Abstract
Background.
According to death certificates, approximately 1800 persons die from hepatitis B annually in the United States; however, this figure may underestimate true mortality from chronic hepatitis B (CHB).
Methods.
We analyzed data from CHB patients seen in the Chronic Hepatitis Cohort Study (CHeCS) between 1 January 2006 and 31 December 2013. We compared overall and cause-specific death rates and mean ages at death between CHeCS CHB decedents and U.S. decedents from the Multiple Cause of Death (MCOD) file.
Results.
Of 4389 CHB patients followed for a mean of 5.38 years, 492 (11%) CHB patients died after a mean follow-up of 3.00 years. Compared to survivors, decedents were older, more likely to be White (40.6%), African-American (27.1%), or male (74.2%); and more likely to have had cirrhosis (59.8%), diabetes (27.2%), alcohol abuse (17.7%), hepatocellular carcinoma (17.5%), or a liver transplant (5.7%); whereas survivors were more likely to be Asian (48.8%; all P < .001). CHB patients died at an average age of 59.8 years—14 years younger than the general U.S. population—and at higher rates for all causes (relative risk [RR] = 1.85, 95% confidence interval [CI], 1.851–1.857) and liver-related causes (RR = 15.91, 95% CI, 15.81–16.01). Only 19% of CHB decedents and 40% of those dying of liver disease had hepatitis B reported on their death certificates.
Conclusions.
Compared to the general population, CHB patients die at younger ages and higher rates from all causes and liver-related causes. Death certificates underrepresent the true mortality from CHB.
Keywords: chronic hepatitis B, mortality, liver disease, risk factors, cause of death
Chronic hepatitis B (CHB), an indolent viral infection, progresses over decades to cirrhosis, liver failure, hepatocellular carcinoma, and premature death [1–3] in some patients. In the United States, about 1800 death certificates annually list hepatitis B virus (HBV) as an underlying or contributing cause of death [4]. However, accurately quantifying mortality related to hepatitis is difficult because of the prolonged period between infection and death and because death is not always linked to the underlying infection [5]. Manos et al. [6] found that only 48% of the deaths that were identified as HBV-associated in the medical records of patients with chronic liver disease were also captured as HBV-associated on their death certificates.
Underestimating the true prognosis of CHB infections may have real consequences for patients. Adults known to have high prevalences of CHB infections—including foreign-born Asian-Pacific Islanders [7], racial and ethnic minorities [8], and adults diagnosed with sexually transmitted infections [9]—are often not screened for HBV as recommended. Many high-risk adults are also not vaccinated against hepatitis B [9–11]. Even patients diagnosed with CHB receive less than the recommended monitoring to document changes in the disease activity or any development of hepatocellular carcinoma [12–14], which would allow them to receive timely and effective treatment.
The National Academies of Sciences, Engineering, and Medicine recently determined that hepatitis B elimination is feasible through vaccinations at birth, to infants, and to high-risk adults, and through screening and treatment of HBV-infected adults [15]; they recommended enhanced estimates of morbidity and mortality to guide elimination efforts [16]. Therefore, using data from the Chronic Hepatitis Cohort Study (CHeCS), we sought to (1) describe baseline demographic and clinical characteristics of CHB decedents and survivors, and (2) compare mean ages at death and all-cause and cause-specific mortality rates between CHB decedents and other U.S. decedents through the Multiple Cause of Death (MCOD) file.
METHODS
Study Population and Follow-up
Based on a study of hepatitis C mortality [17], we hypothesized that reported deaths underrepresent the true mortality from hepatitis B. Methods for CHeCS data collection and inclusion criteria for patients with chronic hepatitis B (CHB) have been described [18–20]. Briefly, the cohort is based on electronic health records (EHR) for patients aged ≥18 years who received outpatient, inpatient, emergency department, or laboratory services on or after 1 January 2006 at 1 of 4 large health systems: Geisinger Health System in Danville, Pennsylvania, serving approximately 2.6 million Pennsylvania residents in 44 counties; Henry Ford Health System in Detroit, Michigan, serving over 1 million southeast Michigan residents; Kaiser Permanente–Northwest in Portland, Oregon, serving approximately 500 000 members; and Kaiser Permanente–Honolulu in Honolulu, Hawaii, serving about 220 000 patients, or approximately one-sixth of the Hawaiian population. Patients were classified as having CHB primarily by laboratory results and secondarily based on International Classification of Diseases, Ninth Revision, Clinical Modification (ICD-9-CM) diagnosis codes [18–21]. Electronically-available data were supplemented with reviews of EHR text fields to collect additional information on demographics, hepatitis treatments, non-electronic laboratory results from other medical facilities, and liver biopsy results. Following standard procedures, trained personnel performed medical record abstraction for all patients in the original 2006–2008 cohort. Due to capacity and funding constraints, patients meeting electronic cohort criteria after 2008 were selected by simple random sampling for chart abstraction as funding allowed [19]; a final total of 69% of all patients meeting HBV cohort eligibility criteria had data abstraction. After data abstractors confirmed CHB infection from medical records or an electronically-available test that detects viral hepatitis nucleic acid, all available retrospective data back to the first health system visit was included for enrolled patients.
Each study site compares cohort patient records with the National Death Index, Social Security Death Index, or electronic state death registries annually [17] for those patients known to have died or for those not seen in the health system for 2 years. Deaths are also ascertained from hospitals and clinics reporting through the EHR. EHR-reported deaths without a matching death certificate do not have MCOD data.
Follow-up was defined to begin on 1 January 2006, regardless of the date of first visit, and age was calculated from date of birth as of that date. The study population included patients receiving their first clinical services through 31 December 2013. Follow-up was truncated on 31 December 2014 or the date of death, whichever came first.
