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?

Aim: To clarify the question of whether patients with Kawasaki disease suffer a higher mortality rate after the incidence of the disease in comparison with age-matched healthy individuals. Methods: Between July 1982 and December 1992, 52 collaborating hospitals collected data on all patients having a new, definite diagnosis of Kawasaki disease. Patients were followed up until 31 December 2001 or their death. The expected number of deaths was calculated from Japanese vital statistics data and compared with the observed number. Results: Of 6576 patients enrolled, 29 (20 males and 9 females) died. The standardized mortality ratio (SMR: the observed number of deaths divided by the expected number of deaths based on the vital statistics in Japan) was 1.15 (95% CI: 0.77-1.66). In spite of the high SMRs during the acute phase, the mortality rate was not high after the acute phase for the entire group of patients. Although the SMR after the acute phase was 0.75 for those without cardiac sequelae, six males (but none of the females) with cardiac sequelae died during this period; and the SMR for the male group with cardiac sequelae was 1.95 (95% CI: 0.71-4.25). The mortality from congenital anomalies of the circulatory system was elevated, but no increase in cancer deaths was observed.

Conclusion: Although it was not statistically significant, the mortality rate among males with cardiac sequelae due to Kawasaki disease appeared to be higher than in the general population. On the other hand, the mortality rates for females with the sequelae and both males and females without sequelae were not elevated.     

Mortality among patients with a history of Kawasaki disease: The third look

Abstract Background: Long-term prognosis of Kawasaki disease is still unclear.
Methods: In a cohort study, 6576 patients with Kawasaki disease were observed from their first medical encounter because of the disease through the end of 1994, or until death. Standardized mortality ratios (SMR) with 95% confidence intervals (CI) were calculated with vital statistics data of Japan used for the control. Results: Of 6576 patients who met the eligibility criteria, 6550 (99.6%) were followed through either the end of the study or the date of death. Twenty patients (14 male, 6 female subjects) died during the study period; an overall SMR of 1.35 (95% CI 0.82–2.08) was calculated. The SMR was 1.45 (95% CI 0.79–2.44) for male subjects and 1.15 (95% CI 0.42–2.52) for female subjects. During the acute phase of the disease (the first 2 months after the first visit to hospital), the SMR was higher, particularly in male subjects (SMR 10.13, 95% CI 3.72–22.08). After the acute phase, however, both boys and girls had low SMR. Nine of the 20 deaths were caused by Kawasaki disease; there were three deaths as a result of congenital heart diseases and two subjects died of malignant neoplasms of lymphatic or hematopoietic tissues.
Conclusions: Although the mortality rate among those with a history of Kawasaki disease was elevated in Japan, many of the deaths that caused the elevation occurred during the acute phase of the disease. The mortality rate was not increased after the acute phase of the disease.     

Incidence of Kawasaki disease in Japan: the nationwide surveys of 1999–2002

BACKGROUND: The purpose of the present study was to describe the results of nationwide epidemiologic surveys of Kawasaki disease for the 4 year period 1999-2002. METHODS: The design is a retrospective incidence survey. The patients reported in these two surveys are all new patients who were reported during the two study periods (1999-2000 and 2001-2002), although the data were collected retrospectively. A questionnaire was sent to all pediatric departments of hospitals with 100 beds or more throughout Japan, requesting data on patients with Kawasaki disease, such as sex, age, date of first hospital visit, recurrence, and cardiac lesions. RESULTS: The total number of patients reported in the 4 year period 1999-2002 was 32 266 (18 604 male, 13 662 female), with an average annual incidence of 137.7 per 100 000 children younger than 5 years old. The male/female ratio was 1.30. The incidence peaked at 9-11 months of age, and the proportion of patients under 1 year of age was 26%. The monthly distribution had a high peak in January and a gradual increase in summer. Geographically, the high-incidence areas were limited to certain prefectures and moved from year to year. The cardiac lesions at acute stage and cardiac sequelae occurred more in children under 1 year and older than 4 years. Among the principal symptoms, fever persisting >or=5 days occurred most commonly, followed by conjunctival congestion, changes in lips and oral cavity, polymorphous exanthema, and changes of extremities. Cervical lymphadenopathy occurred less. CONCLUSION: More than 32 000 patients with Kawasaki disease during the 4 year period 1999-2002 were reported to the nationwide incidence surveys. The number of patients is steadily increasing despite the decrease of children. The seasonal variation, geographical distribution, and age-specific distribution support the infection theory for the etiology of Kawasaki disease.     

