| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
 |
ABSTRACT |
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
|
INTRODUCTION |
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Long-term disability is an important component of the JE disease burden. Such disease burden cannot be fully estimated by short-term follow-up studies because some neurologic disorders resolve over time and others appear at varying intervals after the acute illness.1 Studies with suitably long periods of follow-up are lacking, and past studies of this topic have not been properly controlled or observer-blinded. Here we describe a controlled, blinded assessment of long-term neurologic disability in persons who acquired JE 6–27 years earlier in Shanghai, China.
|
MATERIALS AND METHODS |
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
JE and non-JE study subjects. In all, we identified 329 persons (study subjects) who had been hospitalized when < 15 years old (1973–1994) for putative JE, in one of two major Shanghai pediatric hospitals (Shanghai Children’s Hospital and Shanghai Medical University-affiliated Pediatric Hospital). These subjects had been discharged alive. We reclassified these patients as JE or non-JE based on whether laboratory results in the index hospitalizations fulfilled certain criteria (vide infra). These two groups constituted historic cohorts that were examined and compared in 2000 for neurologic disability.
Collection and interpretation of hospitalization data. Hospital records were evaluated of encephalitis cases. The records were scrutinized to obtain demographic and clinical data. Because of the diagnostic fashion at the time, irrespective of laboratory test results, all subjects in the two cohorts had received clinical diagnoses of JE at discharge on the basis of seasonal occurrence and clinical manifestations of acute encephalitis.17
For this study, we examined the laboratory results during the hospitalizations to reclassify patients as either JE or non-JE (all surviving patients with discharge diagnoses of JE had had one or more immunoassays done on serum obtained during their hospitalizations). Before 1990, complement fixation (CF) and hemagglutination inhibition assays (HIA) were used for laboratory diagnosis in the two study hospitals. After 1990, IgM-capture enzyme-linked immunosorbent assay (ELISA) became the standard. We reclassified a patient’s illness as JE if acute or convalescent serologic assays showed 1) HI antibody titer 
1:80,18,19 2) CF antibody titer 1:4,18 or 3) positive results in the IgM-capture ELISA. These diagnostic criteria were formulated before the study.18,19 Although these criteria do not discriminate JE from other flavivirus infections with complete accuracy, we judged the criteria to be sufficient because neither dengue fever nor West Nile virus has been observed in Shanghai.20
In all, 93% of the study subjects were tested by CF, whereas 4% and 3% cases were tested by IgM-capture ELISA and HIA, respectively. Overall, 160 of the 329 study subjects were defined as JE and 169 cases were defined as post-non-JE; of the non-JE cases, 92% had no detectable CF antibody in both admission and discharge specimens. In the course of assembling these 329 survivors of encephalitis classified clinically as post-JE, we discovered an additional 15 cases who died during the index hospitalizations for putative JE. Five tested positive for JE and five tested negative. The remaining five died before serologic specimens were obtained. Because the patients and their families were unaware of these reclassified diagnoses and because the JE patients were contemporaneous with the non-JE patients, the non-JE group served as a useful control group for follow-up assessments of the prevalence of neurologic disability attributable to JE.
Assembly of neighborhood controls. Because the non-JE hospital controls consisted of persons who had had acute encephalitis and thereby might be expected to have long-term neurologic sequelae, we also assembled a second control group from the general population. For this second set of controls, we selected one person, matched to each confirmed JE case by sex and age (±1 year) from the residence immediately to the right of the case residence. We excluded from this control group any subject with a history of an acute neurologic illness. If an eligible consenting subject was not available next door, we searched consecutive residences to the right of the case residence until one eligible consenting subject was identified. No subjects had to be excluded because of past neurologic events, and all subjects approached for participation gave their consent.
Follow-up of subjects. The follow-up of study subjects from the three groups was done in 2000. To assemble patients who had earlier survived encephalitis (JE or non-JE), we attempted appointments at the Huashan Hospital clinics. Home interviews were made if the patients did not come to the appointments. We used the registered permanent residence records of the Shanghai Police Bureau to search for patients who could not be contacted by home visit or by postal communication.
