Socioeconomic inequality in Hepatitis B vaccination of rural adults in China
Dawei Zhua, Na Guob, Jian Wangc*, Stephen Nicholasd, Zhen Wange, Guojie Zhangf, Luwen Shig, Knut Reidar Wangenh
a Center for Health Policy and Management, Institute of Medical Information & Library, Chinese Academy of Medical Sciences & Peking Union Medical College, No. 3 Yabao Road, Chaoyang District, 100020, Beijing, China;
b China Population and Development Research Center, No. 12 Dahuisi Road, Haidian District, Beijing 100081, China;
c School of Public Health, Shandong University, No. 44 Wenhuaxi Road, Lixia, District, 265400, Jinan, China;
d School of Economics and School of Management, Tianjin Normal University, Tianjin, China;
Beijing Foreign Studies University, Beijing, China; Research Institute of International Strategies, Guangdong University of Foreign Studies, Guangzhou, China; Newcastle Business School, University of Newcastle, Newcastle, NSW, Australia;
e Qingdao center Hospital, Qingdao, China;
f Peking Union Medical College Hospital, No.1 Shuaifuyuan Wangfujing Dongcheng District, 100730 Beijing, China;
g School of Pharmaceutical Sciences, Peking University, No. 38 XueyuanLu Road, 100191, Beijing, China;
h Department of Health Management and Health Economics, University of Oslo, Oslo, Norway.
*Corresponding author:
Jian Wang
Address: School of Public Health, Shandong University, No. 44 Wenhuaxi Road, Lixia, District, 265400, Jinan, China;
Tel: +86 13864157135
E-mail: wangjiannan@sdu.edu.cn
Abstract
Hepatitis B (HB) vaccination is the most effective way to prevent HB virus infection. While measures taken to control the prevalence of HB have achieved significant results, HB prevalence in rural China among adults remains problematic. This study sheds new light on the determinants of HB vaccine uptake and its inequality according to socioeconomic status in rural areas of China. We interviewed 22,283 adults, aged 18-59 years, from 8444 households, in 48 villages from 8 provinces. Vaccination status was modeled by using two logistic models:
whether take at least one HB vaccine and whether to complete the entire vaccination regime.
The Erreygers’ concentration index (𝐸𝐶𝐼) was used to quantify the degree of inequality and the decomposition approach was used to uncover the determinants of inequality in vaccine uptake.
We found that the coverage rate of HB vaccination is 20.2%, and the completion rate is 16.0%.
The 𝐸𝐶𝐼 of at least one dose (0.081) and three doses (0.076) revealed a substantial pro-rich inequality. Income contributed the largest percentage to HB vaccination inequalities (52.17%
for at least one dose and 52.03% for complete vaccinations). HB awareness was another important cause of inequality in HB vaccination (around 30%). These results imply that rich had a greater tendency to vaccinate and inequality favouring the rich was almost equal for the complete three doses. While the factors associated with HB vaccination uptake and inequalities were multifaceted, income status and HB awareness were the main barriers for the poor to take HB vaccine by adults in rural China.
Key words: Hepatitis B vaccination; Inequality; Concentration Index; Decomposition; China
Introduction
Hepatitis B virus (HBV) infection is a global public health issue, infecting over 240 million people worldwide and causing over 700 thousand deaths per year. Attacking the liver, HBV leads to both acute and chronic disease, especially in children1. China has a high prevalence for HBV, with over 120 million HBV virus carriers and about 300 thousand deaths per year due to HBV infection and HBV related diseases2. Vaccination is the most effective way to prevent HBV infection. The recommend standard schedule for hepatitis B (HB) vaccination consists of three doses, given within the first 24 hours of birth (timely birth dose), month 1 and month 6 after birth. In 1992, the World Health Organization recommended global vaccination against HBV as part of the Expanded Program on Immunization3.
China’s HBV vaccination strategy has achieved remarkable results with more than 200 million newborn children inoculated. The uptake rates increased gradually and reached high levels after 2005, when the implementation of a new policy offered infant HBV vaccine at no cost to the families. Prior to this policy, families had to cover vaccination service fees (2002-2004), or both service fees and the cost for the vaccine doses (1992-2001). Due to relatively low uptake rates in early birth cohorts, a catch-up vaccination program was implemented in 2009-2010 for pre- immune children who were born between 1994 and 2001 (4). General HBV vaccination of adults has not been prioritized yet. Among adults, HBV incidence is still high and with increasing age, the HBV infection rate is increasing while the HBV vaccination rate is sharply decreasing (5, 10). Studies have shown that new hepatitis B inpatients were mainly concentrated in the age group 15 to 44 years old (6-9), and that the HBsAg carrier rate of adults has not
decreased significantly. A HBV vaccination program for adults would have important health benefits and would most likely be cost-effective (11). In addition, free vaccination offered to the general adult population would contribute to health equity due to notable social gradients, both in the burden of HBV related disease and in the present HBV vaccination coverage (15,12).