Demographic and Clinical Data
Baseline demographic data and mean follow-up time in years were analyzed for CHB decedents and survivors based on data extracted from the EHR. Geocoded addresses obtained at the time of entry into the cohort were matched to U.S. census data in order to estimate each household’s income. Insurance status was defined as of the most recent visit. Positive hepatitis B e antigen (HBeAg) was defined as any positive laboratory value reported during follow-up. HBV deoxyribonucleic acid levels were defined as the first available level. Baseline FIB-4 (fibrosis) score was calculated as the average of [Age(years)*aspartate transaminase (AST) (U/L)]/{Platelet count(109/L)*[alanine transaminase (ALT) (U/L)]1/2} from clinical laboratory values measured within 1 week of each other during the first 2 years of observation, beginning on or after 1 January 2004 and excluding values generated during hospitalizations. If no FIB-4 value could be calculated in the first 2 years of observation, the first available value was used. FIB-4 values greater than 5.17 predict cirrhosis with a high degree of accuracy in the CHeCS CHB population; patients with FIB-4 less than or equal to 1.58 are unlikely to have cirrhosis [22, 23]. We defined cirrhosis as (1) a FIB-4 value greater than 5.17; (2) cirrhosis noted at any time during follow-up by liver biopsy; (3) ICD-9-CM codes 571.2 and 571.5, for uncomplicated cirrhosis, documented in the EHR; or (4) ICD-9-CM and CPT codes for 1 or more of the following documented in the EHR: decompensated cirrhosis, including liver failure with hepatorenal syndrome (572.4), hepatic encephalopathy (572.2), portal hypertension/portal decompression procedures (572.3, 37 140, 37 160, 37 180, 37 181, 37 182, 37 183), esophageal varices complications with bleeding (456.0, 456.20, 42.91, 44.91, 96.06, 43 204, 43 205, 43 243, 43 244, 43 400 and 43 401), and ascites/paracentesis procedures (789.5, 789.59, 54.91, 49 080, 49 081) [24].
Body mass index (BMI) was defined as the first value recorded in the EHR. We defined other diagnoses as present if they were recorded in the EHR at any time during observation, including alcohol abuse (ICD-9-CM 291.0, 291.1, 291.2, 291.3, 291.4, 291.5, 291.8, 291.81, 291.82, 291.89, 291.9, 303.00, 303.01, 303.02, 303.90, 303.91, 303.92, 305.00, 305.01, 305.02, 980.0) [25]; diabetes (ICD-9-CM 250.Xx), if recorded in the EHR twice at 30 or more days apart; hepatitis C [18]; human immunodeficiency virus (HIV; any positive laboratory test, including third-generation enzyme-linked immunosorbent assay antibody test, fourth-generation combined antigen/antibody test, quantitative or qualitative viral load, and/or Western blot test); or liver transplantation (ICD-9-CM and CPT 996.82, 50.5, 50.51, 50.59, 47 135, 47 136, V42.7). Hepatocellular carcinoma was ascertained through validated tumor registry data. We defined “ever treated for CHB” as any prescription for medication licensed for treatment of hepatitis B.
Comparisons With Multiple Cause of Death Data
We calculated all-cause and cause-specific death rates for CHeCS CHB patients by dividing the number of deaths from 2006 through 2014 among CHB patients by the total person-years among CHeCS CHB population seen between 2006 and 2013. We used groupings of International Classification of Diseases, Tenth Revision (ICD-10) codes to classify causes of death as hepatitis B (B16, B16.x, B17.0, and B18.0, B18.1); hepatitis C (B 17.1, B18.2); liver-related, non-alcohol (K71-K77); liver-related, alcohol (K70); liver cancer (C22, D37.6); other hepatitis-related (B15, B15.x, B17.2, B17.8, B17.9, B18.8, B18.9, B 19, B94.2); HIV (B20-B24); cancer, except liver cancer (C00–C97 [except C22] and D37–48 [except D37.6]); circulatory (I00–I99); respiratory (J00–J99); diabetes (E10–E14); genitourinary (N00–N99); injuries and trauma (S00–S99, T00–T98, V01-V99, W00–W99, X00–X99 and Y00–Y36); mental and behavioral disorders (F00–F99); digestive, extra-hepatic (K00–K67 and K80–K93); and other (ICD-10 codes not specifically listed in any of our other classifications) [17]. Deaths were classified as belonging to a group if the underlying cause of death or MCOD fell into that category; decedents could be classified in more than one category if they had more than one cause of death. All-cause and cause-specific death rates were calculated per 100 000 person-years among CHeCS CHB patients and adjusted to the age distribution of the 2010 U.S. census population. For comparison, we calculated all-cause and cause-specific death rates from the U.S. MCOD database using data from 2006–2014 and the same ICD-10 grouping and age criteria, per 100 000 US population age-adjusted to the U.S. 2010 Census.
We calculated mean age at death for CHeCS CHB decedents for all-cause and cause-specific deaths, and compared that number with the mean age at death for adults aged ≥18 years from the 2006–2014 MCOD file.
Death Certificate Reporting of Hepatitis B Virus Infection
For all-cause and cause-specific deaths of CHeCS CHB decedents, we calculated the proportion of deaths for which hepatitis B was listed as underlying cause or contributing cause of death, defined as ICD-10 codes B16, B17.0, B18.0, and B18.1.
We also compared coding on the EHR for cirrhosis, hepatocellular carcinoma, or liver transplantation versus coding on the death certificate for hepatitis B or any liver disease. Liver disease included ICD-10 codes for liver-related conditions (K70–77), hepatocellular carcinoma (C22, D37.6), and hepatitis (B15-B19).
Statistical Analysis
SAS version 9.4 (SAS Institute Inc., Cary, NC) was used to prepare data for analysis, calculate unadjusted descriptive statistics on baseline characteristics and age-adjusted mortality rates, and compare the mortality rates and mean ages at death between CHeCS and MCOD data. All-cause and cause-specific mortality rates were calculated per 100000 person-years within the follow-up period. We used the Pearson Chi-square test for testing differences in percentages and t-tests for testing differences in mean ages and follow-up years. All tests were 2 sided. A P value < .05 was considered statistically significant.