After efforts to trace the discharged patients, we located 185 of the 329 subjects or their guardians, and 158 agreed to participate in the study and signed informed consent forms. Of these 158 subjects, 85 (78 alive, 7 died after discharge) had confirmed JE and 73 (71 alive, 2 died after discharge) had non-JE encephalitis (Figure 1
).
|
Structured neurologic examinations included evaluation of seizures, speech, autonomic nerve function, cranial nerve function, motor and sensory functions, and reflexes.21 Intelligence Quotient (IQ) measurement and Mini-Mental State Examination (MMSE) were used for intelligence evaluation. We evaluated children aged 6–16 years with the Wechsler Intelligence Scale for Children (WISC) and used the Wechsler Adult Intelligence Scale (WAIS) for subjects 17 years and older. Adults were tested with an urban or a rural version of test according to their residence location. Both of these tests have been adapted for use among Chinese populations.22,23 A proportion of patients refused IQ testing, which took 2 or more hours.
Definitions of endpoints. Neurologic disability was defined as any abnormal finding on structured neurologic examination. IQ scores of 70–84 were classified as mildly abnormal (> 1 but < 2 SE below normal range) and those < 70 (> 2 SE of normal range) were considered severely abnormal.24 An MMSE score < 21 was regarded as abnormal.25 For the activities of daily living (ADL) scale, an assessment of overall functioning that considers 20 basic functions of daily activities,25 moderate and severe disabilities were categorized by scores of 21–60 and 61–80, respectively.
Statistical analysis.
Comparisons of the JE group with each of the two comparison groups for dimensional variables were evaluated with the Student t test or the Mann-Whitney U test when the data were not normally distributed. Those for dichotomous variables were assessed by 
2 test or Fisher exact test when data were sparse. A 2 test for trend was used to test for trend in odds ratios among ordered categories.26 To compare dichotomous variables, after adjusting for confounding factors, we fitted multiple logistic regression models. For contrasts of the two post-encephalitic groups, we forced the following potential confounders into the models: age (years) at discharge, sex, years since the index hospitalization, and rural versus urban residence at the time of the index hospitalization. For comparisons of the JE group with the neighborhood control group, we controlled for age (years) at follow-up, sex, and rural versus urban residence at the time of follow-up. The coefficient for the variable for the group to which the subject belonged in these models was exponentiated to estimate the adjusted odds ratio for the association, and the SE for the coefficient was used to estimate the 95% confidence interval (CI) for the adjusted odds ratio.27–29 All P values and CIs were estimated in a two-tailed fashion. Differences were considered to be statistically significant at P < 0.05.
|
RESULTS |
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
), although comparison of traced patients between the two groups revealed that more years had elapsed from the time of index hospitalization to follow-up for the JE than for the non-JE group (17.2 versus 14.5 years, P < 0.001), and more JE subjects than non-JE patients lived in rural areas (93% versus 45%, P < 0.01) that is characteristic for a mosquito-borne disease such as JE (Table 1). Moreover, traced JE patients had been hospitalized significantly longer and had a higher proportion of neurologic deficits at discharge than the traced non-JE patients that indicates severer course of the disease for the JE than for the non-JE group (Table 1).
|
; Figure 2). Eight of the 17 had motor defects: 4 with paralysis, 2 with hemiplegia, and 1 each with monoplegia or chorea. Six of these eight had aphasia combined with motor deficits. These deficits were evident at discharge from the index hospitalizations and showed no apparent improvement with time after discharge. In addition, 10 JE patients had seizures of diverse types: 9 with generalized tonic-clonic seizure (GTC) and 1 with focal seizure.
|
|
), both appearing after discharge. Another two patients who had motor deficit during hospitalization recovered after discharge and were normal at follow-up examinations. The odds ratio for the presence of at least one neurologic deficit in the JE versus non-JE groups, adjusted for age, sex, residence, and years elapsed since index hospitalization, was 6.30 (95% CI: 1.24–31.91, P < 0.05). None of neighborhood controls exhibited a neurologic abnormality on physical examination (Table 2). Neurologic abnormalities such as muscular rigidity and motor weakness, evident at discharge, had resolved in 11 JE cases (13%) at the time of follow-up. On the other hand, eight (10%) JE patients developed seizure disorders 3 to 17 years after discharge; another two (3%) JE patients showed dysfunction of autonomic nerve or optic atrophy at the follow-up examination that had not been noted on examinations during index hospitalizations.
