• Open Access

The burden of chronic hepatitis B virus infection in Australia, 2011

Authors

  • Jennifer H. MacLachlan,

    1. Epidemiology Unit, WHO Regional Reference Laboratory for Hepatitis B, Victorian Infectious Diseases Reference Laboratory; University of Melbourne, Victoria
    Search for more papers by this author
  • Nicole Allard,

    1. Epidemiology Unit, WHO Regional Reference Laboratory for Hepatitis B, Victorian Infectious Diseases Reference Laboratory; University of Melbourne, Victoria
    Search for more papers by this author
  • Vanessa Towell,

    1. National Policy and Education Division, Australasian Society for HIV Medicine, New South Wales
    Search for more papers by this author
  • Benjamin C. Cowie

    1. Epidemiology Unit, WHO Regional Reference Laboratory for Hepatitis B, Victorian Infectious Diseases Reference Laboratory; University of Melbourne, Victoria
    Search for more papers by this author

Correspondence to: Jennifer MacLachlan, Epidemiology Unit, Victorian Infectious Diseases Reference Laboratory, 10 Wreckyn Street, North Melbourne, Victoria 3051; e-mail: jennifer.maclachlan@mh.org.au

Abstract

Objective: The number of Australians living with chronic hepatitis B (CHB) is thought to be increasing, as are adverse outcomes including cirrhosis and liver cancer, however, robust, up-to-date estimates of this burden are limited. Contemporary estimates of the prevalence of CHB in Australia are essential to guide appropriate public health and clinical responses.

Methods: This study used census-based methodology attributing risk of CHB by country of birth and Aboriginal and Torres Strait Islander status, augmented with priority risk-group based estimates. Deterministic mathematical modelling was used for comparison and for validation of census-derived estimates.

Results: An estimated 218,000 Australians (plausible range 192,000–284,000) are living with CHB, a significant increase over previous estimates. The prevalence derived using mathematical modelling was similar, at 204,000. Notable differences were observed by geographic area in both prevalence and the populations predominantly affected. It is estimated that only 56% of people living with CHB in Australia have been diagnosed and notified.

Conclusions: The prevalence of CHB in Australia is increasing, with 1% of the population now estimated to be affected. The majority of the burden is experienced by people born overseas in endemic areas, with more than 95% of new cases of CHB entering the population through migration.

Implications: It is imperative that more attention and greater resources are devoted to addressing CHB in Australia; to increase the proportion of Australians affected who have been diagnosed and who are on treatment, in accordance with the First National Hepatitis B Strategy.

“The epidemiology of hepatitis B virus (HBV) infection in Australia is inadequately understood. Previous studies have defined groups within the population at higher risk of HBV infection … However, current estimates of HBV prevalence among these groups and the overall population in Australia are needed.”1

Since 2004, Australia has made significant progress in developing an improved understanding of the epidemiology and control of hepatitis B virus (HBV) infection. As a result, public health policy has also seen notable advances, best exemplified by the production and endorsement of Australia's First National Hepatitis B Strategy in 2010.2 However; robust, accurate data regarding the burden of this emerging national health priority remain scarce, restricting the evidence base on which to build healthy public policy.

Australia has made great progress in preventing incident HBV infection through infant and adolescent catch-up programs implemented within the National Immunisation Program in all States and Territories by the year 2000.3 However, our domestic vaccination program will not have any noticeable effect on the burden of chronic hepatitis B (CHB), the major determinant of which is migration from high and intermediate prevalence countries.2,4 It will also, therefore, have negligible impact on the burden of HBV-attributable liver cancer in Australia, now the fastest increasing cause of cancer death in Australians.5,6

As the 2011 Australian Census demonstrated, Australia's migration patterns are rapidly changing. Since 2006, immigration from China, India and the Philippines has increased by 54%, 101% and 42% respectively; migrants from the Asia Pacific region now make up one-in-three Australians born overseas.7,8 In 2010, China overtook England for the first time as the major source of permanent additions to the country.9

These demographic changes have particular relevance for estimating the burden of health issues that are more prevalent in some of Australia's overseas-born communities, such as CHB. Although nationally representative (albeit convenience sample) serosurveys remain the gold standard for measuring the prevalence of CHB infection in the community, the most recent estimates are derived from samples collected in 200210 (see Table 1). When considering the profound population changes that have occurred in the past 10 years, a population prevalence estimate based on residual sera collected a decade ago is unlikely to be an accurate representation of current burden.

