a bed capacity of at least 100, or hospitals with a bed capacity of less than 100 beds but specializing in pediatrics. The survey questions were created by the Japan Kawasaki Disease Research Committee. Response to the surveys has been about 70% [6]. The last reported surveys included the years 1999-2002 [6]. Since the inception of the epidemiological study 186 069 KD patients have been reported.

Figure 1. Evaluation of suspected incomplete Kawasaki disease (KD). In the absence of gold standard for diagnosis, this algorithm cannot be evidence-based, but rather represents the informed opinion of the expert committee. Consultation with an expert should be sought anytime assistance is needed. (1) Infants s 6 months old on day a 7 of fever without other explanation should undergo laboratory testing and, if evidence of systemic inflammation is found, an echocardiogram, even if the infants have no clinical criteria. (2) Patient characteristics suggesting KD are listed in Table 1. Characteristics suggesting diseases other than KD include exudative conjunctivitis, exudative pharyngitis, discrete intraoral lesions, bullous or vesicular rash, or generalized adenopathy. Consider alternative diagnoses. (3) Supplemental laboratory criteria include albumin s 3.0 g/100 ml, anemia for age, elevation of alanine aminotransferase, platelets after 7 days a 450 000/mm3, white blood cell count a 15 000/mm3, and urine a 10 white blood cells/high-power field. (4) Can treat before performing echocardiogram. (5) Typical peeling begins under nail bed of fingers and then toes. (6) Echocardiogram is considered positive for purposes of this algorithm if any of three conditions are met: z score of LAD or RCA a 2.5, coronary arteries meet Japanese Ministry of Health criteria for aneurysms, or a 3 other suggestive features exist, including perivascular brightness, lack of tapering, decreased LV function, mitral regurgitation, pericardial effusion, or z scores in LAD or RCA of 2-2.5. (7) If the echocardiogram is positive, treatment should be given to children within 10 d of fever onset and those beyond day 10 with clinical and laboratory signs (CRP, ESR) of ongoing inflammation. Taken from [5]. Copyright ©2004 American Heart Association.

During the most recent study period, 32 266 patients were reported with an annual incidence of 137.7 per 100 000 children < 5 years old in 1999 and 151.2 per 100 000 in 2002. The male to female ratio was 1.30 [6]. The annual incidence of KD in Japan has increased progressively from 1987 to 2002 from 73.8 to 151.2 per 100 000 < 5 years of age [6-10].

Over the 32-year period of surveillance, three nationwide epidemics of KD have been observed, in 1979, 1982 and 1986 [11]. The incidence rate in the last epidemic in 1986 was 176.8 per 100 000 children < 5 years of age. No national epidemic outbreaks have been reported since 1987 but regional outbreaks continue to occur.

United States

Surveillance of KD in the United States is through passive reporting of cases to the Centers for Disease Control and Prevention, where a database has been maintained since 1984. Unfortunately, only a fraction of cases are identified through this system. More robust estimates of the incidence of KD has come from reports from regional investigators [12-14], surveys conducted by specialty societies and more recently through the use of administrative databases [15] of childhood hospitalizations.

Taubert [16] conducted surveys of 440 general hospitals with at least 400 beds that included a pediatric section and of 63 children's hospitals. The survey periods covered the years 1984-1987, 1988-1990 and 1991-1993. During the latter period only the children's hospitals were surveyed. The surveys yielded rates of 7.6 cases per 100 000 children < 5 years of age and 9.2 cases per 100 000 for the first two periods. Following the third survey, the reported cases throughout the 10-year period were totaled and a minimum estimate of the annual rate for the period was calculated which was 8.9 cases per 100 000 children < 5 years of age. This rate was similar to rates found previously reported in regional studies [13, 14].