RESULTS
A total of 4389 patients with CHB, seen at 1 of 4 participating medical centers during 2006 through 2013, were followed for a mean of 5.38 years through 31 December 2014. During that timeframe, 492 (11%) CHB patients died, after a mean follow-up of 3.00 years.
Demographic and Clinical Descriptive Data
Demographic and clinical variables for CHeCS CHB decedents and survivors are shown in Table 1. Compared to survivors, decedents were older and twice as likely to be White or African-American. Survivors were more than twice as likely as decedents to be Asian (48.8% vs. 18.4%). Decedents were significantly more likely than survivors to have evidence of cirrhosis (59.8% vs. 12.0%, P < .001). Decedents were also more likely to have evidence of diabetes, a history of alcohol abuse, a hepatitis C or HIV coinfection, hepatocellular carcinoma, a liver transplantation, or a history of treatment for CHB (all P < .001, compared with survivors).
Table 1.
Baseline Demographic and Clinical Characteristics of CHeCS CHB Decedents and Survivors Aged ≥18 Years, 2006–2014
| Number (%) | |||
|---|---|---|---|
|
| |||
| Baseline Characteristics | CHeCS CHB Decedents, n = 492 | Surviving CHeCS CHB Population, n = 3897 | P Value |
| Age as of 1/1/2006, in years (86 missing) | |||
| 18–29 | 12 (2.4) | 796 (20.9) | |
| 30–44 | 91 (18.5) | 1495 (39.2) | |
| 45–59 | 223 (45.4) | 1195 (31.4) | |
| 60–74 | 129 (26.3) | 298 (7.8) | |
| 75+ | 36 (7.3) | 28 (0.7) | <.001 |
| Sex | |||
| Male | 365 (74.2) | 2072 (53.2) | |
| Female | 127 (25.8) | 1825 (46.8) | <.001 |
| Race (23 missing) | |||
| White | 199 (40.6) | 788 (20.3) | |
| Black | 133 (27.1) | 519 (13.4) | |
| Hispanic | 7 (1.4) | 65 (1.7) | |
| Asian | 90 (18.4) | 1890 (48.8) | |
| Hawaiian/Pacific Islander | 29 (5.9) | 277 (7.2) | |
| American Indian/Alaskan Native | 32 (6.5) | 337 (8.7) | <.001 |
| Site | |||
| Kaiser Permanente–Northwest, Portland, Oregon | 85 (17.3) | 1277 (32.8) | |
| Kaiser Permanente–Honolulu, Hawaii | 82 (16.7) | 1063 (27.3) | |
| Henry Ford Health System, Detroit, Michigan | 267 (54.3) | 1260 (32.3) | |
| Geisinger Health System, Danville, Pennsylvania | 58 (11.8) | 297 (7.6) | <.001 |
| Median household income | |||
| <$15000 | 16 (3.4) | 55 (1.5) | |
| >$15000–30 000 | 97 (20.4) | 477 (12.9) | |
| >$30 000–<50 000 | 231 (48.5) | 1502 (40.5) | |
| >$50 000–<75 000 | 93 (19.5) | 1162 (31.3) | |
| >$75 000 | 39 (8.2) | 513 (13.8) | <.001 |
| Insurance status | |||
| Medicaid | 44 (9.4) | 367 (9.6) | |
| Medicare only | 94 (20.0) | 337 (8.8) | |
| Medicare Plus | 125 (26.6) | 364 (9.5) | |
| Private | 175 (37.2) | 2649 (69.1) | |
| None | 32 (6.8) | 119 (3.1) | <.001 |
| Baseline clinical characteristics | |||
| HBeAg status | |||
| Ever HBeAg(+) | 127 (25.8) | 701 (18.0) | |
| Always HBeAg(−) | 169 (34.3) | 2080 (53.4) | |
| HBeAg status missing | 196 (39.8) | 1116 (28.6) | <.001 |
| Median HBV DNA level | |||
| <2000 IU/mL | 69 (14.0) | 1046 (26.8) | |
| 2000–20 000 IU/mL | 36 (7.3) | 510 (13.1) | |
| >20 000 IU/mL | 144 (29.3) | 878 (22.5) | |
| HBV DNA positive; level unknown | 23 (4.7) | 116 (3.0) | |
| HBV DNA missing | 220 (44.7) | 1347 (34.6) | <.001 |
| Cirrhosis | |||
| FIB-4 > 5.17 | 132 (26.8) | 119 (3.1) | <.001 |
| FIB-4 > 1.58 to ≤5.17 | 176 (35.8) | 808 (20.7) | |
| FIB-4 ≤ 1.58 | 78 (15.9) | 2086 (53.5) | |
| FIB-4 missing | 106 (21.5) | 884 (22.7) | |
| Liver biopsy evidence of cirrhosis | 16 (3.2) | 28 (0.7) | <.001 |
| ICD-9-CM and CPT codes for cirrhosis | 60 (12.2) | 93 (2.4) | <.001 |
| ICD-9-CM and CPT codes for hepatic decompensation | 182 (37.0) | 116 (3.0) | <.001 |
| Cirrhosis by any criteria (above) | 294 (59.8) | 467 (12.0) | <.001 |
| Selected risk factors and outcomes | |||
| BMI < 25 | 115 (23.4) | 1607 (41.2) | <.001 |
| BMI 25 to < 30 | 106 (21.5) | 1056 (27.1) | |
| BMI ≥ 30 | 102 (20.7) | 814 (20.9) | |
| BMI missing | 169 (34.3) | 420 (10.8) | |
| History of alcohol abuse | 87 (17.7) | 235 (6.0) | <.001 |
| History of diabetes | 134 (27.2) | 485(12.4) | <.001 |
| Hepatitis C coinfection | 52 (10.6) | 118 (3.0) | <.001 |
| HIV coinfection | 67 (13.6) | 214 (5.5) | <.001 |
| Hepatocellular carcinoma | 86 (17.5) | 45 (1.2) | <.001 |
| Liver transplantation | 28 (5.7) | 46 (1.2) | <.001 |
| Ever treated for CHB | 200 (40.7) | 873 (22.4) | <.001 |
| Mean (median) follow up (years) from 1 January 2006 | 3.0 (2.3) | 5.68 (6.0) | <.001 |
Abbreviations: BMI, body mass index; CHB, chronic hepatitis B; CHeCS, Chronic Hepatitis Cohort Study; CPT, Current Procedural Terminology; DNA, deoxyribonucleic acid; FIB-4, fibrosis-4 score; HBeA, hepatitus B e antigen; HBV, hepatitis B virus; HIV, human immunodeficiency virus; ICD-9-CM, International Classification of Diseases, Ninth Revision, Clinical Modification.