Table 1. Population-based studies estimating CHB prevalence in Australia.
StudyStudy Type and sample yearsCHB prevalence range* (% HBsAg)Study estimate of number of people living with CHBEstimated number of people living with CHB in 2011
  1. *Ranges represent uncertainty in point prevalence estimates based on sample contamination/exhaustion

O'Sullivan et al, 20041National opportunistic serosurvey (1996–1999)0.49 / 0.87%91,500 / 163,000105,388 / 187,117
Gidding et al, 200710National opportunistic serosurvey (2002)0.7 / 0.8%137,640 / 157,300150,554 / 172,062
Cowie et al, 201025Victorian opportunistic serosurvey (1995, 2000 and 2005)1.1%211,200236,585

This study uses contemporary estimates of the number of people in each of the priority populations affected by CHB2,8,11,12 and of the prevalence of CHB in these groups,13–19 together with novel mathematical modelling approaches, to develop a picture of the burden of hepatitis B in Australia for this decade, and of where and by whom this burden is disproportionately borne. Estimating the true burden of CHB in Australia is crucial to effective service delivery, public health planning, identification of priority populations and evaluation of the efficacy of population-based interventions.

Methods

The number of people living with CHB in Australia in 2011 was calculated by estimating the prevalence of chronic infection in both the overall population and in identified higher-risk groups, multiplied by the estimated population in each of these categories.

Parameters used for CHB prevalence (defined as the proportion of the population positive for hepatitis B surface antigen, HBsAg) and population sizes of the given groups were derived from review of the literature, together with recently released data from the 2011 Census. Local estimates of CHB prevalence were derived for each Australian State and Territory using the 2011 Census data for country of birth and Aboriginal and Torres Strait Islander status. As estimates of the number of people who inject drugs and men who have sex with men were not available according to region, proportional figures were applied based on population for each State and Territory.

To derive an indication of the plausible range of the estimated number of people living with CHB, the uncertainty around estimates of prevalence for each of the groups was incorporated into an upper and lower estimate where available, as outlined below.

People born overseas

Country of birth information for the Australian population was obtained from the 2011 Australian Bureau of Statistics (ABS) Census of Population and Housing,8 in which almost 95% of respondents indicated their country of birth. Non-reporting of country of birth was conservatively assumed to be random; that is, those who did not report their country of birth were assumed to have the same country-of-birth distribution (and resultant CHB prevalence) as those who reported country of birth.

Estimates of CHB prevalence by country of birth were obtained from two sources: antenatal seroprevalence data collected from hospitals in Sydney's South-West Area Health Service (SSWAHS) from Australian women born in 27 countries;13 and a recent global systematic review14 used to estimate CHB in overseas-born people in the United States. The Australian-derived antenatal estimates were applied preferentially in all COB categories and were lower in all cases (except Taiwan) than the systematic review estimates. In the case of Zimbabwe, seroprevalence for South African born women in the SSWAHS study was applied, due to source country diversity in both CHB prevalence and likelihood of migration, which would lead to overestimation of the prevalence in migrants to Australia.13

The 27 countries for which Australian antenatal data was available represent 62% of Australia's overseas-born population, resulting in overall conservative estimates for CHB prevalence according to country. The midpoint between the prevalence estimates from these two sources was used as the upper limit for plausible range calculations.

Aboriginal and Torres Strait Islander people

The number of Australians who identify as being of Aboriginal and/or Torres Strait Islander descent was obtained from the 2011 census, and was found to be 548,366 people.

The prevalence of CHB in Aboriginal and Torres Strait Islander people was extrapolated from a 2011 review of studies of prevalence, with a midpoint estimate of 3.7% prevalence applied,19 with lower and upper estimates of 2% and 5% respectively, predominantly based on existing community CHB estimates1 and available antenatal screening data from the last decade.20,21

People who inject drugs

The number of Australians aged over 14 years who have ever injected drugs was obtained from the Australian Institute of Health and Welfare (AIHW) National Drug Strategy Household Survey (NDSHS) 2010, and was estimated to be 1.8%. 12 This proportion was applied to the 2011 Australian population aged over 14 years8 for a figure of 314,000 people having ever injected drugs.

