INTRODUCTION
Hepatitis E, caused by hepatitis E virus (HEV), is an important public health concern as a common cause of enterically transmitted hepatitis worldwide. It is usually self-limiting but may develop to fulminant hepatitis, especially in the pregnant population [Reference Hamid1, Reference Banait2]. Two billion people, representing one third of the world's population, have been exposed to the virus [Reference Khuroo3, Reference Mirazo4]; and according to Rein et al.’s estimate, there are 20 million HEV-infected cases, 3·4 million symptomatic cases, 70 000 deaths, and 3000 stillbirths in nine endemic regions [Reference Rein5].
Estimating the disease burden of hepatitis E is important to increase scientific and social awareness of the disease and to inform decisions about health policy priorities and disease-control technologies [Reference Conteh, Engels and Molyneux6], such as vaccines [Reference Wu7]. However, data on the disease burden are scarce [Reference Xiu8, Reference Jiang9]. Therefore, we conducted a community-based study to evaluate the disease burden of hepatitis E cases reported between 2008 and 2015 in a rural region of China.
METHODS
Hepatitis surveillance
An active hepatitis surveillance system was enacted between 2008 and 2015 in 11 townships (Anfeng, Fuan, Hougang, Liangduo, Nanshenzao, Qindong, Shiyan, Tangyang, Wulie, Xinjie, Xuhe) of Dongtai City, Jiangsu Province, China, and the annual numbers of registered residents recruited in the surveillance system between 2008 and 2015 were 486417, 483685, 482342, 478734, 473260, 471470, 466160, and 462 813, respectively. The working procedure of the surveillance system has already been discussed in previous studies, and a hepatitis E case was defined as a patient who fulfilled three conditions [Reference Zhu10, Reference Zhu11]: acute illness lasting for at least 3 days; abnormal serum alanine transaminase concentration ⩾2·5 times the upper limit of normal range; and positive HEV IgM and RNA, ⩾4 times increase in HEV IgG, or both. Further, all hepatitis E cases reported in the active surveillance system were investigated face to face, and written informed consent was obtained from each participant before conducting research. Approval of the study was obtained from the Ethics Committee of the Jiangsu Provincial Center for Disease Control and Prevention.
Prospective survey
All participants were interviewed using a uniform questionnaire to collect information on basic demographic characteristics, economic burden, and health burden, and admission record review was used to assess inpatients’ direct medical cost and expense afforded by medical insurance.
Demographic characteristics. Demographic characteristics were gender, age, degree of education, and profession, among others.
Economic burden. Economic burden included direct, indirect, and intangible economic burden. Direct economic burden was regarded as the sum of direct and indirect medical cost. Direct medical cost comprised expenses of outpatient services/visits, pharmaceutical drugs, nursing care, and other factors directly related to diagnosis and treatment; indirect medical cost covered transportation, nutritional services, and others [Reference Xie12]. Indirect economic burden monetarily assessed productivity loss of patients and individuals caring for patients [Reference Tarricone13]; however, we assumed that patients aged >65 years and their parents, wives/husbands, or friends caring for them did not lose work time due to being retired. The indirect economic burden was equal to disposable income per capita in Dongtai City in 2014 (365·25 × productivity loss of the patient and individuals caring for the patient). We assumed that the value of disposable income per capita in Dongtai City in 2014 was approximately equal to that in 2015; intangible economic burden was calculated by converting ‘disvalue’ to an individual from pain, anxiety, and fear to monetary values based on the willingness-to-pay method [Reference Ogwulu14, Reference Wundes15]. Moreover, Consumer Price Index (CPI) values obtained from the website of Yancheng City Bureau of Statistics were used to convert direct and intangible costs that occurred in other years into 2015 RMB, and RMB was converted to US dollars ($) with an exchange rate of 6·2284 based on data from the National Bureau of Statistics of China.
Health burden. The EQ-5D-3L Chinese version was used to evaluate health burden of hepatitis E patients [16], EQ-5D-3L includes the descriptive system and the visual analogue scale (EQ-VAS) [17]. Quality-adjusted life year (QALY) and VAS score were indices of health burden. QALYs were gained from the conversion table for Japan (EQ-5D-JP) using time trade-off (TTO)-based preference scores [Reference Tsuchiya18].