Kawasaki disease in parents and children

Aim: To estimate the probability that the parents of patients with Kawasaki disease also had a history of the same disease. Methods: Self-reported parents' histories of Kawasaki disease were collected from data of the 16th nationwide survey of the disease conducted in Japan from January 1999 to December 2000. The incidence of Kawasaki disease was calculated by using data reported in all 16 nationwide surveys and live births in the Japanese vital statistics. The expected number of parents with a history of Kawasaki disease in the general population, which was calculated by using the assumed number of parents in the vital statistics and the incidence of this disease, was compared with the observed number. Results: Among 14 163 parent pairs of patients with Kawasaki disease, 33 parents (25 mothers and 8 fathers) had a history of the disease. The number of parents expected to have a history of Kawasaki disease was 16.1 (8.4 mothers and 7.7 fathers). From a Poisson distribution, the probability of the observed number was less than 0.001 among parents or mothers. The prevalence of a recurrence of Kawasaki disease and incidences involving siblings of patients whose parents had a history of the disease were five or six times higher than those of all patients who were reported in the 16th survey.

Conclusion: When compared with parents in the general population, the probability of a history of Kawasaki disease was significantly higher in those parents whose children suffered from the same disease. This suggests that, epidemiologically, a genetic predisposition to Kawasaki disease may be implicated in its occurrence.     

Cardiac sequelae of Kawasaki disease in Japan over 10 years

To observe the secular trend of a proportion of Kawasaki disease patients with cardiac sequelae in Japan, we analyzed patients with Kawasaki disease reported to nationwide surveys of the disease during 10.5 years from July 1982 to December 1992. Of 69 382 patients reported to the surveys, 10 596 (15.3%) were reported to have cardiac sequelae such as dilatation or stenosis of coronary arteries, myocardial infarction or valvar lesions, 1 month or more after onset. The percentage of cardiac sequelae was particularly high in males, infants younger than 1 year and children older than 5 years of age. The overall prevalence declined steadily over the observed period. However, the percentage for children older than 5 years of age did not decrease, whether treated with intravenous gamma globulin or untreated. As a consequence of the increased number of patients treated with intravenous gamma globulin, the proportion of Kawasaki disease patients with cardiac sequelae decreased annually. However, the proportion of children older than 5 years of age did not decrease.     

Epidemiologic and clinical characteristics of Kawasaki disease in Shaanxi Province, China, 1993-1997

The objective of this paper is to describe the epidemiological and clinical characteristics of Kawasaki disease (KD) in Shaanxi Province, China during the 5-year period from January 1993 to December 1997. A province-wide epidemiological survey on KD was made by the China-Japan Kawasaki Disease Study Group. The questionnaire form and the diagnostic criteria of KD, which were prepared by the Japan Kawasaki Disease Research Committee and translated into Chinese, were sent to the departments of pediatrics of all the hospitals with 100 beds or more in Shaanxi province. All the KD patients who were diagnosed during the observation period from 1993 to 1997 were asked to take part in this survey. The databases of reported KD in this survey were analysed at the Department of Pediatrics of the Shaanxi Provincial People's Hospital, Xi'an, China. All the patients that satisfied the diagnostic criteria were included in the report. A total of 105 (70 per cent) hospitals responded and 376 cases of KD were confirmed. More cases were reported in 1993 and 1994. Of the total patients reported, 69 per cent were children under 3 years old with a male to female ratio of 1.6:1. The proportion of patients with cardiac sequelae was 19 per cent with a male to female ratio of 3:1. There were four fatal cases with a fatality rate of 1 per cent. It is concluded that KD is common in China. Continuous surveillance is necessary to maintain high awareness of KD so as to find possible risk factors and their association with the disease.     

Nationwide survey on Kawasaki disease in Japan

Results of the statistical analysis of a nationwide survey of patients with Kawasaki disease diagnosed within a 2-year 6-month period, from July 1982 to December 1984, are as follows. (1) The cumulative number of patients reported by the end of 1984 was 63,399 (36,891 boys and 26,508 girls; male to female ratio 1.4). (2) There were two epidemic years, 1979 and 1982, in which the numbers of patients were more than twofold that of the previous years. (3) A curve plotted for age-specific incidence rate showed a unimodal peak at age 1 year. (4) Steroid therapy was used for 6.3% of the patients, aspirin for 89.8%, antibiotics for 57.6%, and gamma-globulin for 11.4%. (5) The proportion of sibling cases was 1.4% and that of recurrent cases was 3.9% of all cases reported. (6) The incidence of cardiac sequelae in 1 month after disease onset was 17.2%.     