The prevalence of CHB in Australians who inject drugs was estimated to be 4% (range 2.9–5%), taken from a global systematic review published in 2011.17

Men who have sex with men

Estimates of the number of men who have sex with men in Australia were derived from the Australian Study of Health and Relationships conducted in 2001–02 by the Australian Research Centre in Sex, Health and Society (ARCSHS), which involved telephone interviews with a representative sample of more than 19,000 Australian adults aged 16–59 years. This study found that 5% of men interviewed reported genital sexual contact with other men.11 This proportion was applied to the Australian male population aged 16–59 years in 20118 to obtain an estimate of 322,000 men who have sex with men.

Limited data are available regarding the prevalence of CHB in men who have sex with men. A conservative prevalence figure of 3% was used based on the three studies available, which found CHB prevalence of 3.3%, 15 5.7%16 and 3%, 18 with 5.7% used as the upper limit for the plausible range.

Non-Indigenous people born in Australia

The prevalence of CHB in Australians not of Aboriginal or Torres Strait Islander descent and not belonging to other risk groups such as those who inject drugs or men who have sex with men was estimated as 0.3%, in keeping with antenatal screening results.1,13 A range of 0.2–0.4% prevalence was used as the lower and upper bounds of this estimate.

Retrospective estimates of the number of people living in Australia with CHB were also calculated for both 2001 and 2006, based on historical Census7,22 data.

Mathematical modelling

A deterministic, dynamic mathematical model of hepatitis B in the Australian population from 1951 onwards was constructed.23 In summary, the Australian population over the duration of the model was divided into compartments, representing HBV infection status (Figure 1). Flow between these compartments occurs in response to HBV infection, or by vaccination. Separate age strata were modelled to account for differences in risk of infection, outcomes of infection, and uptake of vaccination; these strata were 0–4 years, 5–14 years, 15–44 years and 45 years and over.

Figure 1.

Schematic representation of modelled transition between HBV infection states. Blue boxes represent these states, and arrows represent the flow of individuals defined by the natural history of HBV infection, with the white boxes defining the respective transitions.

Entry into the population includes births and also age-specific net overseas migration by source country CHB prevalence; exit from the simulation is via HBV-attributable or all-cause deaths. The model was parameterised using a wide range of data sources (including Census, Births, Net Overseas Migration, Population Mortality and Migration Projections from the Australian Bureau of Statistics, published evidence regarding the natural history of hepatitis B by age group, surveillance notifications since 1971, source country seroprevalence estimates, and existing mathematical models in the international literature) to incorporate these, and other important factors including the ageing of the population, the variable natural history of CHB, and vaccination uptake and outcomes.4 A dynamic force of infection was used to reflect both the increased risk of transmission with rising prevalence of CHB in the population, and also the impact of herd immunity from vaccination programs. Assumptions including the force of infection, homogeneous mixing, migration projections, and the uptake of vaccination were subjected to sensitivity analysis. Model-derived estimates were compared with annual hepatitis B surveillance notifications to estimate the proportion of people living with CHB who have been diagnosed.

Data for the risk-group based estimates was handled in Microsoft Excel 2011, which was also used to produce graphs. The mathematical model was constructed using Berkeley Madonna v.8.3.14.

Results

The risk group method estimated that there are approximately 218,000 people living with chronic hepatitis B in Australia (1.02% population prevalence), the majority having been born overseas, particularly in the Asia Pacific region and in Sub-Saharan Africa (Table 2). Of those born in Australia, specific priority populations including Aboriginal and Torres Strait Islander people (9.3%), people who inject or have injected drugs (5.7%) and men who have sex with men (4.4%) represent the majority of the remaining people living with CHB.

Table 2. Estimates of the number of people living with CHB in Australia.
Group2011 populationPrevalence of CHB*Number of people living with CHBProportion of total
  1. * Prevalence figures for those born overseas are averages derived by number of people living with CHB divided by total population born in that region.

