Trends in mortality after diagnosis of hepatitis B or C infection: 1992–2006

Research Article Trends in mortality after diagnosis of hepatitis B or C infection: 1992–2006 Scott R. Walter1,2, Hla-Hla Thein1, Janaki Amin1, Heather F. Gidding1, Kate Ward2, Matthew G. Law1, Jacob George3, Gregory J. Dore1,⇑ 1 National Centre in HIV Epidemiology and Clinical Research, The University of New South Wales, Sydney, Australia; 2New South Wales Department of Health, Sydney, Australia; 3Storr Liver Unit, Westmead Hospital and Westmead Millennium Institute, University of Sydney, Sydney, Australia Background & Aims: Chronic hepatitis B (HBV) or C (HCV) virus infection has been associated with increased risk of death, particularly from liver- and drug-related causes. We examined specific causes of death among a population-based cohort of people infected with HBV or HCV to identify areas of excess risk and examine trends in mortality. Methods: HBV and HCV cases notified to the New South Wales (NSW) Health Department between 1992 and 2006 were linked to cause of death data and HIV/AIDS notifications. Mortality rates and standardised mortality ratios (SMRs) were calculated using person time methodology, with NSW population rates used as a comparison. Results: The study cohort comprised 42,480 individuals with HBV mono-infection and 82,034 with HCV mono-infection. HIV co-infection increased the overall mortality rate three to 10-fold compared to mono-infected groups. Liver-related deaths were associated with high excess risk of mortality in both HBV and HCV groups (SMR 10.0, 95% CI 9.0–11.1; 15.8, 95% CI 14.8– 16.8). Drug-related deaths among the HCV group also represented an elevated excess risk (SMR 15.4, 95% CI 14.5–16.3). Rates of hepatocellular carcinoma (HCC)-related death remained steady in both groups. A decrease in non-HCC liver-related deaths was seen in the HBV group between 1997 and 2006, but not in the HCV group. After a sharp decrease between 1999 and 2002, drug-related mortality rates in the HCV group have been stable. Conclusions: Improvements in HBV treatment and uptake have most likely reduced non-HCC liver-related mortality. Encouragingly, HCV drug-related mortality remained low compared to pre-2002 levels, likely due to changes in opiate supply, and maintenance or improvement in harm reduction strategies. Keywords: Hepatitis B; Hepatitis C; Linkage; Mortality; Australia. Received 1 June 2010; received in revised form 30 July 2010; accepted 19 August 2010; available 23 October 2010 ⇑Corresponding author. Address: National Centre in HIV Epidemiology and Clinical Research, The University of New South Wales, Ground floor, CFI building, Cnr. Boundary and West Streets, Darlinghurst, NSW 2010, Sydney, Australia. Tel.: +61 2 9385 0900; fax: +61 2 9385 0920. E-mail address: gdore@nchecr.unsw.edu.au (G.J. Dore). Abbreviations: HBV, hepatitis B virus; HCV, hepatitis C virus; HCC, hepatocellular carcinoma; NSW, New South Wales; NDD, Notifiable Diseases Database; RBDM, Registry of Births, Deaths, and Marriages; ABS, Australian Bureau of Statistics; ICD, International Classification of Diseases; NHD, National HIV Database; NAR, National AIDS Registry; NCHECR, National Centre in HIV Epidemiology and Clinical Research; SMR, standardised mortality ratio; NSPs, Needle and Syringe Programs. Ó 2010 European Association for the Study of the Liver. Published by Elsevier B.V. All rights reserved. Introduction An estimated 500 million people worldwide have chronic infection with hepatitis B virus (HBV) or hepatitis C virus (HCV) [1]. Both infections are associated with progressive hepatic fibrosis, cirrhosis, and subsequent complications including hepatocellular carcinoma (HCC) [2–3]. The extent of liver disease and the risk of liver disease-related complications in chronic HBV and HCV have been outlined in many clinic-based cross-sectional and prospective cohort studies [4–8]. Few, however, have examined the excess morbidity and mortality attributed to chronic HBV [9– 10] or chronic HCV [9–11] at a population-level compared to background uninfected populations. New South Wales (NSW), the most populous state in Australia, is currently estimated to have 7.1 million residents, making up about 32% of the country’s population [12]. In a previous population-based study we found a 40% increased mortality risk for chronic HBV and 3-fold increased mortality risk for chronic HCV [9]. The excess HCV-related mortality was largely explained by the high rate of drug-related deaths, which outnumbered those related to liver disease. Since 2002, there have been major advances in antiviral therapy for chronic HBV and chronic HCV supported by strategies to enhance treatment uptake in Australia [13]. Other temporal trends over the last decade in Australia have included a decline in the number of injecting drug users resulting in declining HCV incidence [14–15], which may have produced an ageing cohort of people with chronic HCV [16]. People with chronic HBV or HCV have increased liver disease progression if they are co-infected with both viruses, or co-infected with HIV [2,17]; however, there is limited information at the population-level on the impact of co-infection on overall and disease specific mortality. In this study we extended our previous population-level data linkage study to examine ongoing trends in HBV- and HCV-related mortality including the impact of co-infection. The mandatory notification of HBV, HCV, and HIV in Australia, together with estimated high proportions of infected populations being diagnosed Journal of