Original Article
Mortality among Children with Kawasaki Disease in Japan
N Engl J Med 1992; 326:1246-1249May 7, 1992
- Abstract
- Abstract
Background and Methods.
It is not certain whether patients with Kawasaki disease have a higher death rate than the age-matched healthy population. We therefore undertook a study to investigate this question. Between July 1982 and December 1988, 53 collaborating treatment centers collected data on all patients who had an unequivocal new diagnosis of Kawasaki disease; patients who had recurrent disease or whose first visit to the treatment center occurred more than 14 days after the onset of symptoms were excluded. Patients were followed from the time of the first visit to the treatment center until December 31, 1989, or until death, whichever occurred first. The expected number of deaths was calculated from Japanese vital-statistics data and compared with the number observed.
Results.
Of 4676 patients who met the eligibility criteria, 4608 (98.5 percent) were followed through either the end of the study or the date of death. Thirteen patients (10 boys and 3 girls) died during the study period. The number of deaths expected was 7.61 (ratio of observed to expected deaths, 1.71; 95 percent confidence interval, 0.91 to 2.92). The ratio was 2.04 (95 percent confidence interval, 0.98 to 3.76) for boys and 1.11 (95 percent confidence interval, 0.23 to 3.23) for girls. During the acute phase of the disease (the first two months after onset), the ratios of observed to expected deaths were higher: 9.86 overall (95 percent confidence interval, 3.95 to 20.31), 13.33 for boys (95 percent confidence interval, 4.89 to 29.07), and 3.85 for girls (95 percent confidence interval, 0.10 to 21.42). After the acute phase, however, both sexes had ratios of observed to expected deaths that were lower than 1, and the difference from the control population was not statistically significant.
Conclusions.
The mortality rate among boys with Kawasaki disease in Japan is twice that among healthy boys of the same age, and most deaths occur within two months of diagnosis. The mortality rate among girls with the disease appears similar to that among healthy girls, although the numbers are very small. (N Engl J Med 1992;326: 1246–9.)
Media in This Article
Table 1
Sex and Age Distribution of the 4676 Study Patients.Table 2
Observed and Expected Deaths According to Disease Phase and Sex.*Article Activity
- Article
SINCE the first description of Kawasaki disease by one of us in 1967,1 case reports from more than 50 countries have appeared. This febrile disease, the cause of which is unknown, primarily affects infants and toddlers, in whom it causes widespread vasculitis. Cardiac sequelae, notably coronary aneurysms and carditis, occur in 10 to 20 percent of patients.2 During a two-year period (1985 through 1986), 28 deaths from Kawasaki disease (a case fatality rate of 0.1 percent) were reported in Japan.3
Many investigators4 5 6 7 8 9 10 11 12 13 14 have conducted long-term follow-up studies that have focused on survival and cardiac sequelae among patients with Kawasaki disease. These studies, however, included only patients with cardiac lesions, and follow-up was conducted only to assess the efficacy of treatment. Although such studies yield valuable information about the outcomes of severely affected patients, the results may be biased with regard to the prognosis for the total group of patients with Kawasaki disease.
To clarify the overall prognosis, we conducted a large multicenter study in which virtually all patients who received a diagnosis of Kawasaki disease were followed, including those who did not return to the treatment center. Our goal was to determine whether patients with Kawasaki disease were more likely to die than age-matched healthy children.
Methods
Entrance Criteria
To date, 11 nationwide surveys of Kawasaki disease have been conducted by the Japanese Kawasaki Disease Research Committee. All patients in whom Kawasaki disease was diagnosed during a 2 1/2 year period from July 1982 through December 1984 were included in the eighth survey.15 Patients included in the 9th survey were given the diagnosis during a two-year period from January 1985 through December 1986,3 and those included in the 10th survey during a two-year period from January.1987 through December 1988.2
In 1990, a collaborative research group was formed to conduct a follow-up study of patients with Kawasaki disease. The research group included members of the Japanese Kawasaki Disease Research Committee and pediatricians from 53 treatment centers (see the Appendix). All patients in whom Kawasaki disease was diagnosed who were seen at the 53 treatment centers were included in the 8th, 9th, and 10th nationwide surveys and served as the patient population for this study. We included only patients who met the following criteria: the diagnosis of Kawasaki disease had to be unequivocal according to diagnostic guidelines developed by the committee,15 the disease had to be in its initial episode, and the patient had to have presented at the treatment center less than 15 days after the onset of symptoms. Thus, patients were excluded from the study if the diagnosis was only probable, if they had recurrent disease, or if they were first seen after the 14th day of symptoms. These limitations were imposed to avoid bias due to the large numbers of patients with late cardiac sequelae seen at large referral hospitals.