A total of 10 patients with seizure disorders were found in the JE group compared with 2 in the non-JE group (Figure 2). The odds ratio of having a seizure disorder in JE versus non-JE patients, adjusted for age, sex, years of follow-up, and residence, was 3.03 (95% CI: 0.52–17.76).
Cognitive function for the three groups at follow-up.
In all, 78 JE patients, 71 non-JE encephalitis patients, and 78 neighborhood controls were evaluated by the MMSE. Eight (10%) JE patients and one (1%) non-JE patient had an MMSE score < 21, whereas all members of neighborhood control group had normal MMSE scores. The differences between JE patients versus non-JE patients and between JE patients versus neighborhood controls were all statistically significant (Table 2).
IQ tests were administered to 60 (76%) JE patients, 53 (74%) non-JE patients, and 67 (85%) neighborhood controls. Of those, 17 (28%) JE patients had IQ scores < 85, 11 (18%) had severe mental retardation with IQ scores < 70, and 6 (10%) had IQ scores of 70–84. Only one (2%) non-JE encephalitis patient had an abnormal IQ (IQ = 78), and only one (2%) neighborhood control had an abnormal IQ (IQ = 82; Table 2). All differences in the prevalence of low IQ scores between JE patients versus non-JE patients and between JE patients versus neighborhood controls were statistically significant (Table 2). IQ scores were significantly lower in the JE group than in the post non-JE group (P < 0.001) and in the neighborhood control group (P < 0.001; Table 2). The odds ratios of having a low IQ, adjusted for age, sex, years elapsed since index hospitalization, and residence were 20.72 (95% CI: 2.15–199.26; P < 0.01) for JE patients versus non-JE patients and 26.64 (95% CI: 3.40–208.97; P < 0.001) for JE patients versus neighborhood controls.
Overall function status by ADL score at follow-up.
All 71 surviving non-JE patients and 78 neighborhood controls had normal ADL scores (Table 2). In contrast, only 66 (85%) JE patients had normal ADL scores (P < 0.01 versus post-non-JE; P < 0.001 versus neighborhood controls). Of importance, five (6%) JE patients had ADL scores indicative of near incapacitation.
|
DISCUSSION |
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Relation to other studies. Schneider et al4 reported the neurologic abnormalities in 37% of pediatric JE cases and 26% of adult cases after 1 year, and Hoke et al5 reported abnormal neurologic status in 45% of patients 3 months after acute JE illness. No longer follow-up evaluations were done.
Kumar et al10 found major neurologic sequelae in 46% of JE patients 12–18 months after discharge, but this rate was exaggerated because only 55 of 90 patients returned to the hospital for evaluation. Home visits were not done and the authors admitted that subjects without sequelae were less likely to return for follow-up.
Lin30 followed 50 JE cases in China for 5–8 years and compared them with 50 neighborhood control children. He found that 18% of the JE patients had low IQs. Unfortunately, the JE diagnosis in the study of Lin was not confirmed by laboratory assays.
Wang and Chen31 followed 54 of 77 serologically proved JE patients for 3–8 years after initial diagnosis. Of the 54, 23 (43%) exhibited neurologic sequelae, 24 (44%) showed psychiatric abnormalities, and 34 (63%) had mental deficit. However, the hospital in which the patients were seen cared for more severe JE cases who were referred by rural clinics. Because mild cases were treated at the rural clinics, the cases seen at the urban hospital may not have been representative. Goto32 reported neurologic sequelae in 76% of 43 JE cases at 5 years of follow-up. This rate was also an overestimate because all 43 patients had sequelae 1 month after the disease diagnosis; patients without sequelae at the 1-month point were not included for follow-up.