Hepatology 2011 vol. 54 j 879–886 Research Article makes Australia an ideal setting to conduct population-level studies of mortality risk [18–20]. Methods Data sources The study group of interest consisted of all people recorded in the NSW Notifiable Diseases Database (NDD) with either HBV or HCV between 1 January 1992 and 31 December 2006. Since 1991, state government legislation has mandated that all incident cases of HBV and HCV be reported to the NSW Department of Health (NSW Public Health Act 1991). A notifiable HBV case requires detection of HBV surface antigen or HBV DNA, while a notifiable HCV case requires detection of anti-HCV antibody or HCV RNA. Personal identifiers were first recorded in the NDD in 1992. The NSW Registry of Births, Deaths, and Marriages (RBDM) records the date of death for all deaths occurring in NSW. The RBDM supplies the Australian Bureau of Statistics (ABS) with the Medical Certificate of Cause of Death for coding of the underlying cause of death according to the International Classification of Diseases (ICD) [21]. In Australia, HIV is notifiable by the diagnosing laboratory and AIDS is notifiable by physicians. Information about HIV and AIDS is recorded in the National HIV Database (NHD) and National AIDS Registry (NAR) respectively, both of which are maintained by the National Centre in HIV Epidemiology and Clinical Research (NCHECR). Reporting of HIV has been mandatory in NSW since 1985 and has been nationally administered since 1989 [22]. HIV and AIDS data sources use a four letter name code made up of the first two letters of the first and last name. Both data sources also record sex and date of birth information. Linkage Data linkage occurred in two stages. In the first stage, HBV and HCV notifications in the NDD were matched internally to allow identification of cases co-infected with HBV and HCV. All notifications were then matched to RBDM death records. In these steps, linkage was done probabilistically using full name, sex, date of birth and address by means of ChoiceMaker software [23]. ABS cause of death records were linked deterministically to RBDM death records. In the second stage, data were matched deterministically to notifications from the NHD and the NAR using name code, sex, and date of birth. All linkage was performed by the NSW Centre for Health Record Linkage [24]. Table 1. Characteristics of people diagnosed with viral hepatitis in New South Wales: 1992–2006. Viral hepatitis notification Hepatitis B virus n = 42,480 Hepatitis C virus n = 82,034 Hepatitis B and HIV Hepatitis B and C co-infection n = 3,285 co-infection n = 269 Hepatitis C and HIV co-infection n = 620 Year of viral hepatitis diagnosis, median (IQR) 1999 (1995-2002) 1999 (1995-2001) 2000 (1996-2002) 1997 (1994-2002) 1999 (1995-2002) Age at viral hepatitis diagnosis (years), median (IQR) 35.1 (26.9-44.5) 34.5 (27.6-41.6) 35.6 (28.8-42.7) 36.2 (30.0-44.2) 35.8 (30.0-41.5) Males [n], (%) 22,766 (54%) 51,458 (63%) 2,365 (72%) 254 (94%) 558 (90%) Data missing [n], (%) 441 (1%) 390 (<1%) 19 (1%) 0 1 (<1%) Linked deaths [n], (%) 1181 (3%) 4626 (6%) 236 (7%) 63 (23%) 95 (15%) Follow-up time (years), median (IQR) 7.8 (4.6-11.5) 7.6 (4.7-10.9) 6.6 (4.4-9.7) 6.3 (2.7-10.5) 6.1 (3.2-9.5) Table 2. Causes of death by ICD-10 chapter heading among people diagnosed with HBV or HCV mono-infection in New South Wales: 1992–2006. Viral hepatitis notification Description Hepatitis B virus 1 Hepatitis C virus SMR 95% CI Observed deaths Rate1 SMR 95% CI All cause 1181 37.6 1.1 1.0-1.2 4626 78.4 2.6 2.6-2.7 A00-B99 Infections 126 4.0 6.2 5.2-7.4 416 7.0 10.8 9.8-11.8 C00-D48 Neoplasms 521 16.6 1.4 1.3-1.6 866 14.7 1.7 1.6-1.8 D50-D89 Blood/Immune 6 0.2 1.6 0.7-3.6 24 0.4 3.8 2.5-5.7 E00-E90 Endocrine 43 1.4 1.4 1.1-1.9 108 1.8 2.2 1.9-2.7 F00-F99 Mental and behavioural 19 0.6 0.7 0.5-1.2 581 9.8 11.2 10.3-12.2 G00-G99 Nervous system 17 0.5 0.5 0.3-0.9 58 1.0 1.1 0.9-1.4 I00-I99 Circulatory system 225 7.2 0.7 0.6-0.8 710 12.0 1.3 1.2-1.4 J00-J99 Respiratory system 33 1.1 0.5 0.3-0.6 138 2.3 1.2 1.0-1.4 K00-K93 Digestive system 58 1.8 1.5 1.1-1.9 379 6.4 5.7 5.1-6.3 L00-L99 Skin 0 10 0.2 2.7 1.4-5.2 M00-M99 Musculoskeletal 5 0.2 0.9 0.4-2.2 19 0.3 2.1 1.4-3.4 N00-N99 Genitourinary 25 0.8 1.6 1.0-2.3 80 1.4 3.0 2.4-3.7 O00-Q99 Pregnancy/perinatal/congenital 6 0.2 1.3 0.6-3.0 15 0.3 1.7 1.0-2.9 R00-R99 Other 6 0.2 0.9 0.4-2.0 41 0.7 3.0 2.2-4.0 V00-Y98 External 91 2.9 0.7 0.6-0.9 1181 20.0 4.5 4.2-4.8 Observed deaths Rate 1 Rate per 10,000 person-years. SMR standardised mortality ratio. 880 Journal of Hepatology 2011 vol. 54 j 879–886 JOURNAL OF HEPATOLOGY Table 3. Causes of death related to viral hepatitis, HIV infection and drug use among people diagnosed with HBV or HCV mono-infection in New South Wales: 1992– 2006. Description Sex Viral hepatitis notification Hepatitis B virus Observed deaths Rate1 B94.2 C22.0 K70 K71-K77 B20-B24 SMR 95% CI 10.0 9.0-11.1 943 16.0 15.8 14.8-16.8 Men 294 17.9 10.6 9.5-11.9 665 18.2 13.7 12.7-14.8 Women 59 4.0 7.8 6.0-10.1 274 12.4 24.6 21.8-27.6 50 1.6 32.2 24.3-42.6 153 2.6 49.1 41.9-57.5 Men 41 2.5 33.3 24.5-45.2 100 2.7 37.4 30.7-45.5 Women 8 0.6 27.7 13.9-55.4 53 2.4 119.6 91.4-156.6 43 1.4 21.4 15.8-28.9 204 3.5 47.2 41.2-45.2 Men 36 2.2 22.4 16.2-31.1 137 3.8 36.9 31.2-43.6 Women 6 0.4 16.6 7.5-36.9 66 3.0 113.4 89.1-144.4 199 6.3 34.9 30.4-40.2 238 4.0 26.3 23.1-29.8 Men 164 10.0 34.0 29.2-39.6 170 4.7 21.7 18.6-25.2 Women 32 2.2 40.6 28.7-57.5 67 3.0 57.3 45.1-72.8 19 0.6 1.4 0.9-2.3 169 2.9 7.5 6.4-8.7 Men - 1.0 1.5 0.9-2.5 128 3.5 6.7 5.7-8.0 Women - 0.2 1.2 0.4-3.8 41 1.8 11.1 8.1-15.0 19 0.6 2.2 1.4-3.5 154 2.6 10.8 9.2-12.6 Men 12 0.7 1.8 1.0-3.2 109 3.0 9.9 8.2-12.0 Women 7 0.5 3.4 1.6-7.1 43 1.9 13.8 10.2-18.5 5 0.2 0.9 0.4-2.2 17 0.3 1.3 0.8-2.1 Men - 0.2 0.8 0.3-2.1 17 0.5 1.3 0.8-2.1 Women - 0.1 4.3 0.6-30.4 0 Viral hepatitis Sequelae of viral hepatitis3 4 HCC Alcoholic liver disease Non-alcoholic liver disease HIV5 Observed deaths Rate1 11.4 49 1.6 1.4 1.1-1.9 90 1.5 1.7 1.4-2.1 Men 32 2.0 1.4 1.0-2.0 61 1.7 1.7 1.4-2.2 Women 17 1.2 1.5 1.0-2.5 29 1.3 1.7 1.2-2.5 24 0.8 0.7 0.5-1.1 1150 19.5 15.4 14.5-16.3 Men - 1.2 0.8 0.5-1.3 870 23.8 14.2 13.3-15.2 Women - 0.3 0.5 0.2-1.3 270 12.2 20.9 18.6-23.6 C81-C96 Lymphoid ABS 95% CI 358 All liver-related2 B15-B19 Hepatitis C virus SMR 6 Drug-related 1 Rate per 10,000 person-years. 