Protocol
The patients were followed from the time they first came to the treatment center until December 31, 1989, or the time of death, if it occurred before that date. The status of the patients on January 1, 1990, was confirmed by review of records of patient visits to the treatment center after January 1, 1990; by requests for information from the patients' parents if there was no visit between January and September 1990; and by checks of resident registration records in municipal offices when no reply was received. Deaths were confirmed by letters from the parents or by residential registration records. Copies of death certificates were obtained to determine the causes of death.
Statistical Analysis
The duration of observation was calculated according to sex, age, and calendar year for each patient. Age was classified according to month during the first year of life, according to year up to 5 years of age, and then by 5-year intervals up to 19 years of age. The expected number of deaths was calculated by multiplying the observation time for each patient by the death rate calculated from Japanese vital-statistics data for each group defined by sex, age, and calendar year. The sum of the expected deaths in each category constituted the total expected number of deaths. The observed number of deaths was then compared with the expected number to determine the ratio of observed to expected deaths. We calculated 95 percent confidence intervals for this ratio with use of a table of confidence limits for the Poisson parameter.16 A 95 percent confidence interval that did not include 1.0 was considered to indicate statistical significance.
Patients who were lost to follow-up before December 31, 1989, were observed from the time of the first visit to the last visit. Data obtained within two months of the first visit were compared with those obtained more than two months after the first visit in order to evaluate the prognosis for patients in the acute phase of the disease and for those in the later phase.
Results
Between July 1982 and December 1988, 4993 patients with Kawasaki disease of new onset were seen at the 53 treatment centers. Of these patients, 4676 (93.7 percent) were seen within 15 days of the onset of symptoms and were enrolled in the study.
The sex and age distributions of the patients are shown in Table 1Table 1
Sex and Age Distribution of the 4676 Study Patients.. The male-to-female ratio of the patients was 1.3; 92.0 percent of the patients were less than five years of age. The sex and age distributions of the study patients were comparable to those reported in other epidemiologic studies of Kawasaki disease in Japan.2 Of the 4676 patients, 794 (17.0 percent) had evidence of cardiac sequelae; this prevalence is similar to that among all patients reported in the nationwide surveys during the same period (16.3 percent).2 Thirteen patients (0.3 percent) had died by December 31, 1989. Survival after January 1, 1990, was confirmed in 4595 cases (98.3 percent) as follows: 2343 (51.0 percent) by patients' visits to the treatment center, 1265 (27.5 percent) by responses to our letters to parents, and 987 (21.5 percent) by reviews of municipal records. The total follow-up rate was 98.5 percent.
A total of 68 patients (1.5 percent) were lost to follow-up, mainly as a result of their moving to other municipalities. Three of the patients were not Japanese and may have returned to their home countries. Since foreign visitors are not included in the Japanese resident-registration system, we could not confirm their status if their parents did not reply to the mail survey.
We accounted for 19,535 person-years of follow-up (11,144 person-years for boys and 8391 person-years for girls); the average period of observation was 4.2 years.
As shown in Table 2Table 2
Observed and Expected Deaths According to Disease Phase and Sex.*, the number of deaths observed among the boys was higher than the expected number throughout the observation period. The overall ratio of observed to expected deaths for both sexes was elevated (1.71; 95 percent confidence interval, 0.91 to 2.92), as was that for the boys (2.04; 95 percent confidence interval, 0.98 to 3.76), although neither ratio was statistically significant. The ratio for the girls was not increased (1.11).Although we found elevated overall ratios of observed to expected deaths during the acute phase of Kawasaki disease, the ratio was statistically significant only for the boys (13.33; 95 percent confidence interval, 4.89 to 29.07) (Table 2). After the acute phase, the overall ratio of observed to expected deaths was not statistically significant; the same was true for the ratio for each sex.