In 1957, Pieper and Kurland33 followed the survivors of a 1947 JE outbreak in Guam and found that 40% had neurologic sequelae; 11% were considered severe. However, in that study, only 13 JE patients were followed; another 14 JE patients had mumps encephalitis concurrent with JE infection. In addition, 11 patients were considered normal on the basis of correspondence without neuro-psychological evaluations.
In summary, these clinical reports disclose a wide range of outcomes. These may be attributed to design deficiencies, including a relatively short follow-up after JE diagnosis, small sample size, absence of a suitable comparison group, and selection bias.
Laboratory diagnosis of JE. CF antibody appears 1–2 weeks after the onset of JE; therefore, tests of both acute and convalescent sera are needed as the evidence for JE diagnosis. We cannot rule out JE from the non-JE encephalitis group since 8% of the CF-based non-JE diagnoses were based on negative test results from a single specimen. However, the negative CF results were obtained from both acute and convalescent specimens from the majority of non-JE cases. Therefore, the possibility of misclassification of JE as non-JE was low and this misclassification would have only minimized the differences in the outcomes of disability between JE and non-JE encephalitis groups and thus made the conclusions more conservative.
Positive HIA results could represent previous JE infection and could be false for determination of new JE infection. Among true JE patients, however, 80–100% had HIA antibody titers 1:80.18,19 This cut-off titer was found in only 2% of rural healthy residents in Shanghai in the 1970s.19 Only 3% of our study subjects were diagnosed as JE or non-JE by HIA test. IgM capture ELISA is sensitive and specific for JE diagnosis; however, only 4% of our cases were diagnosed as JE or non-JE by the presence of IgM anti-JE in the sera because this test was introduced in Shanghai in 1990s.
Potential limitations.
Despite efforts to capture both severe and mild subjects by home visits, we were able to follow only 48% of the discharged patients. Patients lost from the follow-up because they had changed their addresses during the recent 20 years. We tried to measure the bias because of the dropout by comparing the characteristics of those followed with those lost to follow-up and could not find any statistically significant difference (Table 1). To control for information bias in this observational study, we assembled a non-JE encephalitis group as a control for the JE group. Both groups were followed for neurologic and mental deficits in a double-blind fashion. Both groups had similar social-demographic characteristics and similar proportions of dropout, yet exhibited differences in the frequency and severity of neurologic abnormality (Tables 1 and 2). Undoubtedly, non-JE encephalitis patients could also have permanent sequelae.33–37 The etiology of non-JE encephalitis, however, remains unknown.
Of the groups that we reviewed, the JE patients had longer follow-up (17.2 years) than the non-JE encephalitis patients (14.5 years). The difference in length of follow-up alone may have led to the finding of more neurologic deficits for JE than for non-JE patients. Also, more JE patients than non-JE encephalitis patients lived in suburban areas. To adjust for these variables, we used logistic regression model to estimate the odds ratio of the long-term disability associated with JE; and the adjusted odds ratio for the presence of at least one neurologic deficit in the JE versus non-JE groups, was 6.30, and it was 20.72 for a low IQ. On the other hand, however, there were too few seizures and deaths occurring as new events during the JE period for a meaningful statistical analysis.
Another limitation of this study is that the disability reported here may have been incompletely ascertained; further neurologic dysfunction may occur later in subjects who were considered normal during this study period.
Latency of deficits. We found that seizures and optic atrophy appeared 3–17 years after recovery from the acute illness. This indicates the potential for considerable latency in the onset of JE-associated neurologic dysfunction. Generalized tonic-clonic seizure is seen mostly after acute encephalitis and can be difficult to treat and control.36 Two post-discharge patients died of drowning during generalized tonic-clonic seizure. Therefore, unconsciousness caused by generalized tonic-clonic seizure could indirectly increase mortality. In this study, the time needed for intelligence decline could not be ascertained because of the absence of repeated examinations of cognitive functions after hospital discharge. In addition, mild mental retardation is usually unnoticeable until school age.