Statistical analysis
EpiData v. 3.1 (EpiData Association, Denmark) and SPSS v. 19.0 (IBM Corp., USA) were used for data management and analysis. Incidence rate was standardized using the age data of the rural population in the Sixth National Population Census provided by The National Bureau of Statistics of China. Missing data of economic burden was completed with the median data, we assumed that outpatients expended $0 where questions concerning economic burden were left blank; economic burden of inpatients was completed by the median cost of subjects with the same hospital and hospitalization time. Median and interquartile range (IQR) were used to describe concentrative and discrete trends of quantitative data. Wilcoxon and Kruskal–Wallis H rank tests were used to compare disease burden between categorical subgroups when the outcome is not normally distributed. P < 0·001 was considered statistically significant.
RESULTS
Incidence of hepatitis E
A total of 489 hepatitis E cases were investigated in the active hepatitis surveillance system between 2008 and 2015. The eight years’ cumulative incidence rate was 102·7/100 000, and the age-standardized cumulative incidence rate was 107·9/100 000. Moreover, the median annual incidence rate was 12·6/100 000, and the median age-standardized annual incidence rate was 16·5/100 000 (Fig. 1).
Fig. 1. Incidence rate of hepatitis E in a rural region of China. The x-axis corresponds to the onset time (year), the y-axis corresponds to the annual incidence rate.
Characteristics of hepatitis E patients
The median age of hepatitis E patients was 58·0 (IQR 48·5–67·0) years. The male-to-female ratio was 2·3:1, and 67·9% (332/489) of HEV-infected patients were admitted to hospital for treatment. Moreover, 75·3% (368/489) of patients were farmers, 84·5% (413/489) of infected patients had a low level of school education (no more than primary school education; grades 1–9), and 93·5% (457/489) of cases were involved in the New Rural Cooperative Medical Scheme.
Economic burden (Table 1)
Direct economic burden. The median direct medical cost was $818·8/case, and the median direct medical cost for an outpatient and inpatient were $19·3 and $1813·8, respectively; cost of prescription drugs accounted for 70·4% of hospitalization expenses, and 54·8% of hospitalization expenses was reimbursed by medical insurance. Moreover, the median indirect medical cost for an outpatient, inpatient, and hepatitis E case were $0·7, $241·7, and $110·1, respectively, and 82·6% of indirect medical costs came from nutritional services. The median direct economic burden for one patient was $1046·0, which accounted for 29·2% of disposable income per capita in Dongtai city. The ratio of direct medical cost to indirect medical cost was 7·4:1, and there were statistically significant differences between gender and types of cases.
Table 1. Economic burden for hepatitis E
Values given are median (interquartile range).
* Wilcoxon rank test was used to analyse economic burden.
† Kruskal–Wallis H rank test was used to analyse economic burden.
Indirect economic burden. The median duration of work loss was 5 days/case, and the median hospitalization time of inpatients was 16·5 days. The median indirect cost for one HEV-infected case was $49.1, which accounted for 1·4% of disposable income per capita in Dongtai city. There was a significantly higher number of inpatients than outpatients; moreover, the cost of indirect economic burden differed statistically significantly in age groups.
Intangible economic burden. The median intangible cost was $77·3/case, which accounted for 2·2% of disposable income per capita in Dongtai city. There were statistically significant differences between inpatients and outpatients.
In total, the median economic burden in a patient infected with HEV was $1836·5, which accounted for 51·2% of disposable income per capita in Dongtai city. The economic burden of inpatients was significantly higher than that of outpatients. Further, a multivariate linear regression model indicated inpatient/outpatient and age were significant factors for economic burden (gender, age, onset time, inpatient/outpatient and township were included in the analysis) (P < 0·001).
Health burden (Table 2)
The median QALY and VAS score were 0·7 and 70·0/patient, respectively. Moreover, QALY and VAS scores had a positive correlation (r s = 0·521, P < 0·001). Health burden was greater for inpatients than outpatients. The health burden in older patients aged >60 years was higher than for other ages (QALY: 0·7 vs. 0·7, z = −3·219, P = 0·001; EQ-VAS score: 60·0 vs. 70·00, z = −3·962, P < 0·001). A multivariate linear regression model showed inpatient/outpatient to be significant factors for QALY and VAS score (gender, age, onset time, inpatient/outpatient and township were included in the analysis) (P < 0·001).