上海地区1998－2002年川崎病流行病学特征

目的 对上海地区川崎病发病情况进行调查，了解中国发达地区川崎病的流行病学特征。方法 参照日本流行病学调查方案，制作统一调查表和诊断指南，调查对象为1998年1月1日－2002年12月31日上海地区50家有儿科临床服务的二级及二级以上医院收治的所有川崎病病例。结果 调查表回收率100％，将18例不符合表格填写要求者剔除后，对768例进行分析。上海地区5岁以下儿童川崎病发病率呈逐年增高趋势， 1998年为16.79/10万，1999年为25.65/10万，2000年为28.16/10万，2001年为28.05/10万，2002年为36.76/10万。男女性别比为1.83:1；发病年龄为1个月～18.8岁（中位数1.8岁），发病年龄高峰为9.6个月；全年均可发病，但以春夏之交较多见。主要症状表现为发热持续5 d以上者最为常见（99.3％），其后依次为口唇、口腔黏膜损害（83.5％）、指（趾）端脱皮（82.9％）、皮疹（81.0％）、结膜充血（78.4％）、颈部淋巴结肿大（69.3％）、肢端充血硬肿（48.1％），45.2％的病例有肛周脱皮。心血管损害发生率24.3％，以冠状动脉扩张最为常见，占68％，其次为冠状动脉瘤，占10%。急性期病死率为0.26%，死因为冠状动脉瘤破裂和急性心力衰竭。再发率为1.82%。 结论 上海地区川崎病发病率明显低于日本，但高于西方国家。发病率呈增高趋势，性别分布和心血管损害与其他报道相似。发病季节分布与北京地区的报道相似，但与其他报道不同。     

Increasing incidence of Kawasaki disease in Japan: Nationwide survey

Background: The last nationwide survey of Kawasaki disease in Japan was conducted in 2003; the epidemiologic features of the disease in Japan since then are unknown.
Methods: All the hospitals with a pediatric department and a bed capacity of ≥100, or hospitals having a bed capacity of <100 but specializing in pediatrics in Japan were asked to report (by mail) all patients with Kawasaki disease who visited the hospital due to the disease in 2003 and 2004.
Results: Of the 2335 hospitals asked to participate, 1058 reported 19 138 patients with Kawasaki disease during the 2 year period 2003–2004. The annual incidence in 2004 reached 174.0 per 100 000 children under 5 years of age. The yearly number of patients and annual incidence increased significantly after the last nationwide epidemic of the disease in 1986. The number of patients was largest in January and lowest in October. The age-specific incidence was highest for those aged 9–11 months, and 88.9% of the patients were under 5 years of age. Of the reported patients, 2595 (13.6%) had cardiac lesions in the acute phase of the disease and 844 (4.4%) had cardiac sequelae 1 month after the onset of the disease.
Conclusion: The number of patients and the incidence of Kawasaki disease in Japan has increased year by year in the last two decades.     

A multicenter collaborative study on the risk factors of cardiac sequelae due to Kawasaki disease: a one-year follow-up study

Objective: To measure the prevalence of cardiac sequelae 1 y after the onset of Kawasaki disease and determine the risk factors associated with these cardiac sequelae. Material and methods: 1594 patients who initially visited one of the 87 target hospitals in 1996 for Kawasaki disease participated. Selection of the target hospitals was based on a nationwide survey. The patients were followed‐up and information concerning cardiac sequelae occurring within 1 y of onset was obtained by mail survey. Results: The prevalence of cardiac sequelae 1 mo after onset was 10.2% and decreased to 4.2% in 1 y. The prevalence was higher among males than females and higher in patients less than 1 y and 5 ys or older than in 1–4 year‐olds. Of the patients with cardiac sequelae at 1 mo, the sequelae disappeared in 60.7% after 1 y. Analysis revealed low serum albumin as a risk factor related to the occurrence of cardiac sequelae 1 y after onset. Of the 1594 patients, 10 had giant anuerysms and 3 had a fatal outcome. Conclusions: Approximately 60% of cardiac sequelae due to Kawasaki disease that developed within 1 mo after onset disappeared in 1 y. The odds ratio was significantly higher among patients with a low serum albumin level 1 y after onset.     