Public health implications. By any standard, JE is a major public health problem that can potentially be controlled by proven effective vaccines. However, policymakers in most Asian countries do not consider immunization against JE a priority in their public health programs. Our study reveals a high frequency of JE disability and discloses a far greater burden of long-term disability caused by JE than has been appreciated previously. Our results thereby support the case for introducing JE vaccine into countries in which JE is endemic but without a JE immunization program.
Received April 21, 2006. Accepted for publication March 28, 2007.
Acknowledgments: The authors thank Dr. Zhu Qi-Rong, Children’s Hospital, Fudan University, Shanghai, China, and Dr. Shen Bing and Dr. Xu Rui-Fang, Jing An and Qing Pu District Center of Disease Control, Shanghai, China for their great support.
Financial support: This project was supported financially by the Bill and Melinda Gates Children’s Vaccine Program at the Program for Appropriate Technology in Health, Seattle, WA (contract number 00-GAT.770-790-01149-LPS). IVI is supported by the governments of the Republic of Korea, Sweden, and Kuwait.
Disclaimer: The authors do not have a commercial or other association that might pose a conflict of interest. The sponsor of this study played no role in the design, execution, and analysis of the study, and was not involved in the decision to submit the paper for publication.
* Address correspondence to Zhi-Yi Xu, Division of Translational Research, International Vaccine Institute, SNU Research Park, San 4-8, Bongcheon 7-dong, Kwanak-gu, Seoul 151-818, Republic of Korea. E-mail: xuzhiyi{at}ivi.int 
Authors’ addresses: Ding Ding and Zhen Hong, Institute of Neurology, Huashan Hospital, Fudan University, 12 Wulumuqi Zhong Road, Shanghai, China 200040, Telelphone and Fax: 86-21-62483421, E-mails: dding99{at}yahoo.com and zhong{at}shmu.edu.cn. Shou-jun Zhao, Department of Neurology, University of California San Francisco, 505 Parnassus Ave., Box 0114, San Francisco, CA 94143-0114, Telephone: 1-415-476-1488, Fax: 1-415-476-3428, E-mail: zhao538{at}itsa.ucsf.edu. John D. Clemens, Zhi-yi Xu, and Wei Liu, International Vaccine Institute, SNU Research Park, San 4-8, Bongcheon 7-dong, Kwanak-gu, Seoul, Republic of Korea 151-818, Telephone: 82-2-871-2801, Fax: 82-2-871-2803, E-mails: jclemens{at}ivi.int, xuzhiyi{at}ivi.int, and wliu{at}ivi.int. Bin Zhou, Bei Wang, Mao-sheng Huang, Jun Zeng, and Qi-hao Guo, Institute of Neurology, Fudan University, 12 Wu-lumuqi Zhong Road, Shanghai, China 200040, Telelphone and Fax: 86-21-62483421, E-mails: zhoubinjx{at}hotmail.com, huangquewb{at}163.com, mshuang{at}hotmail.com, junzeng{at}hotmail.com, and dr.guiqihao{at}126.com. Fang-biao Tao, School of Public Health, An-hui Medical University, Meishan Rd. Hefei, China 230032, Telephone and Fax: 86-551-5161179, E-mail: taofangbiao{at}yahoo.com.
Reprint requests: Zhi-yi Xu, International Vaccine Institute, SNU Research Park, San 4-8, Bongcheon 7-dong, Kwanak-gu, Seoul, Republic of Korea 151-818, E-mail: xuzhiyi{at}ivi.int.
|
REFERENCES |
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  S. Ompusunggu, S. L. Hills, M. S. Maha, V. A. Moniaga, N. K. Susilarini, A. Widjaya, A. Sasmito, A. Suwandono, E. R. Sedyaningsih, and J. A. Jacobson Confirmation of Japanese Encephalitis as an Endemic Human Disease Through Sentinel Surveillance in Indonesia Am J Trop Med Hyg, December 1, 2008; 79(6): 963 - 970. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