2Consists of viral hepatitis, sequelae of viral hepatitis, HCC, alcoholic and non-alcoholic liver disease, and non-HCC liver cancer. 3Not coded prior to 1st Jan 1997 in this cohort. 4Code included all primary liver cancer. prior to 1st Jan 1997. 5Refers to those who do not appear on the NAR or NHD. 6As defined by the Australian Bureau of Statistics [21]. Counts less than five have been suppressed. SMR standardised mortality ratio. Statistical methods Where a person had multiple records for the same infection the earliest diagnosis date was used. Among co-infected people, the date of diagnosis was defined by the later infection. Due to the potential for higher rates of diagnosis among those with significant morbidity, people who died within six months of the date of hepatitis diagnosis were excluded. Consistent with this exclusion, all other people remaining in the study group had their time at risk shortened by six months. Causes of death in the ABS mortality data were defined using ICD-9 codes for deaths occurring prior to 1 January 1997, and thereafter ICD-10 codes were used. Drug-related deaths were defined according to methods set out by the ABS [21]. This refers to deaths involving dependence disorders due to psychoactive substances, as well as abuse of non-dependence producing substances, the ICD-10 codes for which belong to the mental and behavioural disorders chapter. Drugrelated deaths also include poisoning or overdose by exposure to legal or illegal drugs, the codes for which belong to the external-causes chapter. Death rates after hepatitis diagnosis were estimated using person time methodology. Person time at risk was defined as the time from six months after the date of diagnosis until either the date of death or until December 31st, 2006, if there was no death record. Death rates in the study population were compared to rates in the NSW population for each cause of death using standardised mortality ratios (SMRs). SMRs were adjusted for sex, 5-year-age-group, and calendar year, with the latter two treated as time dependent covariates. Confidence intervals (CIs) for SMRs were calculated assuming a Poisson distribution. Tests for differences between rates and SMRs were performed using Poisson regression. Trends in rates were examined between 1997 and 2006 as small counts prior to 1997 meant rates were unstable. While the change in the cause of death coding systems from ICD-9 to ICD-10 on January 1, 1997 had minimal impact on classifying conditions of interest in this study, it provided a further reason for looking at trends from 1997 onwards. Poisson regression was used to test for trends in rates, with a p-value less than 0.05 considered significant. Ethics approval for the study was granted by the University of NSW Human Research Ethics Committee and the NSW Population & Health Services Research Ethics Committee. Results The final study cohort consisted of 128,726 people who had a HBV or HCV notification between 1992 and 2006. Within the cohort, 42,480 (33.0%) people had HBV mono-infection, 82,034 (63.7%) had HCV mono-infection and 3285 (2.6%) were coinfected with both HBV and HCV (Table 1). Two hundred and sixty-nine (0.2%) and 620 (0.5%) people were co-infected with HBV/HIV and HCV/HIV, respectively. There were 38 (<0.1%) people co-infected with HBV, HCV, and HIV. Approximately 90% of those with HIV co-infection and 72% of HBV/HCV co-infected Journal of Hepatology 2011 vol. 54 j 879–886 881 Research Article Table 4. Causes of death among people with HBV/HCV, HBV/HIV or HCV/HIV co-infection in New South Wales: 1992–2006. Description Sex Viral hepatitis notification Hepatitis B and HIV2 co-infection Hepatitis B and C co-infection Observed Rate1 deaths 95% CI Observed deaths Rate1 SMR 95% CI Observed Rate1 deaths SMR 95% CI All cause 236 111.1 4.0 3.5-4.6 63 378.6 14.5 11.3-18.5 95 255.0 13.1 10.7-16.0 All liver-related 69 32.5 28.8 6.4-15.2 5 30.1 21.4 8.9-51.4 7 18.8 16.2 7.7-34.0 Men 58 37.6 27.2 21.0-35.2 5 31.5 21.7 9.0-52.1 - 18.0 14.3 6.4-31.9 Women 11 19.5 41.6 23.1-75.2 0 - 26.1 77.8 11.0-552.4 8 3.8 14.5 7.2-29.0 0 0 Men - 4.5 14.6 6.9-30.6 Women - 1.8 13.9 2.0-98.6 - 1.4 5.7 1.8-17.7 44 264.4 690.1 513.6-927.4 58 155.7 449.8 347.7-581.8 Men - 1.3 3.9 1.0-15.4 - 258.1 664.5 474.5-875.2 52 155.6 405.3 308.9-531.9 Women - 1.8 115.2 16.2-817.9 - 398.5 22,292 7,189.8-69,119.0 6 156.4 9,231 4,146.9-20,545.9 66 31.1 23.2 18.2-29.5 - 12.0 7.2 1.8-28.9 10 26.8 17.0 9.1-21.5 Men 58 37.6 23.0 17.8-29.7 - 12.6 7.3 1.8-29.3 - 26.9 15.9 8.3-30.6 Women 8 14.2 24.9 12.5-48.8 0 - 26.1 42.9 6.0-304.3 K71-K77 N-a* liver-disease B20-B24 HIV ABS SMR Hepatitis C and HIV2 co-infection Drug-related3 1 Rate per 10,000 person-years. 