Leukocyte counts in patients with Kawasaki disease: from the results of nationwide surveys of Kawasaki disease in Japan

In the latest survey of Kawasaki disease in Japan, questionnaires on maximal leukocyte counts of the patients were included to clarify whether leukocyte counts could be of value for the diagnosis and prediction of outcome. A questionnaire form and diagnostic guidelines for Kawasaki disease were sent to all pediatric departments of hospitals with ≥100 beds throughout Japan, and information including maximal leukocyte counts was obtained on patients with Kawasaki disease diagnosed during the 2-y period from January 1993 to December 1994. Of the 11 458 patients diagnosed during the 2-y period, maximal leukocyte counts were reported in 11 062 patients (96.5%). The mean value and the distribution of maximal leukocyte counts were lower in the age group under 1 y. The mean values and the distribution of leukocyte counts were lowest in suspected cases among three diagnostic categories: typical cases of Kawasaki disease, atypical cases, and suspected cases. The mean values of maximal leukocyte counts of the patients with cardiac sequelae were significantly higher than those without cardiac sequelae in each age group. The proportion of patients with cardiac sequelae increased with leukocyte counts in each age group. The Receiver/Response Operating Characteristic (ROC) curve for maximal leukocyte counts in Kawasaki disease revealed that the accuracy of maximal leukocyte counts for prediction of cardiac sequelae was highest in the age group <6 months, and the most accurate cut-off point was 16 × 10⁹/l. The strongest association between higher leukocyte counts (≥16 × 10⁹/l) and cardiac sequelae was observed in the age group <6M. A large-scale analysis of leukocyte counts in patients with Kawasaki disease revealed age-dependent relationship between maximal leukocyte counts, diagnostic categories and outcome. Maximal leukocyte counts may be helpful for the prediction of outcome with the consideration of age.     

Cardiac sequelae of Kawasaki disease in Japan: statistical analysis.

The proportions of Kawasaki disease patients with cardiac sequelae in Japan were analyzed using nationwide survey data from the 6 1/2-year period July 1982 through December 1988. Of 46,864 cases of Kawasaki disease reported in the surveys, 7637 or 16.3% had cardiac sequelae such as dilation or stenosis of coronary arteries, myocardial infarction, and valvar lesions 1 month or more after onset. The prevalence of cardiac sequelae was particularly high in males, infants younger than 1 year, and children older than 5 years of age. In sequential observation, there was no correlation between the prevalence of cardiac sequelae and periods of high or low incidence of the disease. The prevalence of cardiac sequelae overall declined steadily over the observation period, perhaps as a consequence of increasing use of intravenous gamma globulin. However, children older than the age of 5 years manifested increasing prevalence of cardiac sequelae over the observation period, probably as a result of lower rates of intravenous gamma globulin administration.     

Treatment of osteopenia in children with insulin-dependent diabetes mellitus: The effect of 1α-hydroxyvitamin D3

All Austrian patients with insulin dependent diabetes diagnosed between 1979 and 1990 and age at onset below 15 years were followed from manifestation until death or until 31 December 1990 by cross linking the diabetes registry data with the National Mortality database (death certificates). Out of the cohort consisting of 1185 cases, 6 had died during the study period, resulting in a standardized mortality ratio of 1.53 for the total cohort. Of the patients 50% died due to acute diabetic complications including 2 children at onset of the disease. The risk for premature death in this cohort of very young insulin dependent diabetes mellitus patients was only slightly increased. In comparison to previous studies it is rather low, but comparable to recent investigations in northern Europe. Conclusion Although the mortality of diabetic children and adolescents in Austria is only marginally higher than in nondiabetics, there are still deaths which should be preventable in this age group. Received: 20 February 1996 / Accepted: 27 May 1996     

