Etiology of Encephalitis in Australia, 1990–2007 | CDC EID

Volume 15, Number 9–September 2009
Research
Etiology of Encephalitis in Australia, 1990–2007
Clare Huppatz, David N. Durrheim, Christopher Levi, Craig Dalton, David Williams, Mark S. Clements, and Paul M. Kelly
Author affiliations: Hunter New England Population Health, Newcastle, New South Wales, Australia (C. Huppatz, D.N. Durrheim, C. Dalton); Australian National University, Canberra, Australian Capital Territory, Australia (M.S. Clements, P.M. Kelly); and John Hunter Hospital, New Lambton, New South Wales, Australia (C. Levi, D. Williams)

Suggested citation for this article

Abstract
Encephalitis is a clinical syndrome commonly caused by emerging pathogens, which are not under surveillance in Australia. We reviewed rates of hospitalization for patients with encephalitis in Australia's most populous state, New South Wales, from January 1990 through December 2007. Encephalitis was the primary discharge diagnosis for 5,926 hospital admissions; average annual hospitalization rate was 5.2/100,000 population. The most commonly identified pathogen was herpes simplex virus (n = 763, 12.9%). Toxoplasma encephalitis and subacute sclerosing panencephalitis showed notable declines. The average annual encephalitis case-fatality rate (4.6%) and the proportion of patients hospitalized with encephalitis with no identified pathogen (69.8%, range 61.5%–78.7%) were stable during the study period. The nonnotifiable status of encephalitis in Australia and the high proportion of this disease with no known etiology may conceal emergence of novel pathogens. Unexplained encephalitis should be investigated, and encephalitis hospitalizations should be subject to statutory notification in Australia.

Many novel infectious diseases have been reported since 1940; most have been zoonoses originating from wildlife (1). Several novel zoonotic viruses, including Hendra virus, Australian bat lyssavirus, and Nipah virus, have resulted in encephalitic illness in humans (2–4). Australia has witnessed the emergence of several zoonotic and arboviral pathogens associated with encephalitis; these pathogens have been either novel pathogens or pathogens appearing in new geographic locations and include Hendra virus, Murray Valley encephalitis virus, Australian bat lyssavirus, and Kunjin and Japanese encephalitis viruses (5–7). The appearance of these emerging pathogens that can result in encephalitis raises questions about the etiology of encephalitis in the Australian population and about the adequacy of surveillance for novel pathogens.

Encephalitis is an inflammatory process in the brain parenchyma and is associated with clinical evidence of brain dysfunction (8). An infectious etiology of encephalitis is usually suspected in a patient with fever, headache, and signs of diffuse brain dysfunction, often with focal neurologic signs (9,10). Encephalitis generally results in a serious illness requiring hospitalization. Severe encephalitic illness can lead to death, and survivors frequently experience ongoing neurologic sequelae (9,10).

Presumably, the most common etiologies of encephalitis are infectious (11), and viral pathogens account for most diagnosed cases. In developed countries, the most commonly identified pathogens associated with acute encephalitis are the herpes viruses (9,12–15). Herpes simplex encephalitis is believed to account for 10%–20% of cases (12), but pathogen identification is not always possible (16).

Several countries have conducted large epidemiologic studies to assess the impact of disease caused by encephalitis and to determine its etiology. Davison et al. reviewed UK hospital records for a 9-year period (1989–1998) and found that the hospitalization rate for viral encephalitis in UK hospitals was 1.5 cases per 100,000 population; 60% of cases were recorded as unidentified viral infection (17). During the study period in the United Kingdom, 419 deaths were attributed to viral encephalitis (overall case-fatality rate 6.5/100 cases), but an etiologic organism was identified for only 50% of these deaths (17). A smaller prospective study conducted in Finland for a 25-year period (1967–1991) found no etiology for encephalitis in 49% of 322 patients hospitalized with this illness (18).

Khetsuriani et al. analyzed national data for the United States over a similar period (1988–1997) to estimate the impact of both viral and nonviral encephalitis hospitalizations (11). This study found that the hospitalization rate for encephalitis in the United States was 7.3 per 100,000 population, and no specific etiology was identified in 59.5% of cases. During the study period, a case-fatality rate of 7.4 per 100 cases was recorded. For those persons admitted with encephalitis for which an etiology was identified, most specified a viral etiology (11). Also in the United States, the California Encephalitis Study, a large prospective study, found that 63% of hospitalized patients with encephalitis from 1998 through 2005 had encephalitis of unknown etiology, despite extensive laboratory testing (19). In another US study of persons whose deaths were associated with encephalitis, 81.5%–86.2% of the deaths resulted from encephalitis with an unknown etiology (20).

