Epidemiology of Travel-Associated and Autochthonous Hepatitis A in Austrian Children, 1998 to 2005

Authors

  • Pamela Rendi-Wagner MD, DTM&H, MSc,

    Corresponding author
    1. Department of Specific Prophylaxis and Tropical Medicine, Centre for Physiology and Pathophysiology, Medical University Vienna, Vienna, Austria
    Search for more papers by this author
  • Maria Korinek MD,

    1. Department of Specific Prophylaxis and Tropical Medicine, Centre for Physiology and Pathophysiology, Medical University Vienna, Vienna, Austria
    Search for more papers by this author
  • Andrea Mikolasek,

    1. Department of Specific Prophylaxis and Tropical Medicine, Centre for Physiology and Pathophysiology, Medical University Vienna, Vienna, Austria
    Search for more papers by this author
  • Andreas Vécsei MD,

    1. St Anna Kinderspital, Vienna, Austria
    Search for more papers by this author
  • Herwig Kollaritsch MD

    1. Department of Specific Prophylaxis and Tropical Medicine, Centre for Physiology and Pathophysiology, Medical University Vienna, Vienna, Austria
    Search for more papers by this author

Pamela Rendi-Wagner, MD, DTM&H, MSc, Department of Specific Prophylaxis and Tropical Medicine, Centre for Physiology and Pathophysiology, Medical University Vienna, Kinderspitalgasse 15, A-1090 Vienna, Austria. E-mail: pamela.rendi-wagner@meduniwien.ac.at

Abstract

Background In Austria, being an area of low hepatitis A endemicity, every year, several cases of this infectious disease are reported. The aim of the present study was to provide data on disease and hospitalization of children below the age of 15 for imported and autochthonous hepatitis A in Austria.

Methods Nationwide, active, hospital-based surveillance during the period 1998 to 2005.

Results During this 8-year observation period, 413 children below 15 years of age were hospitalized with acute hepatitis due to infection with hepatitis A . The mean annual incidence of hospitalization per 100,000 population was 3.8, with a decreasing trend from 1998 to 2005. The mean length of hospital stay attributable to hepatitis A was 6.5 days. The mean annual number of days of hospitalization attributable to acute hepatitis A infection in children below 15 years of age was 335 days. Information on origin of infection was available in 48% of the reports, the majority of which (69%) were in consequence of infection import. The mean annual incidence of travel-associated, hospitalized hepatitis A cases was 1.3 per 100,000, showing a lesser decrease rate over the observation period than the total hospitalization incidence.

Conclusions In an area of low hepatitis A endemicity such as Austria, hospitalization incidence of children is still at a considerable level. Our findings contribute to an open discussion about universal childhood vaccination.

Incidence rates of hepatitis A, causing worldwide approximately 1.5 million hepatitis A cases per year,1,2 are closely related to standards of hygiene and sanitation. However, in most Western European countries, in well-defined areas of low hepatitis A endemicity, several infections are reported every year, most of which can be traced back to import by travelers, refugees, or immigrants who visited their friends or families at home in countries highly endemic for infection with hepatitis A virus (HAV).3–5

Among persons at highest risk are travelers to endemic regions, persons working in institutions for immigrants and refugees, and, maybe the most important, children who are not aware of sanitary aspects of this infectious disease.6–9 In children, hepatitis A infections often occur asymptomatic, thus facilitating the distribution of HAV.10–12 Universal hepatitis A vaccination of small children can lower incidence of hepatitis A not only in that age-group but also in older children and adults, which can be considered as a herd effect.13 However, in most European countries, the implementation of a universal hepatitis A vaccination program is still under debate.

Up to now, reliable estimates of the incidence of hospitalized hepatitis A cases are available only for a few countries,14 and so far, no such data exist for Austria. We aimed to provide the first nationwide, population-based data on rates of hepatitis A–associated hospitalization in children in Austria.

Methods

Study design

The study is based on an 8-year surveillance of hepatitis A infection in hospitalized Austrian children up to the age of 15 years. The observation period was January 1, 1998, through December 31, 2005. The mean population of children below 15 years of age in Austria declined from 1,388,452 in 1998 to 1,317,707 in 2005.15,16

The data originate from a nationwide, voluntary, hospital-based surveillance network. In the period from 1998 to 2005, the proportion of reporting hospitals as measured by pediatric hospital beds varied between 42.4 (2002) and 100% of pediatric beds (1999).