Mortality
Death certificates with MCOD data were available for 409 (83%) CHeCS CHB decedents. Cause-specific and all-cause age-adjusted mortality rates during 2006–2014 for CHeCS patients with CHB were compared with the general U.S. population through the MCOD file, and are shown as sorted by relative risk in Table 2. Patients with missing death certificates were included in all-cause deaths, but in no other category. Risk of death from all causes (relative risk [RR] = 1.854, 95% confidence interval [CI], 1.851–1.857) and all liver-related causes (RR = 15.91, 95% CI, 15.81–16.01) was elevated in CHeCS CHB patients compared to the general population. While death rates from non–liver-related causes were clustered in the lower range of the distribution, only 3 causes of death were significantly lower among CHB patients compared to the MCOD file: injuries/trauma, circulatory, and respiratory causes. (Note that patients may have more than 1 cause of death listed on a death certificate.)
Table 2.
Age-adjusted Mortality Rates Among CHeCS CHB Patients Compared With the MCOD Data by Cause of Death, 2006–2014
| Annual Mortality Rate, per 100000 Person-Years | ||||
|---|---|---|---|---|
|
| ||||
| Cause of Death | CHeCS CHB (n = 492)a | MCOD (n = 2 464 447)b | Relative Risk | (95% CI)* |
| Hepatitis B virus | 299.105 | 0.761 | 392.7 | (374.9–411.4) |
| Hepatitis-related (other) | 17.901 | 0.102 | 173.3 | (152.8–196.7) |
| HIV | 193.617 | 4.518 | 42.96 | (42.14–43.79) |
| Liver cancer | 281.230 | 9.480 | 29.39 | (29.00–29.78) |
| Any liver-related | 716.571 | 44.876 | 15.91 | (15.81–16.01) |
| Liver, non–alcohol-related | 415.891 | 28.298 | 14.66 | (14.55–14.78) |
| Hepatitis C virus | 42.785 | 7.228 | 5.900 | (5.805–5.997) |
| Liver, alcohol-related | 24.719 | 8.994 | 2.746 | (2.703–2.790) |
| Genitourinary | 233.308 | 114.382 | 2.021 | (2.011–2.030) |
| Overall | 1948.137 | 1041.162 | 1.854 | (1.851–1.857) |
| Other causes | 639.560 | 374.191 | 1.689 | (1.685–1.694) |
| Digestive, non-liver | 80.506 | 49.860 | 1.601 | (1.589–1.612) |
| Diabetes | 159.329 | 100.396 | 1.575 | (1.567–1.584) |
| Malignancy, non-liver | 385.194 | 265.196 | 1.445 | (1.441–1.450) |
| Mental and behavioral disorders | 236.007 | 206.219 | 1.128 | (1.124–1.132) |
| Circulatory | 456.327 | 573.388 | 0.788 | (0.786–0.790) |
| Injury/Trauma | 75.569 | 95.927 | 0.784 | (0.779–0.788) |
| Respiratory | 198.685 | 269.264 | 0.730 | (0.727–0.733) |
Abbreviations: CHB, chronic hepatitis B; CHeCS, Chronic Hepatitis Cohort Study; CI, confidence interval; HIV, human immunodeficiency virus; MCOD, Multiple Cause of Death.
Total number of deaths equals 492; total CHB deaths with a death certificate were 409. Cases could have >1 listed cause of death. A mean mortality rate from 2006–2014 was calculated for CHeCS and MCOD data. CHeCS data were age-standardized to the census population in 2010.
Average deaths per year, 2006–2014, age-standardized to the 2010 U.S. census population.
P < .05 for all variables.
Sorted in ascending order, mean ages at death were compared for CHeCS CHB patients and the U.S. general population (MCOD) in Table 3. On average, patients with CHB died significantly younger from all causes than persons in the general population (59.8 vs. 73.9 years, P < .0001). CHeCS CHB patients died, on average, before age 60 from HIV, alcohol-related liver disease, hepatitis C, other hepatitis causes, and non–alcohol-related liver disease. These differences were only significant for non–alcohol-related liver disease (59.3 years for CHeCS CHB patients versus 63 years for MCOD data, P = .0013) and other hepatitis causes (56.4 years for CHeCS CHB patients versus 62.6 years for MCOD data, P = .0158). For CHeCS CHB patients, the mean age at death for most non-hepatitis causes, such as respiratory illness, circulatory illness, and diabetes, was 60.5 to 63.5 years: for each cause, this mean was 10 to 16 years younger for those same causes in the U.S. population as a whole.
Table 3.
