Am Fam Physician. 2000;62(9):2115-2121
Kawasaki disease is a systemic vasculitis that occurs in children. Although it has been recognized as a clinical entity for about 30 years, its etiology remains unknown, and there are no confirmatory laboratory or other diagnostic tests. Clinical criteria include fever for at least five days; bilateral conjunctival injection; mucosal changes in the oropharynx; changes in the peripheral extremities, such as edema or erythema; a polymorphous rash; and cervical lymphadenopathy of greater than 1.5 cm. A second condition, known as atypical Kawasaki disease, includes patients who do not meet the true case definition of Kawasaki disease but have comparable laboratory findings and no other explanation for the illness. A condition that is often misdiagnosed as Kawasaki disease and poses the greatest diagnostic dilemma is acute adenoviral infection. Barone and colleagues retrospectively compared clinical and laboratory features of the three conditions, particularly emphasizing the use of rapid direct fluorescent antigen (DFA) testing for acute adenoviral infection.
Medical records of all patients diagnosed with Kawasaki disease at a university hospital over a two and one-half-year period were analyzed. A subset of children who did not meet standard diagnostic criteria was classified as having atypical Kawasaki disease. All children with either diagnosis were treated with intravenous gamma globulin (IVIG) and aspirin. Acute adenoviral infection was diagnosed if a positive culture for adenovirus was identified during medical evaluation. A variety of demographic indicators, clinical signs and symptoms, and laboratory data from the children's medical records was analyzed as well.
A total of 43 children met the inclusion criteria; of these, 36 were diagnosed with Kawasaki disease and treated. Thirteen of the 36 had atypical Kawasaki disease; seven others were found to have acute adenoviral infection. Mean patient age, duration of fever, number of physician visits before diagnosis and reports of exposure to sick contacts were similar across groups. For a comparison of clinical findings across groups, see the accompanying table. Conjunctivitis and changes in peripheral extremities were more common in children with either type of Kawasaki disease than in those who had adenoviral infection. Although a number of children in the latter group had conjunctivitis, it was usually purulent rather than exudative, as was characteristic of the children with Kawasaki disease. Clinical findings such as the presence or absence of rash, cervical adenopathy or mucous membrane changes did not differ across groups; however, the rash associated with Kawasaki disease tended to be more accentuated in the perineal region. The only notable difference between children with atypical Kawasaki disease and those with adenoviral infection was that a greater number of children in the former group had conjunctival involvement.
Laboratory testing revealed that children with Kawasaki disease had higher mean white blood cell counts, platelet counts, erythrocyte sedimentation rates and alanine aminotransferase levels compared with children who had adenoviral infection. Pyuria (defined as more than 10 white blood cells per high-power field) was identified in 17 of the children with Kawasaki disease but in none of the children with adenoviral infection. Echocardiography performed on all of the children with Kawasaki disease revealed that none had a coronary aneurysm; whereas only one child with adenoviral infection had an echocardiogram, which was also was normal.
Rapid DFA testing for adenovirus was done in 18 children with Kawasaki disease, including nine of the 13 children with atypical disease. In all 18 children, the results were negative. In this study, the rapid DFA test had a sensitivity of 86 percent and a specificity of 100 percent. None of the children with a DFA test positive for adenoviral infection was inappropriately given IVIG.
The authors conclude that acute adenoviral infection and Kawasaki disease share many clinical manifestations, making the diagnosis difficult. They recommend the use of rapid DFA testing as a way to help distinguish the two conditions. Results are returned within six hours and can be confirmed with a culture. Positive results may prevent the unnecessary use of IVIG in some children and the significant expense incurred with this form of therapy.