More recently, investigators from the Centers for Disease Control and Prevention utilized data from a large inpatient database to determine incidence rates in the United States. The database was designed to generate robust national estimates of pediatric hospitalizations [15]. The rates were determined for the years 1997 and 2000, which were 17.6 per 100 000 < 5 years of age and 17.1 per 100 000 < 5 years, respectively. These rates were comparable to those determined from the regional studies using State health or health maintenance organization data [18-20]. Based on data from this study and others published previously during the decade, the authors concluded that incidence rate for KD in the United States had been stable.

A national epidemic of KD involving ten regions in the United States occurred between August 1984 and January 1985 [21]. Several other regional outbreaks have been reported over the years as well [16]. Similar incidence rates were reported from Canada based on a national health statistic data base. From 1990-1991 to 1995-1996 the mean rates for children < 5 years across Canada were 13.8 per 100 000 children [22].

Global distribution

KD has been reported from every continent and several island groups across the globe (Tab. 2). While the incidence rates in Japan remain the highest in the world, several other Asian nations have posted high rates as well. Several reports from China (Beijing [23], Hong Kong [24], Shanghai [25]), Taiwan [26] and Korea [27] documented rates intermediate between those of Japan and North America (Tab. 3). All but Taiwan appeared to have increasing rates over the study periods.

Case reports or case series have been reported from many countries in Europe, Oceania, Africa and South Africa as well as other Asian countries (Tab. 2). Of those countries reporting incidence rates, only those from Ireland are comparable to those in North America [28].

To summarize the global experience with KD, the highest incidence rates are found in Japan followed by Korea, China, North America and Europe. Local or regional outbreaks have been documented in both Japan and the United States, and national epidemics have been observed in both countries as well as Finland [29]. Incidence rates have trended upward in several countries and have remained stable in others. The effect of ascertainment bias on apparent increases in incidence is not known.


As suggest by higher incidence rates in Asian countries, KD occurs in higher frequency in Asian populations. Numerous studies from the Unites States have shown KD to be over-represented among Asian children [12, 14-16, 18, 19]. An interesting study of the epidemiology of KD in Hawaii dramatically demonstrated this predominance [30]. A retrospective analysis of the State Inpatient Database for Hawaii was performed for KD patients hospitalized during 1996 through 2001. Race classification provided by Census 2000 indicated race listed alone or in combination with other races. This race-specific numerators and denominators could be determined. The average annual incidence for KD was 45.2 per 100 000 children < 5 years of age, highest in the United States. Japanese-American children < 5 years had the highest incidence (197.7 per 100 000) followed by Native Hawaiian (99.1), Chinese (81.3) and Filipino (64.8). Caucasian children < 5 years old had a rate of 35.3 per 100 000 children. These findings suggest there may be true differences in incidence of KD among Asian populations. Since the populations in Hawaii came from relatively similar social and physical environments and have similar access to healthcare and diagnostic practices,

Epidemiology and etiology of Kawasaki disease Table 3. Incidence rate of KD in different areas

Incidence rate

Epidemiology and etiology of Kawasaki disease Table 3. Incidence rate of KD in different areas

Incidence rate


Surveillance period



Beijing [23]




Taiwan [26]




Hong Kong [24]




Shanghai [25]




Korea [27]




the socioeconomic and environmental factors would not introduce bias to the ascertainment. Incidence rates among racial groups in the United States (2000) were Caucasian (non-Hispanic) 11.4 per 100 000 children < 5 years, African American (non-Hispanic) 19.7 per 100 000 children, Hispanic 13.6 per 100 000 and Asian/Pacific Islander 39.0 per 100 000 [15]. In a separate study of American Indian/native Alaskan children the rate was 4.2 per 100 000 [31].

Age and gender

Most series from diverse geographic and racial populations have shown approximately 85% of children with KD are < 5 years of age [2]. Thus, incidences are expressed generally as a proportion of children < 5 years old. The most recent population-based study in the United States indicated 76% and 77% of patients were < 5 years of age in 1997 and 2000. The median age of KD patients in the United States is 2 years. In Japan, the peak age is 9-11 months and 88.9% of KD patients were < 5 years of age [6].