2Refers to those who appear on the NAR or NHD. 3As defined by the Australian Bureau of Statistics [21]. Counts less than five have been suppressed. SMR standardised mortality ratio. ⁄ N-a Non -alcoholic Table 5. Age-specific liver- and drug-related mortality rates among people with HBV and HCV mono-infection in New South Wales: 1992–2006. Age Hepatitis B virus 1 Hepatitis C virus Drug-related Liver-related 1 Drug-related Liver-related 1 Rate1 N Rate 95% CI N Rate 95% CI N Rate 95% CI N ≤14 - 2.7 0.4-19.0 - 2.7 0.4-19.0 - 4.5 1.1-18.2 0 0 - 95% CI 15-19 0 - - - 3.7 1.2-11.5 0 - - 8 17.3 8.6-34.5 20-24 0 - - - 1.4 0.5-4.4 - 0.3 - 77 25.5 20.4-31.9 25-29 5 1.5 0.6-3.6 - 0.3 0.0-2.1 - 0.6 0.2-1.7 153 23.9 20.4-28.0 30-34 6 1.4 0.6-3.1 5 1.1 0.5-2.8 16 1.7 1.0-2.8 253 27.1 23.9-30.6 35-39 14 2.9 1.7-4.8 - 0.6 0.2-1.9 59 5.2 4.0-6.7 258 22.5 20.0-25.5 40-44 32 6.8 4.8-9.6 - 0.8 0.3-2.3 127 11.2 9.4-13.4 240 21.2 18.7-24.1 45-49 35 9.3 6.7-13.0 - 0.8 0.3-2.5 178 21.9 18.9-25.3 111 13.6 11.3-16.4 50-54 32 12.3 8.7-17.4 - 0.4 0.1-2.7 130 34.1 28.7-40.5 38 10.0 7.3-13.7 55-59 49 29.8 22.5-39.5 0 - - 56 37.6 29.0-48.9 7 4.7 2.2-9.9 60-64 45 43.9 32.8-58.8 0 - - 66 78.3 61.5-99.6 - 4.7 1.8-12.6 65-69 50 68.7 52.0-90.6 0 - - 62 86.1 67.1-110.4 - 1.4 0.2-9.9 70-74 28 55.2 38.1-79.9 0 - - 101 154.8 127.4-188.1 0 - - 75-79 29 95.8 66.6-137.8 0 - - 86 171.6 138.9-211.9 0 - - 80-84 26 188.9 128.6-277.5 0 - - 34 124.2 88.8-173.8 0 - - 85+ 6 70.5 31.7-156.8 0 - - 21 125.5 81.8-192.5 0 - - 1 Rate per 10,000 person-years. Counts less than five have been suppressed. people were males. A total of 6201 people died, with the proportion of deaths ranging from 3% among the HBV mono-infected group to 23% among HBV/HIV co-infected group (Table 1). A total of 1367 people died within six months of diagnosis and were excluded from the analysis. Among those with HBV mono-infection, the leading cause of death was neoplasms (Table 2), the most frequent of which was HCC (38%) followed by lung cancer (12%) and lymphoid cancer (9%). The greatest excess mortality risk was for infections (SMR 882 6.2, 95% CI 5.2–7.4), with the underlying cause for most of these being coded as viral hepatitis (61%). For HCV mono-infected people, however, the leading cause of death was external causes; 72% of which were due to either drug overdose or suicide. Mental and behavioural disorders had the greatest excess rate of death among those with HCV (SMR 11.2, 95% CI 10.3–12.2). In 87% of these cases, drug dependence disorders were recorded as the underlying cause of death although many had overdose recorded as an additional cause. Both the rates and SMRs of all causes of Journal of Hepatology 2011 vol. 54 j 879–886 JOURNAL OF HEPATOLOGY 12 A Hepatitis B virus HCC-related Non-HCC-related 9 6 3 Rate per 10,000 person/years Rate per 10,000 person/years A 250 Hepatitis B virus <1920 200 1920-29 1930-39 150 1940-49 1950-59 100 1960-69 50 0 0 0-29 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 30-39 40-49 B Hepatitis C virus 10 HCC-related Non-HCC-related 5 Rate per 10,000 person/years Rate per 10,000 person/years B 15 50-59 60-69 70-79 80+ 70-79 80+ Age Group Year 250 Hepatitis C virus <1920 200 1920-29 1930-39 150 1940-49 1950-59 100 1960-69 50 0 0 0-29 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 30-39 40-49 50-59 60-69 Age Group Year mortality were significantly higher among the HCV group compared to the HBV group (78.4 vs. 37.6/10,000 person-years, p <0.001; SMR 2.6 vs. 1.1, p <0.001). Across all ICD chapters, both rates and SMRs were higher among the HCV group compared to the HBV group, with the exception of neoplasms where HBV infected people had significantly higher rates (16.6 vs. 14.7/ 10,000 person-years, p = 0.03). In general, the highest SMRs were for hepatitis-related deaths and liver cancer, in particular HCC (Table 3). HCC accounted for 88% and 91% of liver cancer deaths and 56% and 25% of all liver-related deaths for HBV and HCV groups, respectively. Overall rates and SMRs for HCC were significantly higher among HBV infected people than HCV infected people (6.3 vs. 4.0/per 10,000 person-years, p <0.001; SMR 34.9 vs. 26.3, p = 0.003). For nonalcoholic liver disease, the opposite relationship was observed (0.6 vs. 2.9/10,000 person-years, p <0.001; SMR 2.2 vs. 10.8, p <0.001). For both HBV and HCV infected people, rates of mortality due to liver, drug, and lymphoid-related causes were higher for men than women (Table 3). For HCV infected people, SMRs were higher among females for drug-related deaths and all sub-categories of liver-related deaths, significantly so for all liver-related deaths combined (p <0.001). Among people co-infected with HBV and HCV, both the rate and SMR for all causes of mortality were considerably higher than mono-infected groups (111.1/10,000 person-years, SMR 4.0) (Table 4). Among those with HBV, co-infection with HIV Fig. 2. Liver-related mortality rates by age group and birth cohort, (A) hepatitis B virus and (B) hepatitis C virus. 