Causes of death in hepatitis B and/or C virus infected–people—lessons for clinical practice


  • Potential conflict of interest: Nothing to report.

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. (Reprinted with permission.)


Background: Hepatitis B and hepatitis C virus infections are common causes of death related to liver disease. In this large study, we aimed to investigate all cause mortality of the viruses in a community-based setting. Methods: In the study population, 39,109 people had hepatitis B, 75,834 had hepatitis C, and 2,604 had hepatitis B and hepatitis C co-infection, notified to the New South Wales state health department, Australia, between 1990 and 2002. Their data were probabilistically linked to the National Death Index. Standardised mortality ratios for all causes of death were calculated and adjusted for age, sex, and calendar year. Results:The number of deaths identified by the linkage were 1233 (3.2%) for hepatitis B, 4008 (5.3%) for hepatitis C, and 186 (7.1%) for hepatitis B and C co-infection. Raised risk of liver-related death (standardised mortality ratios 12.2, 95% CI 10. 7-13.9; 16.8, 15.4-18.3, and 32.9, 23.1-46.7, for hepatitis B, hepatitis C, and hepatitis B and C co- infected patients, respectively) and drug-induced death (1.4, 1.0-2.0; 19.3, 18.1-20.5; and 24.7, 18.2-33.5, respectively) were detected. In people with hepatitis C, raised risk of dying from drug-related causes was significantly greater than from liver-related causes (p= 0.012), with the greatest excess risk in women aged 15-24 years (56.9, 39.2-79.9). Interpretation: All groups had increased risk of liver-related death compared with the standard population, with the greatest excess in people diagnosed with hepatitis B and hepatitis C co-infection. Our data highlight that young people with hepatitis C and with co-infection face a higher mortality risk from continued drug use than from their infection, whereas the main cause of hepatitis B death was liver related.


In this recent paper in Lancet, Janaki Amin and colleagues systematically investigated the causes of death after diagnosis of HBV and/or HCV infection in a community-based setting.1 The authors retrospectively linked data of newly diagnosed HBV (n = 39,109), HCV (n = 75,834), or HBV/HCV coinfection (n = 2,604) reported to their State Health Department with data reported on the causes of death (encoded by the ICD-9/10) in the National Death Index between 1990 and 2002 within their community (New South Wales, Australia). The overall mortality rate, adjusted to sex, age, and calendar year, increased to 1.4-fold (HBV) or 3.1-fold (HCV) compared with the standard population; HBV/HCV-coinfected patients had the highest (5.6-fold) overall mortality (Fig. 1A). The increase in mortality found in this study was lower for HBV-infected patients when compared with previous work from China and Italy,2, 3 but higher for HCV-infected people when compared with literature from Western Europe (Fig. 1A).4–6 Not surprisingly, all groups had significantly increased risks of liver-related death (standardized mortality ratios: HBV, 12.2; HCV, 16.8; HBV/HCV, 32.9). Patients infected with HBV had the highest risk of dying from hepatocellular carcinoma (58% of all deaths in the HBV group). Interestingly, the risk of dying from drug-related causes exceeded liver-related deaths in HCV-infected patients, especially in younger age groups (15-40 years) and in women (Fig. 1B).

Figure 1.

Causes of death after diagnosis of HBV and/or HCV infection. Standardized mortality ratios (SMR) adjusted to age, sex, and calendar year of the general population (SMR = 1, control) were assessed by Amin et al.1 (A) The overall death rate was significantly increased in HBV-infected, HCV-infected, and HBV/HCV-coinfected patients (AUS, Australian population). The increase in mortality found in this study was lower for HBV-infected patients compared with previous work from China (CHN; SMR = 3) or Italy (ITA; SMR = 5.2) but was more pronounced for HCV-infected people compared with literature from Italy (ITA; SMR = 2.5), the United Kingdom (UK; SMR = 1.14), or Germany (GER; SMR = 1.6). References are given in the text. (B) Patients with HCV infection in the study by Amin et al. study had a significantly higher risk of death from drug-related versus liver-related causes (P = 0.012; SMR for the whole population depicted), especially in younger age groups and women (data not shown).

Although the median 5 years of follow-up per person is too short to cover all HBV- or HCV-related complications (because cirrhosis or carcinoma usually develop over longer periods), this large community-based linkage study provides interesting lessons for the clinical hepatologist. First, it once again highlights the risk for HCC in HBV-infected patients, even without underlying cirrhosis. These data from an Australian population with a mixed ethnic background thereby supplement large recent studies conducted in Asian patients. Two studies in HBV-infected cohorts from Taiwan identified novel risk factors for HCC development—specifically HBV-DNA level and HBV genotype C,7, 8 which may prompt the extension of indication for antiviral therapy to asymptomatic HBV carriers with high viral load. The range of drugs for HBV is expanding and currently includes standard and pegylated interferon, lamivudine, adefovir, entecavir, and–likely soon—tenofovir.9 Treatment with antivirals can efficiently suppress HBV replication and subsequently reduces the risk for HCC (e.g., long-term treatment with lamivudine).10 It will be important to implement the new epidemiological findings into current guidelines for diagnosis, treatment, and follow-up of patients infected with HBV.9, 11

Second, illicit drug use is the predominant route of infection with HCV in developed countries today. The study by Amin et al. stresses that young people infected with HCV died disproportionately from complications of drug addiction rather than from liver disease.1 It may therefore be medically and economically more beneficial to primarily focus on the reduction of drug-related harm in these patients; in contrast, older patients with (presumably long-standing) HCV infection had a high risk of death from liver disease and/or HCC and would likely benefit from current therapies based on pegylated interferon and ribavirin.12 Moreover, the decline in standardized mortality ratios for liver-related deaths in HCV-infected (as well as HBV-infected) patients from 1995 to 2002 in this community1 indicates that the latest improvements in the treatment of chronic HBV or HCV infection will reduce mortality rates of HBV-infected or HCV-infected patients in the future.

Third, Amin and colleagues reported an excessive mortality rate among HBV/HCV-coinfected patients.1 Although this excess can be partially attributed to drug-related deaths or additional infection with HIV,1 it confirms that combined chronic HBV and HCV infection is associated with more severe liver disease. However, no standard recommendations exist for treatment of dual HBV/HCV infection.13 Treatment options mostly include interferon with or without lamivudine or ribavirin, and decisions are often based on the determination of the “dominant” hepatitis virus.13 The extraordinarily high mortality among HBV/HCV-coinfected patients underlines the need of further molecular studies to understand the virological interactions as well as large randomized controlled clinical trials to define the best therapeutic strategy.