Reports of hospital admissions due to hepatitis A infection were collected four times a year via standardized questionnaires. Cases were eligible for evaluation if the patient had been admitted to a pediatric hospital and if hepatitis A infection was confirmed serologically [anti-HAV immunoglobulin M (IgM) positive]. The reporting physicians were requested to provide the following patient’s data to the central study notification center at the Department of Specific Prophylaxis and Tropical Medicine, Medical University, Vienna: age, gender, date of onset of symptoms, date of hospitalization, date of hospital discharge, confirmation of positive clinical diagnosis, confirmation of positive hepatitis A serology, and, if known, origin of infection. Hepatitis A serology was performed by commercially available automated enzyme immunoassays of antibodies against HAV according to manufacturer’s instructions (AXSYM HAVAB-M; Abbott, Abbott Park, IL, USA; Cobas Core Anti-HAV IgM EIA; Roche, Basel, Switzerland) or by immunoluminometric capture assay (LIAISON HAV IgM; DiaSorin, Saluggia, Italy) at the local hospital-based laboratories. Only laboratory-confirmed (anti-HAV IgM positive) cases were included into case analysis.

In the case of incompleteness of reports, active surveillance by personal contact to the designated reporting persons at either site was employed on a regular basis. To maintain confidentiality, only anonymous data were used for further analysis.

Statistical analysis

We examined the annual average number of cases (by sex, season, and age-group) and average length of stay in hospital. The disease burden of hepatitis A and the trends of hepatitis A–associated hospitalizations were estimated using the weighted number of hospitalizations extrapolated from the number of hospitalizations reported by a sample of hospitals. Weighting was done for every single year, the weight given as the number of beds within every federal state divided by the number of beds from hospitals reporting hepatitis A cases per analyzed year. Annual incidence rates (per 100,000 population) were based on determination of the number of children of the respective age-group at risk during the surveillance period.15

Results

During the 8-year study period, a total number of 413 children below 15 years of age with acute hepatitis A infection were admitted to hospitals (males 53% and females 47%) in Austria, resulting in an average number of 51 hospitalized children per year of observation.

The mean annual incidence for the age-group below 15 years was 3.8 per 100,000 population. The highest incidence rate was observed in 1999, when up to 6.2 per 100,000 children were hospitalized. In the year 2005, this incidence dropped to 1.5 children per 100,000 population (Table 1). Incidence rates of all hospitalized cases versus imported hepatitis A cases per year of observation are shown in Figure 1.

Table 1.  Annual incidence rate per 100,000 population (<15 y), mean age of patients (±SD) in years, mean annual number of days of hospitalization (±SD), and mean duration of illness (±SD) in days per year of observation
YearIncidence (per 100,000 population)Mean age ± SD (y)Duration of hospitalization ± SD (d)Duration of illness ± SD (d)
19984.317.33 ± 3.306.64 ± 53.4810.24 ± 5.81
19996.178.18 ± 3.546.47 ± 5.209.91 ± 6.89
20005.668.07 ± 3.755.97 ± 3.3112.21 ± 11.36
20012.786.37 ± 2.866.75 ± 5.7511.26 ± 8.74
20023.337.18 ± 3.399.50 ± 12.1810.50 ± 5.08
20032.187.60 ± 3.785.18 ± 2.9312.73 ± 11.30
20044.457.13 ± 3.904.60 ± 2.2117.84 ± 20.45
20051.556.81 ± 3.487.00 ± 5.4012.13 ± 11.43
Total3.817.56 ± 3.566.47 ± 5.4011.5 ± 10.00
Figure 1.

Annual incidence of hospitalized cases of hepatitis A per 100,000 population (<15 y), 1998 to 2005 (gray bars—incidence of all cases; black bars—incidence of imported cases; lines indicating the linear trend over the observation period—all cases: y =−0.4501x + 5.8303, R2= 0.455; imported cases: y =−0.129x + 1.8901, R2= 0.3464).

Seasonal peaks of hepatitis A occurrence were observed annually at all reporting sites, with a maximum of cases notified in September and October (47% of cases) and a second smaller peak in March (8%).

The mean age of the hospitalized children was 7.6 years (SD 3.56, median 7.47 y), observing a minor decrease from 8.2 years in 1999 to 6.3 years in 2001. From 2002 to 2004, the average age was between 7.1 and 7.6 years. In 2005, children were youngest, with an average age of 6.8 years (Table 1). In general, we observed an average peak incidence between 4 and 10 years of age, with about 58.4% of all reported cases, as shown in Figure 2.

Figure 2.

Age distribution of hospitalized cases with acute hepatitis A disease.

During the whole observation period, neither clinical complications nor deaths were reported by any of the network hospitals.