Data were collected on the 13 patients who died (Table 3Table 3
Characteristics of Patients Who Died.). Except for one, who died as a result of an injury, all the patients who died during the acute phase died of Kawasaki disease: five as a result of cardiac disease, and one of encephalopathy. Six patients died after the acute phase: two of congenital heart disease and one of coronary artery disease secondary to Kawasaki disease.Discussion
Although more than 20 years have passed since Kawasaki disease was first reported, there has been no long-term follow-up study of a large sample of patients with this disorder. Many deaths from Kawasaki disease have been reported,3 , 4 , 15 , 17 18 19 including those of patients who lived for many years, but the degree to which Kawasaki disease contributes to excess mortality among children in Japan has not been documented.
Although in our study population the probability of death within two months after the onset of symptoms was high for both sexes, the probability for girls was similar to that expected, but the probability for boys was twice as high as expected. The difference between the sexes was probably a result of the larger proportion of boys who had cardiac lesions.20 This observation agrees with the findings of the nationwide surveys of Kawasaki disease in Japan, in which the incidence of cardiac sequelae caused by the disease was higher among boys than among girls.2 , 15
Akagi et al.21 reported that about 1 percent of their patients with Kawasaki disease had valvular disease; 12 of the 13 with valvular lesions had mitral regurgitation, which suggests that mitral insufficiency is the most common valvular lesion caused by Kawasaki disease. The current study supports these results; two patients (Patients 3 and 5 in Table 3) died of mitral insufficiency during the acute phase of the disease.
After the acute phase, the ratios of observed to expected deaths were approximately 1.0 in both sexes. Two of six patients died of congenital cardiac anomalies during this period. This finding suggests that Kawasaki disease may increase the risk of death from congenital heart disease.
The collaborative study group comprised pediatric cardiologists and pediatricians who were especially interested in Kawasaki disease: thus, all the patients participating in the study received care from specialists, even during the early stages of the disease. Because the current data, which indicate that the excess mortality is limited to the acute stage of the illness, reflect care by these specialists, the possibility exists that the mortality rate may be higher for patients who are not treated by physicians familiar with Kawasaki disease.
Furthermore, although the date of onset of the disease was known for each patient, the data we analyzed were collected beginning with the patient's first visit to the treatment center. Because patients would not have been included in this study if they had died before diagnosis or before their first visit to one of the treatment centers, observations starting with presentation at a center might generate biased results. The results of this study may thus provide a conservative estimate of the mortality associated with this mysterious disease.
Conducted as a research project by the Japanese Kawasaki Disease Research Committee, of which Dr. Kawasaki is the chairperson, and supported by the Ministry of Health and Welfare of the Japanese Government.
Treatment centers participating in this study are listed in the Appendix.
We are indebted to all the members of the collaborative research group for conducting follow-up of the patients with Kawasaki disease and for providing us with data; to all the patients and their parents and guardians for responding to the mail surveys; and to the Japanese Association of Parents of Children with Kawasaki Disease (chairperson, Mr. Mitsuru Asai) for coordinating the study.
Source Information
From Department of Public Health, Jichi Medical School, Tochigi, Japan (Y.N., H.Y.), and the Kawasaki Disease Research Information Center, Tokyo (T.K.). Address reprint requests to Dr. Nakamura at the Department of Public Health, Jichi Medical School, Yakushiji 3311–1, Minamikawachi-machi, Tochigi-ken, 329–04 Japan.