In Australia, encephalitis, as a syndrome, is not a notifiable disease, and data about trends or even clusters of this disease are not routinely collected. Only laboratory-confirmed encephalitis cases due to certain pathogens (e.g., Murray Valley encephalitis virus, Japanese encephalitis virus, or Australian bat lyssavirus) are notifiable to public health units; thus, the occurrence of encephalitis is not well documented.

A single, small study conducted in Australia's tropical Northern Territory found that 18 (53%) of 34 encephalitis patients admitted to the Royal Darwin Hospital during a 5-year period (1992–1996) had encephalitis with unexplained etiology (21). This level of unknown pathogen etiology suggests that Australia has a similar rate of pathogen identification as that reported in the overseas studies. However, the relative importance of different pathogens may differ because Australia has unique pathogens such as Murray Valley encephalitis virus, Hendra virus, and Australian bat lyssavirus) (5).

In this study, we examine the impact of hospitalizations due to encephalitis in New South Wales (NSW), Australia's most populous state. We also describe trends in pathogen identification in patients hospitalized with encephalitis over the 18-year period 1990–2007.

Methods
Data Sources
Data on hospital discharges, deaths, and population for NSW were obtained for 1990–2007 from the Health Outcomes and Information Statistical Toolkit, a collection of databases maintained by the Epidemiology and Surveillance Branch of the NSW Department of Health. The datasets used were from the Inpatient Statistics Collection library. Encephalitis-associated hospital stays were extracted for the period for patients for whom data were complete by using International Classification of Diseases, 9th revision, Clinical Modification codes (ICD-9-CM, Jan 1990–Jun 1998) and International Classification of Diseases, 10th revision, Australian Modification (ICD-10-AM, July 1998–Dec 2007).

Data for encephalitis-associated deaths in NSW were extracted from the Australian Bureau of Statistics death library by using ICD-9 (1990–1997) and ICD-10 (1998–2006) codes. Data for deaths occurring in 2007 were unavailable at the time of extraction. Population statistics from the Australian Bureau of Statistics were accessed through the Health Outcomes and Information Statistical Toolkit for the same period.

Definitions
An encephalitis-associated hospital stay was defined as a hospitalization for which the primary discharge diagnosis was an ICD-9-CM or ICD-10-AM code for acute encephalitis. Relevant ICD-9-CM and ICD-10-AM codes were obtained by searching the alphabetical list of ICD codes for all codes that include the term enceph, but excluding nonencephalitis conditions (e.g., anencephaly) and including rabies. Additionally, 2 ICD-9-CM codes used for encephalitis-related conditions before July 1998 did not include the prefix enceph-, unlike nomenclatures used in ICD-10-AM encephalitis codes. One of these ICD-9-CM codes was 49.8, defined as other non–arthropod-borne viral diseases of central nervous system –other specified non–arthropod-borne viral diseases of the central nervous system. The other code was 49.9, defined as other non–arthropod-borne viral diseases of central nervous system –unspecified non–arthropod-borne viral diseases of the central nervous system. The use of these codes was found by comparing data extracted for 1998 and 1999. For these 2 years, coding was performed in both ICD-9-CM and ICD-10-AM, and data were then compared to identify the additional codes.

The Table shows the ICD-9-CM and ICD-10-AM encephalitis-associated conditions with codes used for encephalitis hospitalizations during 1990–2007. These codes were further classified by investigators into encephalitis with known pathogens or unknown pathogens (Table).

Data Analysis and Ethics
Encephalitis-associated hospitalizations were analyzed by using SAS version 8.0 (SAS Institute, Inc., Cary, NC, USA) by etiologic category, year, age, gender, and hospital location. Hospitalization rates and death rates were calculated by using annualized population estimates from the Australian Bureau of Statistics. Negative binomial regression was used to analyze trends for hospitalization rates of all encephalitis hospitalizations and for those caused by known and unknown pathogens. The regression was repeated with adjustment for age groups over time by using log of the population as an offset and age as a categorical covariate. Trends were represented as annual percentage changes in rates. Ethical clearance was given by Hunter New England Population Health and by the Australian National University Human Research Ethics Committee.

