| |
The persistence of underreporting of hepatitis C as an underlying or contributing cause of death, 2011-2017
|
| |
| |
Download the PDF here
Clinical Infectious Diseases 09 February 2021
Philip R. Spradling1, Yuna Zhong1, Anne C. Moorman1, Loralee B. Rupp2, Mei Lu2, Eyasu H. Teshale1, Mark A. Schmidt3, Yihe G. Daida4, Joseph A. Boscarino5, Stuart C. Gordon2,6, for the CHeCS Investigators*
1Division of Viral Hepatitis, Centers for Disease Control and Prevention (CDC), Atlanta, GA, USA;
2Henry Ford Health System, Detroit, MI, USA; 3The Center for Health Research, Kaiser Permanente
Northwest, Portland, OR, USA; 4The Center for Integrated Health Research, Kaiser Permanente Hawaii, Honolulu, HI, USA; 5Department of Population Health Science
Abstract
Using electronic health records, we found that hepatitis C reporting on death certificates of 2,901 HCV-infected decedents from four U.S. healthcare organizations during 2011-2017 was documented in only 50% of decedents with hepatocellular carcinoma and less than half with decompensated cirrhosis. National figures likely underestimate the U.S. HCV mortality burden.
Background
In 2018, there were 15,713 hepatitis C-associated deaths reported in the United States [1]. U.S. mortality estimates are derived from death certificates and compiled nationally as multiple cause of death data [2,3]. Accordingly, accurate cause of death documentation on death certificates is essential for understanding the burden of a disease and for gauging the effectiveness of interventions to reduce its lethality. In a previous publication using U.S. healthcare data from 2006-2010, we found that hepatitis C was reported as an underlying or contributing cause of death in 19% of patients with identified hepatitis C virus (HCV) infection, and in only 30% of those with hepatitis C and advanced liver disease [4]. We hypothesized that increased provider awareness of hepatitis C since the initial study would result in more frequent death certificate reporting and sought to determine whether such a change occurred during 2011-2017.
Results
A total of 21,378 patients with hepatitis C were followed for a median of 6.0 years, during which 3,413 (16%) died after a median follow-up of 2.3 years; 2,901 (85%) of these decedents had death certificate data on cause of death available. The median age at death was 60.0 years. Overall, hepatitis C was listed as an underlying or contributing cause of death for 783 (27%) decedents, compared with 19% during 2006-2010. Among patients in the 2011-2017 cohort with documentation of liver-related/non-alcohol, liver-related/alcohol, or hepatocellular carcinoma as a cause of death, hepatitis C was listed on 52%, 49%, and 50% of death certificates, respectively. Hepatitis C was listed less frequently for non-liver-related causes, ranging from 9% of deaths from injuries/trauma to 36% of deaths from diabetes. Except for deaths from injuries/trauma, reported frequencies of hepatitis C on death certificates increased for all causes relative to 2006-2010, albeit marginally (Table 1).
Among patients with advanced liver disease according to the EHR, hepatitis C was listed on death certificates of 36% of patients with any cirrhosis, 43% of those with FIB-4 greater than 5.88, 44% of patients with decompensated cirrhosis, and 36% of decedents who had undergone liver transplantation.
Relative to Whites, who had hepatitis C listed for 30% of deaths,reporting was significantly lower among Black decedents (19% [RR 0.62]), and higher among decedents of Hispanic (47% [RR 1.57]) and Asian
(41% [RR 1.38]) descent. Compared to decedents with annual income <$15,000, the likelihood of reporting hepatitis C associated with death increased as income increased (23% [RR 1.86], 27% [RR 2.18], 30% [RR 2.38] for annual income ranges of $15,000-$30,000, >$30,000-$50,000, and >$50,000, respectively) (P<.001, Cochran-Armitage Trend Test). Finally, compared with Medicaid recipients (22%), decedents with Medicare plus supplemental coverage (35% [RR 1.58]) and those with private insurance (31% [RR 1.41]) were more likely to have hepatitis C reported on death certificates.
HCV-infected decedents with hepatitis C reported were more likely affiliated with the two Kaiser Permanente sites, which suggests that more cohesive and centralized care delivery systems (such as staff-model health maintenance organizations) may be better equipped to exercise policies and procedures that favor consistency and accuracy in reporting multiple causes of death.
An increase in the frequency of reporting of a diagnosis on death certificates does not necessarily indicate better reporting. Some conditions, such as heart disease, have been subject to overreporting
[10]. With respect to hepatitis C, it is important to differentiate whether persons die from rather than with HCV infection. Hepatitis C is particularly challenging in this respect. Given the relationship of HCV infection with several extrahepatic conditions [11], infected persons may suffer from its effects in the absence of severe liver disease. With our HCV-infected cohort, however, it seems more plausible that underreporting was the prevailing tendency, since over half of decedents with indisputable end-stage liver disease had no death certificate documentation of hepatitis C.
Nonetheless, our findings may not be generalizable to hepatitis C decedents in other U.S. settings, and 15% of our decedents lacked death certificate data; thus, it is difficult to assess the degree to which underreporting of hepatitis C on a larger scale might affect national mortality estimates. However, given the geographic and demographic diversity of our cohort, it is likely that such underreporting is not uncommon and, consequently, that national figures underestimate the actual mortality burden of hepatitis C. Our data also lacked specificity regarding the setting in which deaths occurred (e.g., at home vs. facility); thus, we could not determine whether such factors might affect cause of death documentation.
Appeals to the medical community for measures to improve death certificate reporting are not new [12]. Ideally, such measures are applied at an early, formative stage of the medical career [13]. Promotion and further study of such accuracy are critical to inform policies, target public health resources, and reduce disparities in mortality.
| |
| |
| |
|
|
|