Comparison of Mean Age in Years Among CHeCS CHB Patients and the MCOD Data, by Cause of Death, 2006–2014
| Mean Age,a Years | |||
|---|---|---|---|
|
| |||
| Cause of Death | CHeCS CHB | MCOD | P Value |
| HIV | 48.3 | 49.6 | .4323 |
| Injuries/trauma | 51.4 | 56.7 | .3592 |
| Liver disease, alcohol-related | 54.3 | 56.0 | .6578 |
| Hepatitis C virus | 56.3 | 58.2 | .4619 |
| Hepatitis-related, other | 56.4 | 62.6 | .0158 |
| Others | 58.4 | 76.2 | <.0001 |
| Liver disease, non–alcohol-related | 59.3 | 63.0 | .0013 |
| All liver-related | 59.7 | 62.7 | .0006 |
| Overall | 59.8 | 73.9 | <.0001 |
| Respiratory | 60.5 | 76.5 | <.0001 |
| Genitourinary | 60.6 | 77.2 | <.0001 |
| Digestive, non-liver | 60.7 | 75.1 | <.0001 |
| Malignancy, non-liver | 61.3 | 71.9 | <.0001 |
| Hepatitis B virus | 61.6 | 59.2 | .1150 |
| Liver cancer | 61.7 | 67.9 | <.0001 |
| Circulatory | 62.3 | 77.0 | <.0001 |
| Diabetes | 63.5 | 73.9 | <.0001 |
| Mental and behavioral disorders | 63.5 | 75.8 | <.0001 |
Abbreviations: CHB, chronic hepatitis B; CHeCS, Chronic Hepatitis Cohort Study; HIV, human immunodeficiency virus; MCOD, Multiple Cause of Death.
Age-adjusted to U.S. 2010 census.
Death Certificate Reporting
Overall, 78 (19%) of 409 CHeCS CHB patients with matching death certificates had hepatitis B reported on the death certificate. In addition, only 40% of CHB patients with non-alcoholic liver disease, 13% of CHB patients with alcoholic liver disease, 29% of CHB patients with liver cancer, 50% of CHB patients with hepatitis C, and 33% of patients with another form of hepatitis had hepatitis B reported on the death certificate. For patients with any liver disease, including hepatitis B, reported on the death certificate, 40% had hepatitis B documented among the causes of death (Table 4).
Table 4.
All Causes of Death With Hepatitis B Listed as an Underlying or Contributing Cause of Death for CHeCS CHB Patients, 2006–2014
| Cause of Death (CHeCS CHB Patients) | Total Deathsa | HBV Listed on Death Certificate (%) |
|---|---|---|
| Liver disease, non-alcohol-related | 111 | 44 (40) |
| Liver disease, alcohol-related | 8 | 1 (13) |
| Liver cancer | 75 | 22 (29) |
| Hepatitis C virus | 14 | 7 (50) |
| Hepatitis-related, other | 6 | 2 (33) |
| All liver-related causes | 195 | 78 (40) |
| HIV | 42 | 6 (14) |
| Cancer, except liver cancer | 98 | 11 (11) |
| Circulatory | 117 | 19 (16) |
| Respiratory | 52 | 9 (17) |
| Diabetes | 37 | 11 (30) |
| Genitourinary | 61 | 8 (13) |
| Injuries/Trauma | 16 | 3 (19) |
| Mental and behavioral disorders | 52 | 8 (15) |
| Digestive (extra-hepatic) | 25 | 7 (28) |
| Other | 158 | 32 (20) |
| All causes | 409 | 78 (19) |
Abbreviations: CHB, chronic hepatitis B; CHeCS, Chronic Hepatitis Cohort Study; HBV, hepatitis B virus; HIV, human immunodeficiency virus.
Total deaths with a death certificate were 409. More than 1 cause of death could be listed on death certificate, so individual causes will total more than 409.
We also evaluated the proportion of CHeCS CHB decedents who had liver disease coded in the EHR (Table 5). Of 249 CHeCS CHB decedents with cirrhosis by any criteria, 69 (27.7%) had hepatitis B and 159 (66.3%) had any liver disease reported on the death certificate. Hepatitis B was also less likely than any liver disease to be reported on the death certificate for patients with liver cancer (29.3% vs. 100%) or liver transplantation (12.0% vs. 80.8%).
Table 5.
Comparison of Liver Disease Diagnosis in EHR and Death Certificate Reporting Among CHeCS CHB Patients, by Liver-related Causes of Death, 2006–2014
| Death Certificate Reporting |
|||
|---|---|---|---|
| EHR Coding | n | Hepatitis B Reported(%) | Liver Disease Reported (%) |
| All CHB CHeCS decedents | 409 | 78 (19.1) | 195 (47.7) |
| No evidence of cirrhosis | 160 | 9 (5.6) | 30 (18.8) |
| Cirrhosis by any method of ascertainment | 249 | 69 (27.7) | 165 (66.3) |
| FIB-4 > 5.17 | 108 | 33 (30.6) | 83 (76.9) |
| Liver biopsy evidence of cirrhosis | 11 | 7 (63.6) | 10 (90.9) |
| Diagnostic code for cirrhosis | 48 | 17 (35.4) | 42 (87.5) |
| Diagnostic / procedure code for decompensation | 160 | 53 (33.1) | 121 (75.6) |
| Liver cancer | 75 | 22 (29.3) | 75 (100.0) |
| Liver transplantation | 25 | 3 (12.0) | 20 (80.0) |
Total number of deaths with a death certificate was 409. Patients could have more than 1 diagnosis or cause of death.
Abbreviations: CHB, chronic hepatitis B; CHeCS, Chronic Hepatitis Cohort Study; EHR, electronic health record; FIB-4, fibrosis-4 scores.
DISCUSSION
Compared to the general population, CHeCS CHB patients died at higher rates from all causes and from liver-related causes, including liver cancer. They also died, on average, 14 years younger than decedents in the general population. Almost half (48%) of CHB decedents died at an average age of <60 years, with liver disease listed as an underlying or contributing cause of death on the death certificate. Despite this, only 40% of CHB decedents dying with liver disease had hepatitis B reported as an underlying or contributing cause on the death certificate. Although the relative contribution of HBV to any individual death cannot always be determined with CHeCS data, our study suggests a meaningful effect of CHB on mortality for people living with this chronic infection.