People born overseas 5,057,039 2.43% 122,694 56.1%
Asia/Pacific2,347,4663.55%83,30038.11%
Europe2,131,1471.03%22,03210.1%
Africa/Middle East578,4262.69%15,5657.12%
Americas223,6370.80%1,7960.82%
People born in Australia 15,021,553 0.56% 84,059 38.5%
Aboriginal and Torres Strait Islander people548,3663.7%20,2909.3%
People who inject drugs314,0134%12,5615.7%
Men who have sex with men322,6163%9,6784.4%
Other Australian-born non-Indigenous13,836,5590.3%41,53119.0%
Other/Not Stated 1,205,490 (Australian Average) 11,814 5.4%
TOTAL21,507,7191.02%218,567 

Results according to State and Territory are shown in Table 3. The burden of CHB is fairly evenly distributed, with slightly increased prevalence of CHB in New South Wales, Victoria, and the Northern Territory compared to the national average. However, the distribution of this burden among the identified priority groups in the population varied widely by region, reflecting local demographics. These results indicate that Aboriginal and Torres Strait Islander people made up over a third of the total number of people living with CHB in the Northern Territory, but 5–15% of people living with CHB in most other states, and less than 3% in Victoria. Two-thirds of Victorians living with CHB were born overseas, as were more than 50% of those with CHB in the other jurisdictions, with the exception of Tasmania – the only State in which non-Indigenous Australian-born individuals are estimated to represent the majority of people living with CHB.

Table 3. Estimated number of people living with CHB according to State or Territory of residence in the 2011 Census.
State/Territory2011 PopulationProportion of Australian populationNumber of people living with CHBPrevalence of CHBProportion of Australians with CHB
  1. *Total includes 59 people with CHB whose state was recorded as ‘other territory’ by the Census

Australian Capital Territory357,2191.7%3,6031.01%1.6%
New South Wales6,917,65532.2%77,0761.11%35.3%
Northern Territory211,9431.0%3,5561.68%1.6%
Queensland4,332,73820.1%37,4270.86%17.1%
South Australia1,596,5717.4%14,4420.90%6.6%
Tasmania495,3522.3%3,5130.71%1.6%
Victoria5,354,03924.9%56,8361.06%26.0%
Western Australia2,239,17210.4%22,0550.98%10.1%
AUSTRALIA TOTAL21,507,719 218,567*1.02% 

Sensitivity analyses indicated that, if using the lower range of all CHB prevalence estimates, the number of people living with CHB in Australia would be 191,965. Due to the varying availability of upper and lower estimates by group, this adjustment would affect the demographic distribution of the burden of CHB, reducing the proportion of people with CHB born in Australia to 30%, with those born overseas making up 64%.

Reciprocal calculations based on the upper estimates of these risk group ranges increased the estimated number of people living with CHB to 249,136; also including increased estimates of prevalence according to country of birth would add an additional 34,425 overseas-born people living with CHB, for a total upper estimate of 283,560 – approximately 60% having been born overseas.

Extrapolation of 2011 results using historical census data indicates that the prevalence of CHB estimated using this method is increasing, with an estimated 161,000 people living with CHB in 2001 (0.85% prevalence) and 186,100 in 2006 (0.94% prevalence).

Results of the mathematical modelling undertaken to validate the Census-based methodology produced an estimate of 204,000 Australians living with CHB at the end of 2011 (population prevalence 0.91%). When compared with cumulative notifications of CHB since 1971 the model-derived estimates suggest that only 56% of people living with CHB in Australia have been diagnosed and notified.

The model-based estimates also reinforce the importance of migration to the epidemiology of CHB in Australia. Over 95% of chronic hepatitis B infections entering the population annually are estimated to do so through migration from endemic countries, and not from domestic infections progressing to chronicity. The number of deaths attributable to CHB in Australia in 2011 was estimated to be 377 (plausible range 289–611).