Table 2. Health burden for hepatitis E
QALY, Quality-adjusted life years; VAS, visual analogue scale.
Values given are median (interquartile range).
* Wilcoxon rank test was used for binary categorical variables; Kruskal–Wallis H rank test was used for multi-categorical variables.
DISCUSSION
Hepatitis A and E are two types of enterically transmitted viral hepatitis; however, hepatitis A is well prevented and controlled by universal hepatitis A vaccination, and hepatitis E accounts for a higher proportion of acute viral hepatitis [Reference Zhang19]. Apart from this, the prevalence of anti-HEV IgG in the population is lower than the level observed for hepatitis A [20]. Surveys on the disease burden of hepatitis E are useful for comprehending the severity of the disease and being aware of the necessity of prevention and control measures in order to make more reasonable and effective public health decisions.
From 2008 to 2015, 489 HEV-infected cases were reported in the active hepatitis surveillance system in Dongtai city, China. The median annual incidence rate was 12·6/100 000, which was lower than in 2006–2007 [Reference Zhu11], because a recombinant hepatitis E vaccine with long-term efficacy was used after August 2007 [Reference Zhu10, Reference Zhang21]; additionally, the incidence in 2015, shown in Figure 1, was underestimated due to the uncompleted 1-year survey.
The median direct, indirect, intangible, and total economic burden were $1046·0, $49·1, $77·3, and $1836·5/case, respectively, which accounted for 29·2%, 1·4%, 2·2%, and 51·2% of per capita disposable income. This suggests that more reasonable measures should be implemented to prevent and control hepatitis E, because the economic burden of hepatitis E accounted for more than 50% of disposable income per capita. Moreover, direct medical cost accounted for 78·3% of direct economic burden, and 54·8% of hospitalization cost would be reimbursed through medical insurance systems [Reference Pan22–Reference Mahmood24], which is higher than in previous studies [Reference Xiu8, Reference Jiang9]. Thus, the scientific and effective health insurance systems are also an important factor in reducing the economic burden of hepatitis E from the patients’ perspective.
In this study, EQ-5D-3L was used to measure the health burden of hepatitis E patients, and the median QALY and VAS score per patient were 0·7 and 70·0, respectively. QALY, which reflects the opinion of the general population, is calculated by health state using TTO valuation techniques, but VAS records the patient's own assessment of their health status, which reflects a minor change in the quality of life of patients [17]. The results of EQ-VAS might be unreliable in this survey because patients with a low level of education might encounter difficulties in adequately understanding the definition of VAS, which could cause a low correlation coefficient between QALY and VAS scores. However, comprehension of the questionnaire was increased based on rapid development of media and close attention to the individual's own health. Older adults make up the high-risk population for HEV genotype 4 infection [Reference Khuroo3, Reference Kamar25], and the proportion of older patients (aged >60 years) reached 45·6% in the survey; furthermore, the median health burden in older patients is higher than in those of other ages. Therefore, older adults represent one targeted population group to prevent and control hepatitis E by health decision makers.
There are some limitations to this study. First, participants were recruited from one hepatitis E-endemic rural region in China where unreported cases were not considered [Reference Jia26], which might not represent the disease burden of hepatitis E in other areas, such as Beijing and Nanjing cities. Second, epidemiological features of hepatitis E cases are not sufficiently shown due to insufficient demographic data for registered residents recruited in the active hepatitis surveillance system. Third, more scales, such as the SF-6D [Reference Jin27] and EQ-5D-5L [Reference Jia28] should be used to confirm the most reasonable tool to measure the health burden of hepatitis E patients.
Currently, studies on disease burden of viral hepatitis mainly focus on hepatitis A, B, and C [Reference Steinke29, Reference Shon30]; however, the hepatitis E vaccine sold in China is safe and effective [Reference Zhu10, Reference Zhang21], an effective immunization strategy is worth considering to prevent and control hepatitis E, and studies on disease burden concerning hepatitis E would provide rational evidence for a dominant immunization strategy.
ACKNOWLEDGEMENTS
This study was supported by the National Natural Fund of China (81573258), the Science & Technology Demonstration Project for Emerging Infectious Diseases Control and Prevention (BE2015714) and the Social Development Fund of Jiangsu Province (BE2013723).
DECLARATION OF INTEREST
None.