Because risk-based screening is incompletely implemented in medical practices [7–9], only 26% of persons with CHB were aware of their infection in the 2011–2014 National Health and Nutrition Examination Survey [26]. In fact, 41% of CHeCS CHB patients reported that they were first diagnosed with hepatitis B because of clinical symptoms [27]. Among patients in ongoing medical care with clinical indications for testing (2 or more abnormal ALT levels), less than half were subsequently tested for hepatitis B and an estimated one-fifth of CHB patients in care went undiagnosed with hepatitis B infection [28]. Patients diagnosed with CHB are unlikely to be followed optimally to detect the development of cirrhosis and hepatocellular carcinoma [12–14]. Finally, our data demonstrate that even patients with documented complications of CHB, including cirrhosis, liver transplantation, or liver cancer, were unlikely to have the diagnosis of hepatitis B recorded on their death certificate. To summarize, many patients with CHB are unlikely to be diagnosed [7–9, 26–28]; their fibrosis or other hepatic complications may go undetected [12–14]; and their chronic infection with hepatitis B is unlikely to be recorded at the time of death.
Others have confirmed that hepatitis B is associated with premature death and with elevated rates of death from all causes and from liver-related causes, including hepatocellular carcinoma [29–31]. Similar to our findings, diabetes [32, 33] and alcohol abuse [33, 34] are associated with an increased risk of death in persons with CHB. HIV coinfection also confers an increased risk of death [30, 35], especially in cirrhotic patients [35]. Although some have postulated a protective effect of hepatitis C viral infection on mortality from hepatitis B [36], other data demonstrates worse outcomes from hepatitis B and C coinfection compared to mono-infection [37, 38]. We were unable to evaluate the contribution of tobacco use in our study; however, smoking has been identified as a risk factor for hepatocellular carcinoma among persons with CHB [39]. Strengths of the study include similarities in age at death between the CHeCS CHB population and hepatitis B decedents in the MCOD data, and similar demographic [6, 26, 29] and prognostic [29, 30, 32, 35, 38] factors as those identified in other U.S. studies, suggesting that the CHeCS CHB decedents are similar to CHB decedents in the U.S. general population.
Limitations to our analysis include missing death certificates for 83 (17%) CHB decedents. BMI was missing for 13% of the CHeCS population and 34% of decedents. Smoking data was available for only 10% of the study population and was omitted from this analysis. In addition, laboratory test results were unavailable for many patients [12], including hepatitis e antigen (29.9%), HBV deoxyribonucleic acid level (35.7%), and FIB-4 (22.5%) results. Also, hepatitis B screening tests may not be ordered for high-risk persons, leading to underreporting of chronic hepatitis B [28]. EHR coding underestimates alcohol abuse [25] and may underestimate other clinical parameters. Finally, the CHeCS cohort represents people who received documented medical care for hepatitis B on at least 2 occasions that were 6 or more months apart, and is thus lacking representation from incarcerated persons, uninsured persons, and persons with limited or sporadic access to health care [18].
In conclusion, we observed increased prevalences of cirrhosis and other liver-related complications among CHB decedents compared to survivors, and increased risks of death from all causes and from all liver-related causes among CHB patients compared to the U.S. population. CHB patients also died, of all causes, on average, 14 years younger than persons in the general population. Almost half of these patients died at an average age of <60, with liver disease listed as an underlying or contributing cause. Despite this, only 40% of CHB decedents dying of liver disease had hepatitis B reported on their death certificates, implying that national vital statistics data substantially underestimate the mortality burden of HBV. These findings strongly support hepatitis B elimination as a national public health goal [15, 16]. Clinicians can contribute to elimination of hepatitis B by (1) vaccinating all unvaccinated high-risk adults for HBV; (2) appropriately screening all high-risk persons for HBV; and (3) assuring that persons with HBV receive posttest counseling and clinical care [40].
Acknowledgments.
The authors thank Blythe Ryerson. The CHeCS Investigators include the following investigators and sites: Scott D. Holmberg, Eyasu H. Teshale, Philip R. Spradling, Anne C. Moorman, Jian Xing, and Yuna Zhong, Division of Viral Hepatitis, National Centers for HIV, Viral Hepatitis, STD, and TB Prevention (NCHHSTP), Centers for Disease Control and Prevention (CDC), Atlanta, Georgia; Stuart C. Gordon, David R. Nerenz, Mei Lu, Lois Lamerato, Jia Li, Loralee B. Rupp, Nonna Akkerman, Talan Zhang, Sheri Trudeau, Yueren Zhou, and Kuan-Han Wu, Henry Ford Health System, Detroit, Michigan; Joseph A. Boscarino, Zahra S. Daar, and Robert E. Smith, Department of Epidemiology and Health Services Research, Geisinger Clinic, Danville, Pennsylvania; Yihe G. Daida, Connie Mah Trinacty, Jonathan W. Lai, and Carmen P. Wong, The Center for Health Research, Kaiser Permanente-Hawaii, Honolulu, Hawaii; Mark A. Schmidt and Judy L. Donald, The Center for Health Research, Kaiser Permanente-Northwest, Portland, OR.
Financial support.
This work was supported by funding to the Henry Ford Health System for CHeCS from the Centers for Disease Control and Prevention and from Gilead Sciences. CHeCS was previously funded through May 2016 by the CDC Foundation, which received grants from AbbVie; Genentech, A Member of the Roche Group; Gilead Sciences; Janssen Pharmaceuticals, Inc.; and Vertex Pharmaceuticals; past partial funders included Bristol-Myers Squibb.