Discussion

This study provides contemporary estimates of the prevalence and demographic distribution of CHB in Australia, based on the most recent seroprevalence estimates for all of Australia's identified priority populations.2 The estimated number of people living with CHB in 2011 is 218,000, for an overall population prevalence of 1.0%. When combined with estimated number of people living with chronic hepatitis C infection,24 the total number of Australians living with chronic viral hepatitis approaches 450,000.

Accounting for uncertainty in CHB prevalence and population size parameters, a plausible range for the number of people infected lies between 192,000 and 284,000. The deterministic mathematical model produced a comparable estimate within this range, of 204,000 people living with CHB in Australia at the end of 2011.

The results obtained here also highlight the key priority populations for CHB prevention and management, and the importance of locally tailored estimates to guide programmatic responses that will have the most impact. As shown in Table 2, the majority of people living with CHB in Australia were born overseas, mostly in the Asia-Pacific region. A substantial proportion of the non-Indigenous Australian born people living with CHB (estimated to represent 19% of all Australians chronically infected) are likely to be Australian-born children of migrants from endemic areas, particularly those born prior to the advent of antenatal screening and neonatal vaccination in the 1980s.3 Other Australian-born groups with higher prevalence include Aboriginal and Torres Strait Islander peoples, people who inject drugs, and men who have sex with men.

The estimates presented here are a substantial increase on the generally accepted figures of the number of Australians living with CHB, with extrapolation of serosurvey estimates from 1999 and 2002 indicating between 105,000–187,000 people living with CHB in 2011, based on a population prevalence of 0.49–0.8%. However, the results are comparable to a recently reported Victorian hepatitis B serosurvey that found a prevalence of 1.1%. 25

  • image(2)

[ Distribution of Australia's burden of chronic hepatitis B by priority population. ]

Extrapolation of the Census method to previous years also supports a substantial increase in CHB prevalence in Australia, with estimates based on the 2001 population finding a CHB prevalence of 0.84%– similar to that found in the 2002 nationwide serosurvey (0.7–0.8%). In addition, the mathematical modelling reported here also supports a rapid increase in the number of people living with CHB, with substantial changes from previous estimates26 due to increases in migration well in excess of the ABS projections based on the 2006 Census.27

The methodology used here has the advantage of being relatively simple and not resource intensive, using data that are available in the public domain. With repeated analysis it can be applied to monitoring temporal and geographic trends in CHB burden. This methodology also allows estimation of the relative proportion of the total burden of CHB experienced by different priority populations, which vary significantly by State and Territory within Australia (see Table 3 and Figure 3).

Figure 3.

Distribution of people living with CHB by state and population.

In contrast, the dynamic mathematical model constructed and presented for comparative purposes is complex and requires specialised techniques and software but accounts for population flows, attributable and all-cause mortality, the influence of vaccination on hepatitis B transmission, and other factors.23 That both approaches produced comparable results both validates the relatively simple census-based methodology for estimating the burden of CHB in a heterogeneous population like Australia, and reinforces the estimates of CHB prevalence for 2011 presented here.

Geographic specificity is a major advantage of the census-based methodology, given the availability of population denominator figures at all statistical divisions. This allows for estimation of CHB prevalence at national, state, and local levels, allowing targeted interventions in areas of greatest need, developed in partnership with identified local priority populations. The ability to set local health priorities is a key element of Australia's National Health Reform agenda.28

Estimates of population size are not available at a local level for certain priority populations (people who inject drugs and men who have sex with men) and so these populations are assumed to be evenly distributed according to state, which may not be the case.

There are a number of assumptions that potentially affect the validity of these estimates and could have led to over- or under-estimation of the true burden. Population CHB prevalence may be founded on limited or unrepresentative data, such as country of birth prevalence estimated from studies representing a small number of individuals; based on antenatal data and therefore limited to women; or conducted several years in the past.