50 Rate per 10,000 person/years Fig. 1. Liver-related mortality rates by year of hepatitis diagnosis, (A) hepatitis B virus and (B) hepatitis C virus. Hepatitis C virus All liver 40 Drug 30 20 10 0 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 Year Fig. 3. All-liver- and drug-related mortality rates among HCV mono-infected people by year of hepatitis diagnosis. increased mortality rates by 10 times (378.6 vs. 37.6/10,000 person-years), while among those with HCV, HIV co-infected people were at least three times more likely to die (255.0 vs. 78.4/10,000 person-years). By far the highest rates and SMRs for all causes of mortality were among the HBV/HCV/HIV co-infected group (586.2/10,000 person-years; SMR 24.2, 95%CI 14.3–40.9). The majority of deaths among HBV/HIV and HCV/HIV co-infected groups were HIV-related; 70% and 61%, respectively. Journal of Hepatology 2011 vol. 54 j 879–886 883 Research Article 50 <1950 Hepatitis C virus 1950-59 1960-69 Rate per 100,000 40 1970-79 1980+ 30 20 10 0 0-19 20-24 25-29 30-34 35-39 40-44 45-49 50-54 55+ Age Group Fig. 4. Drug-related mortality rates by age group and birth cohort among HCV mono-infected people. Age-specific rates of liver-related deaths increased considerably with increasing age from 30 years onwards for both HBV and HCV mono-infected groups (Table 5). The rates peaked in the 80- to 84-year-old group among HBV mono-infected people and in the 75- to 79-year-old group for HCV mono-infected people. However, due to relatively low numbers of notified prevalent HBV and HCV cases in these older age groups, the crude number of liver-related deaths was highest in the 65- to 69-year-old group for HBV (n = 50) and in the 45- to 49-year-old group for HCV (n = 178). Compared to liver-related deaths, rates of drugrelated deaths tended to be elevated among relatively younger ages for both infection groups. The number of age-specific drug-related deaths among HBV mono-infected people was small, while in the HCV mono-infected group counts and rates were elevated from the late teens into the early 40s, after which time they decreased steadily. Rates of HCC-related death showed no clear evidence of change over time for either HBV or HCV mono-infected groups (p = 0.23 for both; Fig. 1). Rates of non-HCC liver deaths in the HBV group decreased by a mean of 5% per year (rate ratio [RR] 0.95, 95% CI 0.89–1.01, p = 0.09), compared to no change in the HCV group (RR 1.00, 95%CI 0.97–1.03, p = 0.84). Overall rates of liver-related deaths decreased by a mean of 4% per year in the HBV group (RR 0.96, 95% CI 0.92–1.00, p = 0.04), while no change was observed in the HCV group (RR 1.01, 95% CI 0.98–1.03, p = 0.44). In Fig. 2A, for each age group, there is a clear decrease in liver-related mortality rates with younger cohorts. A similar effect can be seen in Fig. 2B for the HCV group for those aged under 70; however, the differences between rates for successive cohorts is smaller relative to the HBV group. Among the HCV group, rates of drug-related mortality in 2002 were approximately half those seen prior to 2000 and rates have remained low since 2002 (Fig. 3). In the 5-year period 2002–2006 rates of drug-related mortality in the HCV group were significantly lower than in 1997–2001 (p <0.001). Fig. 4 shows no consistent trend in drug-related mortality rates across age groups or birth cohorts. Discussion Our study outlines the burden of mortality related to HBV and HCV, including important trends in cause-specific mortality in 884 recent years, among notified cases in NSW. Age- and sex-standardised mortality was marginally elevated among those with HBV mono-infection while around two and half times higher for those with HCV mono-infection compared to the NSW population. Among those with a co-infection, the risk was between four and 24 times higher, with those who were co-infected with HIV having a markedly higher risk than other infection groups. Liver conditions, in particular HCC, were a major cause of mortality among HBV infected people. For HCV infected people the leading cause of death was drug-related although liver-related deaths were almost as many in number. High rates of liver-related death were associated with older age, while among HCV infected people, rates of drug-related death were higher in the 15- to 50year-age-group. Mortality rates tended to be higher for males, particularly among the HBV group. Drug-related mortality among people with HCV mono-infection declined markedly between 1999 and 2002 and since then it has been lower than liver-related mortality. Contrasting temporal trends in non-HCC liver-related mortality were found with a stable rate among people with HCV mono-infection and a lower and declining rate among those with HBV mono-infection. Rates of HCC-related mortality have remained approximately stable for both HBV and HCV, although at higher levels for HBV. Age-specific rates of liver-related death declined with younger HBV cohorts; however, for HCV cohorts such a decline was less pronounced. These findings build on those from a previous NSW linkage study [9] in providing additional data on both the impact of HIV co-infection and temporal mortality trends in an era of improving HBV and HCV antiviral therapy, and changing opiate markets. Rates of drug-related deaths decreased considerably between 1999 and 2002, that was thought to be due largely to the Australian heroin shortage in which both supply and purity decreased while the price increased markedly in late 2000 and early 2001 [25–26]. This decrease was of sufficient magnitude to overshadow cohort effects. While a previous linkage study also observed this decrease [9], our study found that rather than return to pre-2001 levels, rates of drug-related deaths have remained low in 2002 to 2006. Although the heroin market stabilised after the shortage, the supply, price and purity had not returned to pre-2001 levels as of 2005 [25]. Wider reaching interventions such as the Needle and Syringe Programs (NSPs) and harm reduction campaigns delivered through the NSPs may also have contributed to the maintenance of improved drug-related mortality since 2001 among those infected with HCV [27]. There is also evidence that the shift