Results
Encephalitis Hospitalizations
From January 1990 through December 2007, encephalitis accounted for 5,926 hospitalizations in NSW. The average number of hospitalizations per year was 329 (range 281–393). The most frequently identified etiology of encephalitis was herpes simplex virus infection, which accounted for 763 admissions (12.9%) (Table). Varicella encephalitis (226 admissions or 3.8%) and Toxoplasma meningoencephalitis (221 admissions or 3.7%) were also common known etiologies. The etiology of 4,126 admissions (69.6% of total admissions) was unknown.

Figure 1

Figure 1. Encephalitis hospitalization rates by year and by known and unknown pathogen etiology, New South Wales, Australia, 1990–2007.


Figure 2

Figure 2. Average rates of encephalitis hospital admissions by 10-year age groups and by known and unknown pathogen etiology, New South Wales, Australia, 1990–2007.


Figure 3

Figure 3. Herpes encephalitis hospitalizations by year, New South Wales, Australia, 1990–2007.


Figure 4

Figure 4. Toxoplasma encephalitis hospitalizations by year, New South Wales, Australia, 1990–2007.


Figure 5

Figure 5. Subacute sclerosing panencephalitis hospitalizations by year, New South Wales, Australia, 1990–2007.

The average annual rate of encephalitis hospitalization was 5.2 per 100,000 population (range 4.2–6.7) (Figure 1). The annual rate for total encephalitis cases was higher for men (average rate 5.7/100,000) than for women (4.7/100,000).

The case-fatality rate for encephalitis during 1990–2006 was 4.6/100 cases. Figure 2 shows the average rate of hospitalization for patients with encephalitis by 10-year age groups during 1990–2007. The highest rates of admission occurred for age groups 0–9 years and >60 years.

Pathogen Identification
The rate of encephalitis hospitalizations declined for admissions of patients with encephalitis with both known and unknown etiologies (Figure 1). A negative binomial regression model showed a statistically significant decline of 1.4% per year (p = 0.0003, 95% confidence interval [CI] 0.7%–2.2%) for total encephalitis admissions during 1990–2007. This trend was similar for cases of encephalitis with unknown etiologies (decline of 1.4% per year, p = 0.0002, 95% CI 0.7%–2.2%) and for encephalitis cases with known etiologies, although the effect was not statistically significant for pathogens with known etiologies (decline of 1.4% per year, p = 0.13, 95% CI 0.4%–3.2%,). This trend did not change when adjusted for age groups.

The proportion of cases with pathogens of known etiologies was higher for men than for women; the average hospitalization rate for men with encephalitis with a known etiology was 1.9/100,000 compared with 1.3/100,000 for women with encephalitis with a known etiology. In contrast, rates for patients with encephalitis with unknown etiologies were similar for men (3.8/100,000) and women (3.4/100,000). When the diagnosis of Toxoplasma encephalitis was excluded, the rate for hospitalizations for men with encephalitis with identified etiologies (1.5/100,000) was similar to the rate for women (1.3/100,000).

The proportion of hospitalizations with encephalitis with known etiology varied little among hospitals; a notable exception was a large Sydney hospital specializing in HIV-related medicine. For this hospital, the higher proportion of patient admissions with encephalitis with known etiology (50%) resulted from a high number of Toxoplasma encephalitis admissions.

Known Causes
Herpes encephalitis accounted for a relatively stable proportion of encephalitis hospitalizations during the study period (Figure 3). Toxoplasma encephalitis hospitalizations increased early in the study period and peaked in 1993 (Figure 4). Few toxoplasmosis hospitalizations occurred during the last 10 years of the study period. Subacute sclerosing panencephalitis (SSPE) was the only other diagnosis that decreased most years after 1994. These decreases appeared to contribute to the overall downward trend in hospitalizations with encephalitis with a known etiology, although this trend was not statistically significant; SSPE hospitalizations declined throughout the 1990s (Figure 5).

abrir aquí para acceder al extenso documento CDC del cual editamos el 40%:
Etiology of Encephalitis in Australia, 1990–2007 | CDC EID