KD is relatively uncommon in children < 6 months old and above 5 years of age [6, 15, 17-19]. Studies have suggested that CAA are more common in these two age groups possibly because the illness is less typical and thus diagnosis is delayed [6, 32-36]. While KD is overwhelmingly a disease of children, rare cases have been reported in adults [37].

KD occurs in males more frequently than females [5-16]. Males are at greater risk of developing CAA as well [33]. In the United States the male: female ratio is 1.5:1 while in Japan it is 1.3:1.

Seasonality and temporal-geographic clustering of KD

In the United States, KD hospitalizations are more frequent in the winter months [16, 38]. The seasonality in Hawaii is less obvious, although fewer cases were seen during April through June [30]. In a 5-year period of active surveillance in San Diego County, California, KD incidence was inversely associated with average monthly temperature and positively associated with average monthly precipitation [39].

In Japan, excluding epidemic years, there appears to be a bimodal seasonal occurrence with peaks in January and early summer and a nadir in October [6, 40]. In China and Korea seasonal peaks appear to be more frequent in spring and summer [23, 26, 29].

Temporal-geographic clustering has been frequently observed in both the United States [15, 32, 38, 41] and Japan [6, 40]. The Japanese experience has been especially well described with "hot spots" occurring in various prefectures on a rotating basis [6, 40].

Familial cases

There appears to be an observable enhanced risk of KD within certain families. In Japan, there is a tenfold increased risk of the illness in siblings of an index case [42, 43]. Parents of children with KD are twice as likely to have had the disease as compared to the general population [42]. In families where parents had KD, sibling cases among children are significantly more common [44]. Similar findings have been reported from the United States [45].

Recurrence of KD

Recurrent cases of KD have been reported in both the United States and Japan [42, 46]. The estimated rate of recurrence in Japan is 3%, while that in the United States is < 1 to slightly over 1% [46, 47].

Socioeconomic factors

KD patients in the United States come from families with a higher median household income and are more likely to have private insurance [15, 39]. An analysis of hospitalization costs for KD in the United States for children

< 5 years of age showed that the median cost was $6189 [48]. The average annual total estimated cost associated with hospitalization for KD patients

Other risk factors

Several other risk factors for KD have been reported in the past. An ante cedent respiratory illness has been a significant association with KD patients as compared to controls in outbreak situations in the United States [32].

Shampooing or spot-cleaning carpets within 30-45 days of onset of KD has been a risk factor in some studies but not others [49, 50]. Other factors associated with KD include use of a humidifier [49], living near a body of water [38] and having preexisting eczema [51].

Synthesis of epidemiological data

KD is an acute self-limited illness of children that is characterized epide-miologically by seasonality and occurring in geographic clusters. It shares many clinical characteristics with known infectious diseases such as scarlet fever, toxic shock syndrome, measles and adenovirus infections. It has been associated with antecedent viral-like illnesses in some epidemic situations. All of these factors suggest an infectious etiological agent or agents as a cause. The relative rarity in the first 3 months of life (possibly due to maternal antibody) and peak occurrence early in childhood is another characteristic shared by many common childhood infections.

Host factors also appear to be important in the disease. Susceptibility to the disease is clearly influenced by ethnicity, familial risk factors and possibly preexisting conditions such as atopy. A genetic predisposition is suggested by these factors.

Finally, environmental influences cannot be ruled out as suggested by apparently recent emergence of the disease in the last half of the 20th century, a socioeconomic bias toward more affluent lifestyle, possibly climatic associations and less well established associations such as rug cleaning.

The exposure of a predisposed host to an infectious pathogen or pathogens with possible environmental contributing factors is a reasonable model to propose for KD.

How To Deal With Rosacea and Eczema

How To Deal With Rosacea and Eczema

Rosacea and Eczema are two skin conditions that are fairly commonly found throughout the world. Each of them is characterized by different features, and can be both discomfiting as well as result in undesirable appearance features. In a nutshell, theyre problems that many would want to deal with.

Get My Free Ebook

Post a comment