in the illicit drug market caused many injecting drug users to change to injecting non-opioid drugs which are less likely to result in a fatal overdose [26,28– 30]. The moderate decline in non-HCC liver disease mortality among people with HBV mono-infection and the decline in agespecific rates of liver-related death with younger cohorts suggest that improved HBV antiviral therapy may have reduced the risk of death from decompensated cirrhosis. HCC was a major contributor to liver-related deaths; however, there was no significant evidence for declining rates of HCC-related deaths for either mono-infected group. During the study period, antiviral therapy licensed for treatment of chronic HBV infection in Australia has been lamivudine (1998), adefovir (2004) and entacavir (2006) [31]. Pegylated interferon alfa-2a was only licensed in 2006, and has had limited uptake [31]. Although lamivudine has been shown to reduce liver-related mortality in chronic HBV infection, Journal of Hepatology 2011 vol. 54 j 879–886 JOURNAL OF HEPATOLOGY subsequently available antiviral therapy has a higher genetic barrier to resistance and generally leads to maintenance of HBV DNA suppression [32–34]. Despite a relatively low HBV antiviral therapy uptake in Australia (an estimated 5% of all chronic infection cases are currently on therapy) [31], this level has increased in more recent years, and the capacity of current therapy to prevent disease progression from advanced liver disease or even reverse decompensated cirrhosis [35–37] means that even low levels of uptake may reduce liver-related deaths. In contrast to HBV treatments, antiviral therapy for chronic HCV infection has both low uptake in Australia (1–2% of chronic infection cases treated per year) [15], and is unable to be utilised in people with decompensated cirrhosis [38–39]. We believe that these factors may explain the non-decrease in both HCC and nonHCC-related mortality rates among those infected with HCV. The contrasting trends in non-HCC liver-related mortality in our study are consistent with findings from a recent linkage study examining hospitalisations in New South Wales [40], which found a significant decrease in hospitalisation rates for non-alcoholic liver disease between 2000 and 2006 for both HBV and HCV infected people. The more striking trends in hospitalisation data are potentially explained by the relatively recent advances in antiviral treatment and uptake having impacted hospitalisation rates more immediately than mortality rates. The stable individual risk of HCC in the setting of expanding populations of both chronic HBV infection [41] and chronic HCV infection [15] is the major factor contributing to the escalating incidence and mortality from HCC in NSW and Australia [42]. Similar trends in HCC mortality have been seen in several other countries, including Taiwan, Canada and the USA [43–45]. The relatively late introduction of universal infant HBV vaccination programs in regional endemic HBV countries such as China and Vietnam (two of the largest immigration sources for Australia), and the long latency of HBV-related HCC means that the programs are thought to have little or no impact on immigrationrelated HCC in Australia [46]. Also, antiviral treatment does not completely remove the risk of developing HCC especially in patients with advanced fibrosis [47–48]. There are several limitations in our study that should be considered. Since no state or national registries exist that collect treatment information, our hypotheses regarding the potentially greater impact of HBV antiviral therapy on liver disease mortality need to be confirmed in other population-based studies examining both mortality and treatment uptake over time. A further limitation is that our study is based on diagnosed and subsequently notified cases. As all diagnosed cases of HBV, HCV, HIV, and AIDS are required by law to be notified, the NDD, NHD, and NAR are very close to complete in terms of diagnosed cases that are notified. It has also been estimated that about 70% of all prevalent HCV infections [19] and 60% of all HBV infections [20] are diagnosed/notified and these estimated high levels of diagnosis should have provided enhanced coverage. While HBV notifications are largely based on evidence of chronic infection, HCV notifications are largely based on the presence of anti-HCV antibodies. Thus, a considerable minority of HCV notifications will not have chronic HCV infection, as an estimated 25% of infections spontaneously clear [49] and remain HCV antibody positive. Less than 2% of NDD records explicitly identified an acute case for either HBV or HCV, while more than 65% of records did not specify whether a case was acute or chronic. For this reason it was not possible to reliably separate acute or chronic cases, hence all the records were included. These limitations are likely to have produced an underestimate of liver disease mortality among people with chronic HCV infection, although this should not have impacted on trends as the surveillance definition has not changed during the study period. This underestimation may contribute to the difference in rates of HCC-related deaths between HBV and HCV mono-infected groups in this study. Information on country of birth was not available for the vast majority of notified HBV and HCV cases and therefore was not included in the analyses. It is likely that there are higher proportions of immigrants from HBV endemic countries among the HBV group in comparison to the NSW population [46], while the HCV group is likely