Cardiac transplantation for Kawasaki disease

OBJECTIVE: Severe cardiac sequelae from Kawasaki disease include coronary ischemia and have been treated with a variety of coronary artery bypass procedures. There is only one published report of a child who underwent cardiac transplantation for severe Kawasaki disease-related cardiac complications. The purpose of this study was to gather the worldwide experience with cardiac transplantation for Kawasaki disease. METHODS: Data were obtained from the United Network for Organ Sharing Registry, the European transplant experience, and a phone survey of many Kawasaki disease investigators. Diagnostic and surgical reports as well as clinical records were reviewed. Results. We identified 13 Kawasaki disease patients who underwent cardiac transplantation and obtained data on 10. In these 10 patients, the timing of transplantation was within 6 months after diagnosis of Kawasaki disease (4 patients), 1 to 5 years after diagnosis (3 patients), and 9 to 12 years after diagnosis (3 patients). Indications for transplantation included severe myocardial dysfunction, severe ventricular arrhythmias including cardiac arrest, and severe distal multivessel occlusive coronary artery disease. Nine of the 10 patients remain alive and healthy, with up to 6 years' posttransplant follow-up. One patient died 10 months posttransplant after severe refractory rejection. In addition, 1 patient required retransplantation at 4 years for severe rejection. CONCLUSIONS: Cardiac transplantation for severe ischemic heart disease as a sequela of Kawasaki disease is feasible and can benefit the small subgroup of patients who are not candidates for revascularization because of distal coronary stenosis or aneurysms and/or those with severe irreversible myocardial dysfunction.     

Effects of Gamma-Globulin on the Cardiac Sequelae of Kawasaki Disease

Our aim was to delineate the effect of various factors, such as sex, age, serum albumin levels, and the timing of gamma-globulin (GG) therapy, on cardiac sequelae of Kawasaki disease. The patients with Kawasaki disease who were reported at the 1995–1996 nationwide survey and received 2000 mg/kg at specified hospitals were selected as the subjects of the study. A total of 2221 patients actually received the basic dose. The relationships of the GG therapy with the cardiac sequelae, sex, age, timing of GG administration (the date of initiation and duration of the regimen following disease onset), and serum albumin levels were examined by using logistic regression analysis. The odds ratios for the cardiac sequelae in patients with Kawasaki disease were high in males (1.63), in those under the age of 1 year (1.54), and in those with a serum albumin level <3.2 g/dl (2.64). The odds ratio was low in those who received GG before day 8 of the illness (0.69) or in those for whom the administration period was for 2 days or less (0.67). To prevent cardiac sequelae of Kawasaki disease it is desirable that GG therapy be started as soon as possible and completed within 2 days.     

Early mortality in childhood diabetes in Austria – a population based cohort study

All Austrian patients with insulin dependent diabetes diagnosed between 1979 and 1990 and age at onset below 15 years were followed from manifestation until death or until 31 December 1990 by cross linking the diabetes registry data with the National Mortality database (death certificates). Out of the cohort consisting of 1185 cases, 6 had died during the study period, resulting in a standardized mortality ratio of 1.53 for the total cohort. Of the patients 50% died due to acute diabetic complications including 2 children at onset of the disease. The risk for premature death in this cohort of very young insulin dependent diabetes mellitus patients was only slightly increased. In comparison to previous studies it is rather low, but comparable to recent investigations in northern Europe. Conclusion Although the mortality of diabetic children and adolescents in Austria is only marginally higher than in nondiabetics, there are still deaths which should be preventable in this age group.     

Cardiac sequelae of Kawasaki disease among recurrent cases

OBJECTIVE: This study was undertaken to clarify whether cardiac sequelae due to Kawasaki disease are more frequent among recurrent cases than initial onset cases. STUDY DESIGN: A cross sectional study using data from nationwide surveys of Kawasaki disease in Japan was conducted. A total of 33,976 patients reported were divided into two groups: initial onset cases (32,923 patients) and recurrent cases (1053 patients). Proportions of cardiac sequelae such as coronary aneurysms/dilatation, coronary stenosis/narrowing, myocardial infarction, and valvular lesions were compared between the two groups. RESULTS: The proportions of patients with the sequelae were significantly more common among recurrent cases. In men 25.5% of the recurrent cases had the sequelae in comparison with 14.9% for initial onset cases, and in women 16.1% of recurrent cases had the sequelae compared with 9.8% of initial onset cases. Giant coronary aneurysms were twice as likely in men in whom the disease was recurring than in initial onset cases, and 1.5 times more likely in women in whom the disease was recurring than in initial onset cases. CONCLUSION: Cardiac sequelae of Kawasaki disease are more likely to appear on recurrent case patients.     