S. C. G. receives grant/research support from AbbVie Pharmaceuticals, Bristol-Myers Squibb, Conatus, CymaBay, Exalenz BioScience, Gilead Pharmaceuticals, Intercept Pharmaceuticals, Merck and Shire; is a consultant/advisor for Abbvie, Bristol-Myers Squibb, CVS Caremark, Dova Pharmaceuticals, Gilead, Intercept, and Merck; and serves as a speaker/teacher in programs sponsored by Intercept Pharmaceuticals. J. A. B., Y. G. D., L. B. R. and M. A. S. report grants from Gilead Pharmaceuticals and Intercept Pharmaceuticals. M. A. S. also reports support from Takeda Vaccines and P-95. All other authors report no potential conflicts of interest. All authors have submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest.
Footnotes
Disclaimer.
The findings and conclusions in this report are those of the authors and do not necessarily represent the official position of the Centers for Disease Control and Prevention. Granting corporations did not have access to data and did not contribute to data analysis or writing of manuscripts.
Potential conflicts of interest.
Conflicts that the editors consider relevant to the content of the manuscript have been disclosed.
References
- 1.Burns GS, Thompson AJ. Viral hepatitis B: clinical and epidemiological characteristics. Cold Spring Harb Perspect Med 2014; 4:a024935. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Ieluzzi D, Covolo L, Donato F, Fattovich G. Progression to cirrhosis, hepatocellular carcinoma and liver-related mortality in chronic hepatitis B patients in Italy. Dig Liver Dis 2014; 46:427–32. [DOI] [PubMed] [Google Scholar]
- 3.Poh Z, Goh BB, Chang PE, Tan CK. Rates of cirrhosis and hepatocellular carcinoma in chronic hepatitis B and the role of surveillance: a 10-year follow-up of 673 patients. Eur J Gastroenterol Hepatol 2015; 27:638–43. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Centers for Disease Control and Prevention. Viral hepatitis statistics and surveillance. Available at: https://www.cdc.gov/hepatitis/statistics/index.htm Accessed 30 August 2017. [Google Scholar]
- 5.Wiktor SZ, Hutin YJ. The global burden of viral hepatitis: better estimates to guide hepatitis elimination efforts. Lancet 2016; 388:1030–1. [DOI] [PubMed] [Google Scholar]
- 6.Manos MM, Leyden WA, Murphy RC, Terrault NA, Bell BP. Limitations of conventionally derived chronic liver disease mortality rates: results of a comprehensive assessment. Hepatology 2008; 47:1150–7. [DOI] [PubMed] [Google Scholar]
- 7.Vijayadeva V, Spradling PR, Moorman AC, et al. Hepatitis B virus infection testing and prevalence among Asian and Pacific Islanders. Am J Manag Care 2014; 20:e98–e104. [PMC free article] [PubMed] [Google Scholar]
- 8.Hu DJ, Xing J, Tohme RA, et al. Hepatitis B testing and access to care among racial and ethnic minorities in selected communities across the United States, 2009–2010. Hepatology 2013; 58:856–62. [DOI] [PubMed] [Google Scholar]
- 9.Hechter RC, Jacobsen SJ, Luo Y, et al. Hepatitis B testing and vaccination among adults with sexually transmitted infections in a large managed care organization. Clin Infect Dis 2014; 58:1739–45. [DOI] [PubMed] [Google Scholar]
- 10.Williams WW, Lu P-J, O’Halloran A, et al. Surveillance of vaccination coverage among adult populations–United States, 2015. MMWR Surveill Summ 2017; 66:1–28. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Yue X, Black CL, O’Halloran A, Lu PJ, Williams WW, Nelson NP. Hepatitis A and hepatitis B vaccination coverage among adults with chronic liver disease. Vaccine 2018; 36:1183–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Spradling PR, Xing J, Rupp LB, et al. ; Chronic Hepatitis Cohort Study (CHeCS) Investigators. Infrequent clinical assessment of chronic hepatitis B patients in United States general healthcare settings. Clin Infect Dis 2016; 63:1205–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Serper M, Choi G, Forde KA, Kaplan DE. Care delivery and outcomes among US veterans with hepatitis B: a national cohort study. Hepatology 2016; 63:1774–82. [DOI] [PubMed] [Google Scholar]
- 14.Singal AG, Tiro J, Li X, Adams-Huet B, Chubak J. Hepatocellular carcinoma surveillance among patients with cirrhosis in a population-based integrated health care delivery system. J Clin Gastroenterol 2017; 51:650–5. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.National Academies of Sciences, Engineering, and Medicine. Eliminating the public health problem of hepatitis B and C in the United States: Phase one report. Washington, D.C.: The National Academies Press, 2016. Available at: http://nationalacademies.org/hmd/Reports/2016/Eliminating-the-Public-Health-Problem-of-Hepatitis-B-and-C-in-the-US.aspx. Accessed 27 July 2017. [PubMed] [Google Scholar]
- 16.National Academies of Sciences, Engineering, and Medicine. A national strategy for the elimination of hepatitis B and C. Washington, D.C.: The National Academies Press, 2017. Available at: http://www.nationalacademies.org/hmd/reports/2017/national-strategy-for-the-elimination-of-hepatitis-b-and-c.aspx. Accessed 27 July 2017. [Google Scholar]
- 17.Mahajan R, Xing J, Liu SJ, et al. Mortality among persons in care with hepatitis C virus infection: the chronic hepatitis cohort study (CHeCS), 2006–2010. Clin Infect Dis, 2014; 58:1055–61. [DOI] [PubMed] [Google Scholar]
- 18.Moorman AC, Gordon SC, Rupp LB, et al. ; Chronic Hepatitis Cohort Study Investigators. Baseline characteristics and mortality among people in care for chronic viral hepatitis: the chronic hepatitis cohort study. Clin Infect Dis 2013; 56:40–50. [DOI] [PubMed] [Google Scholar]