to have higher proportions of Australian born people [15]. Being unable to adjust for this, some bias may have been introduced when calculating SMRs, due to differences in hepatitis-related mortality between various countries of origin. There were 71 HIV notifications and 29 AIDS notifications that matched to more than one NDD notification and so it was not possible to identify which NDD individuals were true matches with the NHD or NAR. Hence the number of people in the cohort co-infected with HIV was underestimated. Approximately 16% of all liver-related deaths were categorised as being due to viral hepatitis; however, deaths related directly to HBV or HCV infection tend to occur from complications of cirrhosis, such as hepatic failure or HCC. Among the cases where viral hepatitis was recorded as the underlying cause of death, many had one of the aforementioned complications listed as an additional cause. However, due to the order in which causes were recorded as well as the absence of additional causes for many cases, it was not possible to further refine the viral hepatitis category. This study identified a positive trend in non-HCC liver-related deaths among those infected with HBV, consistent with improvements in HBV treatment and uptake. There is still a need, however, for increased treatment uptake to reduce cases of advanced liver disease and hence reduce the burden of HCC and other liver-related deaths. Our results also show that changes to opiate supply coinciding with maintenance or improvement in harm reduction strategies have resulted in HCV drug-related mortality remaining stable at well below pre-2002 levels. Financial support This paper was funded from the following sources: the Australian Government Department of Health and Ageing; NSW Cancer Council STREP Grant (SRP08-03). The views expressed in this publication do not necessarily represent the position of the Australian Government. NCHECR is affiliated with the Faculty of Medicine, University of New South Wales. Conflict of interest The authors who have taken part in this study declared that they do not have anything to disclose regarding funding or conflict of interest with respect to this manuscript. Supplementary data Supplementary data associated with this article can be found, in the online version, at doi:10.1016/j.jhep.2010.08.035. Journal of Hepatology 2011 vol. 54 j 879–886 885 Research Article References [1] Marcellin P. Hepatitis B and hepatitis C in 2009. Liver Int 2009;29:1–8. [2] Thomas DL, Seeff LB. Natural History of Hepatitis C. Clin Liver Dis 2005;9: 383–398. [3] McMahon BJ. The Natural History of Chronic Hepatitis B Virus Infection. Hepatology 2009;49:S45–S55. [4] Yu MW, Hsu FC, Sheen IS, Chu CM, Lin DY, Chen CJ, et al. Prospective study of hepatocellular carcinoma and liver cirrhosis in asymptomatic chronic hepatitis B virus carriers. Am J Epidemiol 1997;145:1039–1047. [5] Thein H-H, Qilong Y, Dore GJ, Krahn MD. Estimation of stage-specific fibrosis progression rates in chronic hepatitis C virus infection: a meta-analysis and meta-regression. Hepatology 2008;48:418–431. [6] Dore GJ, Freeman AJ, Law M, Kaldor JM. Natural history models for hepatitis C-related liver disease: different disease progression parameters for different settings. Antivir Ther 2003;8:365–372. [7] Freeman AJ, Dore GJ, Law MG, Thorpe M, Von Overbeck J, Lloyd AR, et al. Estimating progression to cirrhosis in chronic hepatitis C virus infection. Hepatology 2001;34:809–816. [8] Liaw Y-F, Tai D-I, Chu C-M, Chen T-J. The development of cirrhosis in patients with chronic type B hepatitis: a prospective study. Hepatology 1988;8:493–496. [9] Amin J, Law MG, Bartlett M, Kaldor JM, Dore GJ. Causes of death after diagnosis of hepatitis B or Hepatitis C infection: a large community-based linkage study. Lancet. 2006;368:938–945. [10] Duberg AS, Torner A, Davidsdottir L, Aleman S, Blaxhult A, Svensson A, et al. Cause of death in individuals with chronic HBV and/or HCV infection, a nationwide community-based register study. J Viral Hepatol 2008;15: 538–550. [11] McDonald SA, Hutchinson SJ, Bird SM, Mills PR, Dillon J, Bloor M, et al. A population-based record linkage study of mortality in hepatitis C-diagnosed persons with or without HIV coinfection in Scotland. Stat Methods Med Res 2009;18:271–283. [12] Australian Bureau of Statistics. Australian Demographic Statistics, September 2009. Cat. No. 3101.0. Canberra; 2009. [13] Shamliyan TA, MacDonald R, Shaukat A, Taylor BC, Yuan JM, Johnson JR, et al. Antiviral therapy for adults with chronic hepatitis B: a systematic review for a National Institutes of Health Consensus Development Conference. Ann Intern Med 2009;150:111–124. [14] Razali K, Thein HH, Bell J, Cooper-Stanbury M, Dolan K, Dore G, et al. Modelling the hepatitis C virus epidemic in Australia. Drug Alcohol Depend 2007;91:228–235. [15] Gidding HF, Topp L, Middleton M, Robinson K, Hellard M, McCaughan G, et al. The epidemiology of hepatitis C in Australia: notifications, treatment uptake and liver transplantations, 1997–2006. Hepatology 2009;24: 1648–1654. [16] National Centre in HIV Epidemiology and Clinical Research. Australian NSP Survey National Data Report 2003–2007. Sydney: National Centre in HIV Epidemiology and Clinical Research, The University of New South Wales; 2008. [17] Lok AS, McMahon BJ. Chronic Hepatitis B. Hepatology 2007;45:507–539. [18] O’Sullivan BG, Gidding HF, Law M, Kaldor JM, Gilbert GL, Dore GJ. Estimates of chronic hepatitis B virus infection in Australia, 2000. Aust NZ J Public Health 2004;28:212–216. [19] Law MG, Dore GJ, Bath N, Thompson S, Crofts N, Dolan K, et al. Modelling hepatitis C virus incidence, prevalence and long-term sequelae in Australia, 2001. Int J Epidemiol. 