Appendix
The following treatment centers in Japan participated in the collaborative follow-up study of Kawasaki disease: Department of Pediatrics, Hokkaido University School of Medicine, Sapporo; Department of Pediatrics, Hakodate National Hospital, Hakodate, Hokkaido; Department of Pediatrics, Hakodate Chuo Hospital, Hakodate, Hokkaido; Department of Pediatrics, Nikko Memorial Hospital, Muroran, Hokkaido; Department of Pediatrics, Asahikawa City Hospital, Asahikawa, Hokkaido; Department of Pediatrics, Date Red Cross Hospital, Date, Hokkaido; Department of Pediatrics, Yamagata University School of Medicine, Yamagata; Department of Pediatrics, Yamagata City Hospital, Yamagata; Department of Pediatrics, Tsuchiura Kyodo Hospital, Tsuchiura, Ibaragi; Department of Pediatrics, Jichi Medical School, Minamikawachi, Tochigi; Department of Pediatrics, Tokyo University Hospital, Tokyo; Department of Pediatrics, Tokyo Medical and Dental University, Tokyo; Department of Pediatrics, Tokyo Metropolitan Bokutoh General Hospital, Tokyo; Department of Pediatrics, Japanese Red Cross Medical Center, Tokyo; Department of Pediatrics, Jikei University School of Medicine, Tokyo; Department of Pediatrics, Tokyo Women's Medical College Daini Hospital, Tokyo; Department of Pediatrics, Nihon University School of Medicine, Tokyo; Department of Pediatrics, St. Marianna University School of Medicine, Kawasaki; Department of Pediatrics, St. Marianna University School of Medicine Toyoko Hospital, Kawasaki; Department of Pediatric Cardiology, Shizuoka Children's Hospital, Shizuoka; Department of Pediatrics, Nagoya University School of Medicine, Nagoya; Department of Pediatrics, Kyoto Second Red Cross Hospital, Kyoto; Department of Pediatrics, National Cardiovascular Center, Suita, Osaka; Department of Pediatric Cardiology, Kinki University School of Medicine, Osaka-Sayama, Osaka; Department of Pediatrics, Mimihara General Hospital, Sakai, Osaka; Department of Pediatrics, Osaka City Momoyama Hospital, Osaka; Department of Pediatrics, Nishiyodo Hospital, Osaka; Department of Pediatrics, Otemae Hospital, Osaka; Nakano Children's Hospital, Osaka; Department of Pediatrics, Minoh City Hospital, Minoh, Osaka; Department of Pediatrics, Higashi-Osaka Seikyo Hospital, Higashi-Osaka, Osaka; Department of Pediatric Cardiology, Kobe General Hospital, Kobe; Department of Pediatrics, Meiwa Hospital, Nishinomiya, Hyogo; Department of Pediatrics, Hyogo Prefectural Nishinomiya Hospital, Nishinomiya, Hyogo; Department of Pediatrics, Wakayama Medical School, Wakayama; Department of Pediatrics, Kokuho Hidaka General Hospital, Gobo, Wakayama; Division of Pediatrics, Heart Institute, Kurashiki Central Hospital, Kurashiki, Okayama; Department of Pediatrics, Hiroshima City Hospital, Hiroshima; Department of Pediatrics, Hiroshima Red Cross Hospital, Hiroshima; Department of Pediatrics, Mazda Hospital, Mazda Motor Corporation, Hiroshima; Department of Pediatrics, Asa City Hospital, Hiroshima; Department of Pediatrics, Onomichi General Hospital, Onomichi, Hiroshima; Department of Pediatrics, National Fukuyama Hospital, Fukuyama, Hiroshima; Department of Pediatrics, Mitsubishi Mihara Hospital, Mihara, Hiroshima; Department of Pediatrics, Chugoku Rousai Hospital, Kure, Hiroshima; Department of Pediatrics, Matsuyama Red Cross Hospital, Matsuyama; Department of Pediatrics, Saiseikai Imabari Hospital, Imabari, Ehime; Department of Pediatrics and Child Health, Kurume University School of Medicine, Kurume, Fukuoka; Department of Pediatrics, Kokura Memorial Hospital, Kitakyushu; Department of Pediatrics, Kyushu Kosei-Nenkin Hospital, Kitakyushu; Department of Pediatrics, St. Mary's Hospital, Kurume, Fukuoka; Department of Pediatrics, Omuta City Hospital, Omuta, Fukuoka; and Department of Pediatrics, Miyazaki Prefectural Miyazaki Hospital, Miyazaki.