The mean duration of stay in hospital for children admitted for acute hepatitis A disease was 6.5 days (SD 5.4). Comparing single years of observation, the shortest stays in hospitals occurred in 2003 with 5.2 days (SD 2.93) and in 2004 with 4.6 days (SD 2.21). The longest duration of hospitalization occurred in the years 2002 with 9.5 days (SD 12.18) (Table 1). The mean annual number of days of hospitalization attributable to acute hepatitis A infection in Austrian children below 15 years of age was calculated at 334.5 days, resulting in a total of 2,676 days for the whole observation period. No age-related trend was observed.

The mean duration of hepatitis A–associated clinical symptoms was 11.5 days (SD 10.0), resulting in a total of 591.3 days of illness in children below 15 years of age for the 8-year observation period equivalent to an average number of 591.3 days per year (Table 1). No age-specific trend was observed.

Information on origin of infection was available in 47.7% of all reported cases (for the rest, the origin of infection was unknown). A proportion of 30.5% of these infections was acquired during home visits and holidays in Turkey and 15.9% during holidays or home visits in former Yugoslavia. Only 20.7% of these reports could be traced back to be caused by household contacts with close physical contact. Exposure at school, kindergartens, and playgrounds led to acute hepatitis A infection in 9.2% of the cases. All other cases with indicated origin of infection were associated with stays abroad or immigration, except one case, which was traced back to illicit drug abuse (Table 2). The mean annual incidence of travel-associated hospitalized hepatitis A cases was 1.3 per 100,000, exhibiting a lesser decrease rate over the observation period than the total incidence as shown in Figure 1. The highest incidence of travel-related hepatitis A disease was observed in 1999 (2.8 per 100,000) and the lowest in 2003 (0.7 per 100,000).

Table 2.  Origin of infection of hepatitis A in 164 hospitalized cases below 15 years of age
Origin of infectionTotal number%
Turkey5030.49
Household contact3420.73
Former Yugoslavia2615.85
Kindergarten, school, and playground159.15
Romania106.10
Holiday with no specification106.10
Africa (Angola, Congo, Egypt, Ethiopia, Nicaragua)63.66
Asia (India, Afghanistan)53.05
Other (Ukraine, Israel, Dominican Republic)53.05
Italy21.22
Illicit drug abuse10.61
Total164100

Discussion

On the basis of a surveillance system, we have generated information on the epidemiology of hepatitis A among Austrian hospital inpatients up to 15 years of age. This study represents the first systematic investigation of this kind since introduction of active immunization in Austria, including only laboratory-confirmed hepatitis A cases, for the whole country.

During the 8-year study period, the incidence of hospitalized children due to hepatitis A infection declined from 6 per 100,000 in 1999 to 1.8 in 2005—on average, 3.7 per 100,000 children below 15 years of age were hospitalized per year. The overall decreasing trend of hepatitis A diseases is most probably attributable to sanitation and increasing hepatitis A travel vaccination.

However, regarding the fact that active hepatitis A vaccination has been available for more than 10 years now, these numbers still seem rather high. In the United States, the incidence of clinical hepatitis A infection in the total population varies between more than 20 cases per 100,000 population in South and Southwestern states and less than 5 in the East.17 In Germany, an incidence of 2.3 per 100,000 population has been observed in 2004.18 The given numbers, however, reflect detection of hepatitis A in the total population, whereas the estimates of the present study are restricted to inpatients below 15 years of age. Therefore, data from the United States and Germany cannot directly be compared to the incidence rates observed by the present study. Moreover, diverse reporting systems as well as differences in hospital admission policies may be considered. Since this represents only a small part of the total population, it may be speculated that the total incidence in Austria might be higher compared to other industrialized countries. This may at least partly be explained by the geographical closeness to East and South-East Europe and migration from these countries, considered at moderate to high risk.19

In 1994, Prodinger and colleagues studied seroprevalence rate of hepatitis A in western parts of Austria. At that time, they had already observed a changing epidemiological pattern with decreasing prevalence of HAV antibodies in young individuals.20 About 20 years ago, a Swedish study reported very low levels of hepatitis A infection in 2 per 100,000 population, with infections mainly concerning children of parents of foreign extraction.21 In many low and intermediate endemic countries, an increasing trend for adolescents susceptible to hepatitis A infection combined with a consecutive epidemiological shift of hepatitis A disease in adulthood and elderly can be observed.8,22,23 According to a recent evaluation by Martin and Lemon, the disease burden of hepatitis A is increasing due to more infections in older age-groups.24