2003;32:717–724. [20] Cowie B, Kelly H. Outcomes of a complex dynamic mathematical model of hepatitis B virus in Australia (PP-027). In: 20th Conference of the Asian Pacific Association for the Study of the Liver (APASL). Beijing: Hepatology International; 2010. [21] Australian Bureau of Statistics. Drug-induced Deaths – A Guide to ABS Causes of Death Data. Cat. No. 4809.0.55.001. Australian Bureau of Statistics, Canberra; 2002. [22] McDonald AM, Crofts N, Blumer CE, Gertig DM, Patten JJ, Roberts M, et al. The pattern of diagnosed HIV infection in Australia, 1984–1992. AIDS. 1994;8:513–519. [23] ChoiceMaker Technologies, Inc. New York, NY 10010. [24] The Centre for Health Record Linkage. Available from: www.cherel.org.au. [25] Day C, Degenhardt L, Hall W. Documenting the heroin shortage in New South Wales. Drug Alcohol Rev 2006;25:297–305. 886 [26] Degenhardt L, Day C, Dietze P, Pointer S, Conroy E, Collins L, et al. Effects of a sustained heroin shortage in three Australian States. Addiction 2005;100: 908–920. [27] Horyniak D, Higgs P, Lewis J, Winter R, Dietze P, Aitken C. An evaluation of a heroin overdose prevention and education campaign. Drug Alcohol Rev 2010;29:5–11. [28] Maher L, Li J, Jalaludin B, Wand H, Jayasuriya R, Dixon D, et al. Impact of a reduction in heroin availability on patterns of drug use, risk behaviour and incidence of hepatitis C virus infection in injecting drug users in New South Wales, Australia. Drug Alcohol Depend 2007;89:244–250. [29] Topp L, Day C, Degenhardt L. Changes in patterns of drug injection concurrent with a sustained reduction in the availability of heroin in Australia. Drug Alcohol Depend 2003;70:275–286. [30] Gibson A, Day C, Degenhardt L. The impact of illicit drug market changes on health agency operations in Sydney, Australia. J Subst Abuse Treat 2005;28: 35–40. [31] National Centre in HIV Epidemiology and Clinical Research. HIV/AIDS, viral hepatitis and sexually transmissible infections in Australia Annual Surveillance Report 2009. Sydney: National Centre in HIV Epidemiology and Clinical Research, The University of New South Wales; 2009. [32] Zhang FK. Lamivudine treatment of decompensated hepatitis B virus-related cirrhosis. Hepatobiliary Pancreat Dis Int. 2006;5:10–15. [33] Dienstag JL, Schiff ER, Wright TL, Perrillo RP, Hann HW, Goodman Z, et al. Lamivudine as initial treatment for chronic hepatitis B in the United States. N Engl J Med 1999;341:1256–1263. [34] Liaw YF, Sung JJ, Chow WC, Farrell G, Lee CZ, Yuen H, et al. Lamivudine for patients with chronic hepatitis B and advanced liver disease. N Engl J Med 2004;351:1521–1531. [35] Lai CL. Therapeutic advances in chronic hepatitis B. J Gastroenterol Hepatol 2004;19:S5–S9. [36] Leung N. Chronic hepatitis B – treatment with nucleoside analogues. Med J Malaysia 2005;60:22–27. [37] Takeda A, Jones J, Shepherd J, Davidson P, Price A. A systematic review and economic evaluation of adefovir dipivoxil and pegylated interferon-alpha-2a for the treatment of chronic hepatitis B. J Viral Hepatol 2007;14:75–88. [38] Iacobellis A, Andriulli A. Antiviral therapy in compensated and decompensated cirrhotic patients with chronic HCV infection. Expert Opin Pharmacother 2009;10:1929–1938. [39] Zavaglia C, Airoldi A, Pinzello G. Antiviral therapy of HBV- and HCV-induced liver cirrhosis. J Clin Gastroenterol 2000;30:234–241. [40] Gidding HF, Dore GJ, Amin J, Law MG. Trends in all cause and liver-related hospitalizations in people with hepatitis B or C: a population-based linkage study. BMC Public Health 2010; accepted for publication. [41] Nguyen VT, Amin J, Law MG, Dore GJ. Predictors and survival in hepatitis Brelated hepatocellular carcinoma in New South Wales, Australia. J Gastroenterol Hepatol 2009;24:436–442. [42] Amin J, O’Connell D, Bartlett M, Tracey E, Kaldor J, Law M, et al. Liver cancer and hepatitis B and C in New South Wales, 1990–2002: a linkage study. Aust N Z J Public Health 2007;31:475–482. [43] ElSaadany S, Tepper M, Mao Y, Semenciw R, Giulivi A. An epidemiologic study of hepatocellular carcinoma in Canada. Can J Public Health 2002;93:443–446. [44] El-Serag HB. Hepatocellular carcinoma: recent trends in the United States. Gastroenterology. 2004;127:S27–S34. [45] Altekruse SF, McGlynn KA, Reichman ME. Hepatocellular carcinoma incidence, mortality, and survival trends in the United States from 1975 to 2005. J Clin Oncol. 2009;27:1485–1491. [46] Nguyen VT, Razali K, Amin J, Law MG, Dore GJ. Estimates and projections of hepatitis B-related hepatocellular carcinoma in Australia among people born in Asia-Pacific countries. J Gastroenterol Hepatol 2008;23:922–929. [47] Sanefuji K, Kayashima H, Iguchi T, Sugimachi K, Yamashita Y, Yoshizumi T, et al. Characterization of hepatocellular carcinoma developed after achieving sustained virological response to interferon therapy for hepatitis C. J Surg Oncol 2009;99:32–37. [48] Kobayashi S, Takeda T, Enomoto M, Tamori A, Kawada N, Habu D, et al. Development of hepatocellular carcinoma in patients with chronic hepatitis C who had a sustained virological response to interferon therapy: a multicenter, retrospective cohort study of 1124 patients. Liver Int 2007;27:186–191. [49] Micallef JM, Kaldor JM, Dore GJ. Spontaneous viral clearance following acute hepatitis C infection: a systematic review of longitudinal studies. J Viral Hepatol 2006;13:34–41. Journal of Hepatology 2011 vol. 54 j 879–886