- References
References
1
Kawasaki
Medline2
Nakamura
Web of Science | Medline3
Yanagawa
CrossRef | Web of Science | Medline4
Kato
CrossRef | Web of Science | Medline5
Suzuki 6
Anderson
CrossRef | Web of Science | Medline7
Yanagisawa
CrossRef | Web of Science | Medline8
Turner-Gomes 9
Tatara
CrossRef | Web of Science | Medline10
Suzuki
CrossRef11
Suzuki
CrossRef | Web of Science | Medline12
Pahl
CrossRef | Web of Science | Medline13
Kuribayashi 14
Tatara
CrossRef15
Yanagawa
Web of Science | Medline16
Schoenberg
CrossRef17
Becker
CrossRef | Medline18
Kohr
CrossRef | Web of Science | Medline19
Fujiwara
CrossRef | Web of Science | Medline20
Shulman
Web of Science | Medline21
Akagi
CrossRef | Web of Science | Medline
- Citing Articles (11)
Citing Articles
1
Bernard C. Portmann, Eve A. Roberts. 2012. Developmental abnormalities and liver disease in childhood. , 101-156.
CrossRef2
Nethnapha Paredes, Tapas Mondal, Leonardo R Brandão, Anthony KC Chan. (2010) Management of myocardial infarction in children with Kawasaki disease. Blood Coagulation & Fibrinolysis1
CrossRef3
Yosikazu Nakamura, Eiko Aso, Mayumi Yashiro, Ritei Uehara, Makoto Watanabe, Izumi Oki, Hiroshi Yanagawa. (2008) Mortality Among Persons With a History of Kawasaki Disease in Japan. Circulation Journal 72:1, 134-138
CrossRef4
Kouji Higashi, Masaru Terai, Hiromichi Hamada, Takafumi Honda, Masaki Kanazawa, Yoichi Kohno. (2007) Impairment of Angiogenic Activity in the Serum From Patients With Coronary Aneurysms Due to Kawasaki Disease. Circulation Journal 71:7, 1052-1059
CrossRef5
Yosikazu Nakamura, Eiko Aso, Mayumi Yashiro, Ritei Uehara, Makoto Watanabe, Morihiro Tajimi, Izumi Oki, Toshiyuki Ojima, Hiroshi Yanagawa, Tomisaku Kawasaki. (2005) Mortality among persons with a history of Kawasaki disease in Japan: Can paediatricians safely discontinue follow-up of children with a history of the disease but without cardiac sequelae?. Acta Paediatrica 94:4, 429-434
CrossRef6
N Wilson, P Heaton, L Calder, R Nicholson, S Stables, R Gavin. (2004) Kawasaki disease with severe cardiac sequelae: Lessons from recent New Zealand experience. Journal of Paediatrics and Child Health 40:9-10, 524-529
CrossRef7
RUSSELL PIERRE, ROBERT SUE-HO, DONNA WATSON MBBS. (2000) Kawasaki syndrome in Jamaica. The Pediatric Infectious Disease Journal 19:6, 539-544
CrossRef8
Yosikazu Nakamura, Hiroshi Yanagawa, Hirohisa Kato, Kensuke Harada, Tomisaku Kawasaki, . (1998) Mortality among patients with a history of Kawasaki disease: The third look. Pediatrics International 40:5, 419-423
CrossRef9
Yosikazu Nakamura, Hiroshi Yanagawa. (1996) A case-control study of recurrent Kawasaki disease using the database of the nationwide surveys in Japan. European Journal of Pediatrics 155:4, 303-307
CrossRef10
Y Nakamura, K Hirose, H Yanagawa, H Kato, T Kawasaki. (1994) Incidence rate of recurrent Kawasaki disease in Japan. Acta Paediatrica 83:10, 1061-1064
CrossRef11
R. I. Schiff. (1994) Intravenous gammaglobulin, 2: pharmacology, clinical uses and mechanisms of action. Pediatric Allergy and Immunology 5:3, 127-156
CrossRef