A key assumption of the country of birth method is that the prevalence in migrants from a country reflects the prevalence in that country. In some circumstances this may not be the case and use of local data where available is key to improving accuracy. This effect in the estimates from the 2011 census is lessened by the use of local antenatal data for the majority of the overseas-born population in this study. The prevalence figures used in the calculations are likely to be a conservative estimate as women have been demonstrated to have a substantially lower prevalence of CHB than men in a variety of settings.13,25 Alternatively, antenatal samples collected more recently may underestimate general source country prevalence, given that antenatal data disproportionately samples younger women who are more likely to have been vaccinated (as evidenced by the lower prevalence in younger women observed).13

Timeliness of seroprevalence estimates will be increasingly important given the impact of universal vaccination programs that will substantially alter the epidemiology of CHB in previously high prevalence populations.29 Although much progress has been made in improving access to neonatal hepatitis B vaccination, especially in our own Western Pacific Region,30 the impact of such programs on the epidemiology of CHB in Australia will take decades to be realised.4

Conversely, the only estimates available for seroprevalence of CHB in men who have sex with men are derived from studies undertaken nearly two decades ago, and prevalence may have since reduced, under the influence of targeted vaccination and harm reduction programs.

Another limitation is the assumption of mutual exclusivity of priority populations – for example, people who inject drugs and men who have sex with men were removed from the ‘other Australian born’ category to calculate incremental burden of CHB due to higher prevalence. Source data on country of birth and Indigenous status of these groups is limited, and priority populations may overlap, and the effect of this assumption will be to increase the estimate of CHB in the community.

Finally, data on the size of the priority populations is necessarily self-reported. Given the potential for stigma and discrimination associated with identification with many of these groups, it is possible that the parameters used are underestimates of the true number.

Overall, we feel that the estimates produced here are likely to have underestimated the true prevalence, however, a comprehensive serosurvey is required to help resolve this uncertainty.

Our findings derived from census data support previous assertions4 that the majority of CHB in Australia arises from migration, not locally acquired infection progressing to chronicity. It reinforces the need for our domestic vaccination strategies to be supported by secondary prevention through identifying, monitoring and treating people living with CHB in order to prevent adverse outcomes.

The culturally and linguistically diverse nature of the predominantly affected communities – including migrants speaking a language other than English, and Aboriginal and Torres Strait Islander peoples – mandates that awareness and prevention campaigns be designed and implemented in a culturally appropriate and safe fashion.2,31 Reduced access to health care services including preventative health programs and cancer screening in these groups has been demonstrated.32,33 The use of novel strategies developed in partnership with these communities is required to enhance clinical and public health outcomes.2 This unmet clinical need is emphasised by the fact that only 56% of people currently living with CHB in Australia have been diagnosed and notified. Failure to diagnose CHB results in poor outcomes for the individual affected, and ongoing transmission to susceptible contacts.

This increases the urgency of improving responses to management of CHB, which is estimated to have caused 377 deaths in Australia in 2011. Liver cancer is now the fastest increasing cause of cancer death in Australia, most of which is attributable to viral hepatitis.5 The need to increase access to essential care including regular monitoring and antiviral therapy is undeniable, both in Australia and globally.34–36 Antiviral therapy for CHB has been shown to be effective at reducing morbidity and mortality, and to be a cost-effective cancer-prevention strategy.36,37

Conclusions

The burden of chronic hepatitis B is increasing in Australia and is now estimated to affect around 1% of the population, with the majority of those affected having been born overseas, or Aboriginal and Torres Strait Islander people. Taken together with other available evidence, it is apparent that the prevalence of chronic hepatitis B, and the incidence of complications including cirrhosis and liver cancer, is rising relatively rapidly. No longer can these be considered rare conditions in Australia.

While further research is required to monitor geographic and secular trends, the evidence of increasing numbers of people living with CHB demonstrates that a higher priority needs to be given to preventing adverse outcomes in Australians living with CHB, particularly given that CHB predominantly and disproportionately affects communities subject to broader healthcare inequities. The first National Hepatitis B Strategy provides a framework around which to build community, clinical and public health responses, and further implementation of the key priorities in the Strategy is urgently required.

Acknowledgements

This work was supported through the Communicable Diseases Flexible Fund Grant from the Australian Government Department of Health and Ageing to the Australasian Society for HIV Medicine, and the Victorian Infectious Diseases Reference Laboratory.

This research would not have been possible without access to data from the 2011 Australian Census, and the provision of these in the public domain on the Australian Bureau of Statistics website (http://www.abs.gov.au).

Ancillary