Our study involved 142 children with EV71 CNS infections that included 61 (43%) with viral meningitis, 53 (37%) with severe CNS involvement including encephalitis, polio-like syndrome and encephalomyelitis, and 28 (20%) with cardiopulmonary failure after CNS involvement. These children were subjected to physical and neurological examinations 2.85 (range 1.0-7.39) years after disease onset. Those below the age of 6 years took the Denver developmental screening test, while those 4 years and over took the intelligence quotient (IQ) test.

For the purposes of our study we identified all the EV71 pediatric patients at Chang Gung Children's Hospital (CGCH) and at the NTUH between 1998 and 2003. These patients were clinically confirmed to have hand, foot, and mouth disease (HFMD) of herpangina or febrile illness, and a positive laboratory-confirmed EV71 infection. The EV 71 infection was confirmed on the basis of a positive viral isolation of EV71 and/or a positive EV71 IgM and/or a four-fold rise in EV71 neutralizing antibody serotiters between the acute and the convalescent sera.

This study involved 621 EV71 patients in total, 534 of them at the CGCH and 87 of them at the NTUH. Of these, 232 cases (37.4%) had CNS involvement. The clinical severity of the EV71 CNS involvement was classified in terms of the increasing severity of the infection: Group 1: cases with mild CNS involvement i.e. aseptic meningitis; Group 2: cases with severe CNS involvement including encephalitis, polio-like syndrome or encephalomyelitis; and Group 3: cases with cardiopulmonary failure after central nervous system involvement. Patients placed in Group 1, were those experiencing headaches, irritability and CSF pleocytosis (>5x106 leukocytes/L) but no altered level of consciousness or focal signs. Patients placed in Group 2, were encephalitis with an altered level of consciousness plus CSF pleocytosis, poliomyelitis-like syndrome with acute limb weakness, and decreased reflex and muscle strength, and encephalomyelitis with the occurrence of both encephalitis and poliomyelitis-like syndrome. Patients placed in Group 3 were those who had experienced cardiopulmonary failure (defined as a decreased ejection fraction of the left ventricle as assessed by echocardiography with or without pulmonary edema/hemorrhage, necessitating inotropic agent and ventilator support). Patients who experienced cardiopulmonary failure after CNS involvement have been found to have had a sudden onset of cardiopulmonary failure several hours or several days after the manifestations of their EV71 CNS infections.5,16 Of these 232 cases with CNS involvement, 25 (10.8%) patients died of cardiopulmonary failure and brainstem encephalitis during acute illness. 19 of these 25 patients died within 7 days of the onset of acute illness, while 14 (6.0%) patients died due to a deep coma or aspiration pneumonia during the convalescent stage (i.e. more than one month after onset of their disease).

Of the remaining 193 patients with CNS involvement (172 of these at the CGCH and 21 of these at the NTUH), 22 refused to be assessed and 29 could not be located. This means that a total of 142 (73.6%) patients were enrolled in our study and took the assessment between January 2003 and December 2005 after informed consent was obtained from their parents (Figure 1). The clinical severity and the demography of the 142 patients who were assessed and the 51 patients who were not assessed (p=0.22 for clinical severity, p=0.33 for their age of onset, and p=0.35 for genders) were similar. It is therefore clear that the size of the study population involving these 142 patients is representative of the 193 EV71 patients with CNS involvement.

Clinical and neurological outcome and neurodevelopment assessment All the children were physically and neurologically examined by a pediatrician or a pediatric neurologist during an outpatient visit or during their stay at chronic respiratory centers. Their physical handicap or neurological sequelae, the requirement of ventilator support and the need of tube feeding were recorded.

EV71 patients who were younger than 6 years of age during our assessment were tested with the Denver Developmental Screening Test (DDST II). The DDST II has four categories: gross motor, fine motor, language, and personal-social.17 Each test item was scored as either passed or failed. The overall assessment of each child was considered normal when the child passed tasks in 25-75% for his/her age, or delayed when the child failed tasks to the left of his/her age line.

Cognitive function assessment The Wechsler Intelligence Scale for Children-Third Edition (WISC-III, Wechsler, 1991) profile was individually assessed by a child psychologist for all children over four years of age except for 3 of the children who had tracheostomy and could not talk. The WISC-III is a widely used measure of general intelligence for children aged 4-16 years. The WISC-III is composed of 13 subtests to test children's cognitive ability of different dimensions, which are grouped into two scores: the performance IQ score and the verbal IQ score. The Performance IQ score (7 subtests) includes Picture Completion, Block Design, Object Assembly, Picture Arrangement, Coding, Symbol Search and Mazes subtests. The Verbal IQ score (6 subtests) includes Information, Comprehension, Arithmetic, Similarities, Digit Span and Vocabulary subtests.18 Four factorially derived composite subscales have been created: (1) Verbal Comprehension: Information, Similarities, Vocabulary, and Comprehension; (2) Perceptual Organization: Picture Completion, Picture Arrangement, Block Design, and Object Assembly; (3) Freedom From Distractibility: Arithmetic and Digit span; and (4) Process Speed: Coding and Symbol Search. 18 Each of the IQ scores and four composite subscales yield standard scores with a mean of 100 and a standard deviation (S.D.) of 15.

Statistical Analysis Data were analyzed with the SAS Statistical Package (Version 9.1, SAS Institute, Cary, North Carolina). Data are expressed as mean ±S.D., median (range) or number (percentage). The Chi-square test was used for categorical data, and the student t-test and analysis of variance (ANOVA) were used for continuous variables with normal distribution. The Mann-Whitney rank sum test or Kruskal-Wallis test was used for continuous variables without a normal distribution in the univariate analysis. If a significant difference was found with ANOVA, pairwise comparison would be performed with the Scheffe test. Multivariate analysis of variance (MANOVA) was used to identify the most significant factors affecting the cognitive function (IQ and the four composite subscales).. P values less than 0.05 were considered significant.

Inclusion Criteria:

• Virologically-confirmed EV71 cases with CNS involvement

Exclusion Criteria:

• No evidence of EV71 infection
• No evidence of CNS involvement