Underreporting of clinical hepatitis A cases, because of a lack of typical clinical symptoms, such as jaundice, may be higher for younger children and adults than for older ones, contributing to a considerable degree of underreporting in younger age-groups. Moreover, since case detection in the present study was limited to hospital admissions, the data are likely to reflect the incidence of severe disease. Undoubtedly, hospital admissions represent only the smaller part of hepatitis A disease, at least in this particular age-group, as most of these patients may often remain undiagnosed at all due to the atypical manner of clinical symptoms.23,25–27

Our data clearly show, beside the fact that Austria is low-endemic for hepatitis A, that most cases of hepatitis A infection are “imported” either by tourists or by immigrants visiting friends and relatives (VFRs) at home in countries, such as Turkey, with intermediate to high hepatitis A endemicity.

The problem of import of infection by VFRs has been described previously.3,28,29 In Austria, about 70% of all hospitalized hepatitis A infections in children with known origin of infection can be traced back to import. Another fact, which fortifies the direct or the indirect origin of infections from abroad, is that reports show a seasonal occurrence, correlating with the high holiday seasons in Austria: most infections reported occurred 4 to 6 weeks following main tourist seasons. Destinations most frequently associated with HAV infection import are Turkey, former Yugoslavia, and Romania, all of which Austria has high passenger traffic with. According to Duval and colleagues,30 the risk of infection is increasing with birth in countries highly endemic for hepatitis A and travel to such countries. The burden of travel-related hepatitis A infection is also significant with respect to public health. An important aspect in the occurrence of hepatitis A is secondary infection. Children, to some extent because they remain asymptomatic, pose an underestimated public health problem related to imported hepatitis A and often pose the unrecognized source of an outbreak, mostly observed in community institutions such as kindergartens and primary schools. Moreover, children certainly practice less efficient sanitary habits, thus contributing more easily to transmission of infection. Indeed, one has to admit that, according to Steffen, the risk of HAV infection in nonimmune travelers has decreased during the past two decades by a factor greater than or equal to 10, presently being 0.1 to 1.0 per 1,000 per trip to an endemic destination.5 However, the risk may cumulate by repeated exposure as mostly observed in VFRs. Moreover, as clearly documented by the European Airport Survey, hepatitis A travel vaccination, as recommended by most expert authorities, is mostly ignored when traveling to “nearby” destinations such as Turkey.31 Interestingly, according to our observation, the decreasing trend for import–associated hepatitis A disease was much less than for the total numbers of hepatitis A hospitalizations. This indicates that contrary to the overall decreasing trend of hepatitis A infection, current preventive strategies appear to have only negligible impact on the frequency of mobility-associated disease.

The average length of stay in hospital was 6 to 7 days, longer than in other industrialized countries, which have reported a range of 2 to 4 days of duration of hospitalization.17 These variations may partly be due to differences in local hospitalization policies. In total, in Austria, about 25 days of hepatitis A–associated hospitalization per 100,000 children per year can be estimated. According to Nwachuku and Gerba, $86,899 per case of hospitalization for a child with hepatitis A infection arise for healthcare systems.9 Duration of clinical symptoms was reported at a range of 11 to 12 days, resulting in an average number of 591 days of illness per year in children below 15 years of age.

According to the age distribution, the majority of the hospitalized children were older than 3 years, clearly indicating community-acquired infection through kindergarten (usually starting in Austria at the age of 3) or school.

The present data reflect that children are at risk of HAV infection and point out the importance of vaccination against hepatitis A, particularly for children before they enter community institutions as asymptomatic, but infected children play a major role in the distribution of HAV.6 Assumptions for lifetime protection against this disease following active immunization fortify the need for early immunization against hepatitis A.

Following the consensus statement of The International Consensus Group on Hepatitis A Virus Immunity, there is no evidence to support HAV booster vaccination after a complete primary vaccination course in healthy individuals, thus making vaccination even more attractive.32

In conclusion, in an area of low hepatitis A endemicity such as Austria, where active vaccination has been available for more than 10 years now, hospitalization incidence of children is still at a considerable level. Our findings should therefore contribute to an open discussion about universal childhood vaccination.

Acknowledgments

We thank all pediatricians and the staff of the wards from all participating hospitals for their continuing support. The skillful technical assistance of Dr Michael Hofer is gratefully acknowledged. This study was supported by a grant of the Research Foundation of the Austrian National Bank (Grant 6165), the Centre for Travel Medicine, Vienna, and by the Austrian Green Cross for Preventive Medicine.

Declaration of Interests

The authors state that they have no conflicts of interest.

Ancillary