Kawasaki disease in Sweden: incidence and clinical features

The incidence and clinical pattern of patients with Kawasaki disease in Sweden and the outcome of treatment with iv immunoglobulin (Sandoglobulin) and aspirin were examined in a national prospective study over a 2-year period. Cases not referred to the study were identified by inquiry. Ninety-nine children were diagnosed as having Kawasaki disease. The annual incidence rate was calculated to be 2.9per 100 000 in children younger than 16 years of age and 6.2 per 100000 in children younger than 5 years of age. The median age of our patients was 2.2 years and the male-to-female ratio was 2.3:1. Cardiac ultrasonography revealed abnormalities in 33% of all patients, and 14% of cases in the prospective study had pronounced abnormalities. In most cases (91%) treatment had a prompt effect on fever and morbidity in general, and side effects were mild. Two infants, identified from the inquiry, died from rupture of an aneurysm in the coronary artery in the acute phase of the disease. The risk of cardiac involvement is obvious and emphasizes the importance of early diagnosis and treatment. Incidence, Kawasaki disease, vasculitis     

Serum C-reactive protein levels in patients with Kawasaki disease: from the results of nation-wide surveys of Kawasaki disease in Japan

Thirteen nation-wide epidemiological surveys of Kawasaki disease have been carried out successively since 1970 in Japan. In the latest survey, questionnaires on serum C-reactive protein (CRP) levels of the patients were included to clarify whether serum CRP levels could be available for the diagnosis and prediction of prognosis. A questionnaire form and diagnostic guidelines for Kawasaki disease were sent to all paediatric departments of hospitals with 100 or more beds throughout Japan, and information including maximal serum CRP levels was obtained on patients with Kawasaki disease diagnosed during the 2-year period from January 1993 to December 1994. Of the 11 458 patients diagnosed during the 2-year period, maximal serum CRP levels were reported in 11040 patients (96.4%). The values of maximal serum CRP were higher in the age groups < 6 months and >2 years. The mean value and the distribution of serum CRP levels in suspected cases were lowest among the three diagnostic categories and this difference among diagnostic categories was highly significant in the age groups 6M-1Y and 1-2Y. The proportion of patients with cardiac sequelae increased with serum CRP levels in each age group. The mean value and the distribution of serum CRP levels of the patients with cardiac sequelae was higher than those without it and this difference between cardiac prognoses was outstanding in the age groups 6M-1Y and <6M. The Receiver/Response Operating Characteristic (ROC) curve for maximal serum CRP levels in Kawasaki disease revealed that accuracy of maximal serum CRP levels for prediction of cardiac sequelae was highest in the age group 6M-1Y. A large-scale observation and analysis of serum CRP levels of the patients with Kawasaki disease revealed age-dependent relationships among maximal serum CRP levels, diagnostic categories and prognosis. Serum CRP levels may be helpful for the prediction of prognosis with the consideration of age.     

Validation of pediatric index of mortality 2 (PIM2) in a single pediatric intensive care unit of Argentina.

OBJECTIVE: Pediatric Index of Mortality 2 (PIM2) is an up-to-date mortality prediction model in the public domain that has not yet been widely validated. We aimed to evaluate this score in the population of patients admitted to our pediatric intensive care unit. DESIGN: Prospective cohort study. SETTING: Multidisciplinary pediatric intensive care unit in a general university hospital in Buenos Aires, Argentina. PATIENTS: All consecutive patients admitted between January 1, 2004, and December 31, 2005. INTERVENTIONS: None. MEASUREMENTS AND MAIN RESULTS: There were 1,574 patients included in the study. We observed 41 (2.6%) deaths, and PIM2 estimated 48.1 (3.06) deaths. Discrimination assessed by the area under the receiver operating characteristic curve was 0.9 (95% confidence interval, 0.89-0.92). Calibration across five conventional mortality risk intervals assessed by the Hosmer-Lemeshow goodness-of-fit test showed chi5 = 12.2 (p = .0348). The standardized mortality ratio for the whole population was 0.85 (95% confidence interval, 0.6-1.1). CONCLUSIONS: PIM2 showed an adequate discrimination between death and survival and a poor calibration assessed by the Hosmer-Lemeshow goodness-of-fit test. The standardized mortality ratio and clinical analysis of the Hosmer-Lemeshow table make us consider that PIM2 reasonably predicted the outcome of our patients.