- 19.Lu M, Rupp LB, Moorman AC, et al. Comparative effectiveness research of chronic hepatitis B and C cohort study (CHeCS): improving data collection and cohort identification. Dig Dis Sci 2014; 59:3053–61. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Mahajan R, Moorman AC, Liu SJ, Rupp L, Klevens RM; Chronic Hepatitis Cohort Study (CHeCS) investigators. Use of the international classification of diseases, 9th revision, coding in identifying chronic hepatitis B virus infection in health system data: implications for national surveillance. J Am Med Inform Assoc 2013; 20:441–5. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Abara WE, Moorman AC, Zhong Y, et al. The predictive value of international classification of disease codes for chronic hepatitis C virus infection surveillance: the utility and limitations of electronic health records. Popul Health Manag 2018; 21:110–5. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Holmberg SD, Lu M, Rupp LB, et al. ; Chronic Hepatitis Cohort Study (CHeCS) Investigators. Noninvasive serum fibrosis markers for screening and staging chronic hepatitis C virus patients in a large US cohort. Clin Infect Dis 2013; 57:240–6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Li J, Gordon SC, Rupp LB, et al. ; Chronic Hepatitis Cohort Study (CHeCS) Investigators. The validity of serum markers for fibrosis staging in chronic hepatitis B and C. J Viral Hepat 2014; 21:930–7. [DOI] [PubMed] [Google Scholar]
- 24.Xu F, Moorman AC, Tong X, et al. All-cause mortality and progression risks to hepatic decompensation and hepatocellular carcinoma in patients infected with hepatitis C virus. Clin Infect Dis 2016; 62:289–97. [DOI] [PubMed] [Google Scholar]
- 25.Boscarino JA, Moorman AC, Rupp LB, et al. ; Chronic Hepatitis Cohort Study (CheCS) Investigators. Comparison of ICD-9 codes for depression and alcohol misuse to survey instruments suggests these codes should be used with caution. Dig Dis Sci 2017; 62:2704–12. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Kim HS, Rotundo L, Yang JD, et al. Racial/ethnic disparities in the prevalence and awareness of Hepatitis B virus infection and immunity in the United States. J Viral Hepat 2017; 24:1052–66. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Gerbi GB, Rupp LB, Ko SC, Moorman AC, Holmberg SD, Xu F; CHeCS Investigators. Reported reasons for testing among hepatitis B virus-infected patients - Chronic Hepatitis Cohort Study (CHeCS), United States, 2006–2010. Liver Int 2014; 34:e162–3. [DOI] [PubMed] [Google Scholar]
- 28.Spradling PR, Rupp L, Moorman AC, et al. ; Chronic Hepatitis Cohort Study Investigators. Hepatitis B and C virus infection among 1.2 million persons with access to care: factors associated with testing and infection prevalence. Clin Infect Dis 2012; 55:1047–55. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Ly KN, Xing J, Klevens RM, Jiles RB, Holmberg SD. Causes of death and characteristics of decedents with viral hepatitis, United States, 2010. Clin Infect Dis 2014; 58:40–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Pinchoff J, Tran OC, Chen L, et al. Impact of hepatitis B on mortality and specific causes of death in adults with and without HIV co-infection in NYC, 2000–2011. Epidemiol Infect 2016; 144:3354–64. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Montuclard C, Hamza S, Rollot F, et al. Causes of death in people with chronic HBV infection: a population-based cohort study. J Hepatol 2015; 62:1265–71. [DOI] [PubMed] [Google Scholar]
- 32.Stepanova M, Rafiq N, Younossi ZM. Components of metabolic syndrome are independent predictors of mortality in patients with chronic liver disease: a population-based study. Gut 2010; 59:1410–5. [DOI] [PubMed] [Google Scholar]
- 33.Mallet V, Hamed K, Schwarzinger M. Prognosis of patients with chronic hepatitis B in France (2008–2013): a nationwide, observational and hospital-based study. J Hepatol 2017; 66:514–20. [DOI] [PubMed] [Google Scholar]
- 34.Hsu CC, Kowdley KV. The effects of alcohol on other chronic liver diseases. Clin Liver Dis 2016; 20:581–94. [DOI] [PubMed] [Google Scholar]
- 35.Rajbhandari R, Jun T, Khalili H, Chung RT, Ananthakrishnan AN. HBV/HIV coinfection is associated with poorer outcomes in hospitalized patients with HBV or HIV. J Viral Hepat 2016; 23:820–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36.Huang YT, Freeman JR, Yang HI, Liu J, Lee MH, Chen CJ. Mediation effect of hepatitis B and C on mortality. Eur J Epidemiol 2016; 31:625–33. [DOI] [PubMed] [Google Scholar]
- 37.Huang YT, Jen CL, Yang HI, et al. Lifetime risk and sex difference of hepatocellular carcinoma among patients with chronic hepatitis B and C. J Clin Oncol 2011; 29:3643–50. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38.Kruse RL, Kramer JR, Tyson GL, et al. Clinical outcomes of hepatitis B virus coinfection in a United States cohort of hepatitis C virus-infected patients. Hepatology 2014; 60:1871–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 39.Franceschi S, Montella M, Polesel J, et al. Hepatitis viruses, alcohol, and tobacco in the etiology of hepatocellular carcinoma in Italy. Cancer Epidemiol Biomarkers Prev 2006; 15:683–9. [DOI] [PubMed] [Google Scholar]
- 40.Abara WE, Qaseem A, Schillie S, McMahon BJ, Harris AM; High Value Care Task Force of the American College of Physicians and the Centers for Disease Control and Prevention. Hepatitis B vaccination, screening, and linkage to care: best practice advice from the american college of physicians and the centers for disease control and prevention. Ann Intern Med 2017; 167:794–804. [DOI] [PubMed] [Google Scholar]