Various studies have analyzed the factors associated with HBV vaccination coverage of adults, and found that vaccination-related costs (such as user fees and travel costs), and socioeconomic status (measured by income, education and occupation) are factors affecting the adults HBV vaccine uptake12-14. Recent research has found that lower socioeconomic status is linked to lower vaccine uptakes in adults and is one of the most important factors affecting the vaccination completion rates in adults12-15. In other words, lower socioeconomic status points to relatively lower vaccination rates, a major cause of health inequality. In some research, the effect of socioeconomic inequalities on vaccine uptake was found to be more important than other factors relating to vaccinations, such as beliefs and attitudes towards immunization16. On the whole, people with low socioeconomic status are more likely to be infected with HBV15.
While revealing the socioeconomic inequality in HBV vaccination rates among adults, these studies did not measure the magnitude of the inequality, limiting the ability to make comparisons across time periods and regions and hampering the identification of effective policies to target low HBV vaccination uptake. To address this lacuna, we use the concentration index (CI) and its decomposition method not only to provide a measure of inequality, but also to identify the contributing factors of the inequality. Derived from the more well-known Gini coefficient, the CI shows how a health outcome, such as vaccinations, varies according to
measures of socioeconomic status, such as income, providing a single value of any income related inequality18. Because the CI is derived from the Gini coefficient of income inequalities, it requires the health variable to be on the same scale as income, i.e. on a ratio-scaled measure without an upper bound19. Since health measures tend to be bounded and either ordinal or cardinal, a variety of concentration indices have been proposed to suit the measurement properties of the variable, such as Generalized CI20, Wagstaff’s normalization CI (𝑊𝐶𝐼)21, and Erreygers’ normalization CI (𝐸𝐶𝐼) 19. The CI can be decomposed into the contributions of individual determinants to socioeconomic health inequality to allow elements that drive the income-related inequality to be identified separately, where each contribution is the product of the sensitivity of heath with respect to that factor (the elasticity) and the degree of income- related inequality in that factor (the respective concentration index).
Understanding the determinants of inequality is an important step in developing effective policy interventions. Using the 𝑊𝐶𝐼 of non-receipt of vaccination in Ireland, and results from the decomposition analysis, Doherty et al. (2014) found that 85% of the inequality was explained by household level variables, such as the socioeconomic status, household structure, and income22. Lauridsen and Pradhan (2011) found that the CI value for a child not fully immunized was -0.15 at the national level in India, and decomposition results revealed that the majority of the inequality was explained by poor household economic status (38%) and mother’s illiteracy (35%)23. As far as we are aware, there has been no study that has employed the CI decomposition technique to quantify the principal determinants of inequality in HBV vaccine uptake among rural adults in China. The only article to analyse the inequality in the hepatitis B awareness level found the CI of hepatitis B awareness was 0.14 in rural China and the
contribution rate of socioeconomic factors was the largest element (60.8%).24. The present study uses CI to measure the extent of inequality in HB vaccine uptake and shed light on the role that socioeconomic factors (measured as income, education and occupation) and other factors play in explaining vaccine uptake of adults in rural China.
The paper uses data from a household survey, covering 22,282 individuals from 8,444 households, in 48 villages from 8 provinces. The paper quantifies the extent of socioeconomic gradient in adults’ vaccination using a CI and decomposes the determinants of inequality in adults’ uptake using decomposition analysis.
Results
Characteristics of the respondents
Table 1 presents descriptive statistics for all independent variables in the analysis. A total of 70.3% of the sample were under 45, and 49.6% were female. Most have a low or medium education (years of schooling less or equal to 9 years), and half were farmers. The income group variables were defined by five dichotomous variables to indicate how each responder’s income was related to the quintiles of the income distribution. The participators were asked how a person could be infected by HBV. Alternatives consisted of 3 actual transmission routes (A child may be infected during birth if the mother is infected; Use of unclean medical or dental equipment; Unprotected sex). True transmission route index counted the number of authentic transmission routes that were identified (range: 0–3), and respondents identified two true transmission routes on average. The variable “Perceived protection” was divided by how long the participants believe the protection of HBV vaccine remain effective and was divided into
five categories that classified the responders according to how long they believed the protection lasted (perceived protection 1–5). Forty three percent (43.0%) of subjects reported that vaccine’s protection was unknown or perceived protection last 0-1 year, followed by those report 1-5 years (33.5%). The average vaccine fee and service fee per dose was 33 Chinese Yuan (1 CNY = 0.158 USD in 2012). Average time spent traveling or waiting per dose was 28 minutes and round-trip cost of travel to a health facility was 4 CNY per dose.
The inequality of vaccination uptake according to socioeconomic status
In our sample, 20.2% of respondents received at least one dose and 16.0% received three doses of the HB vaccine. Concentration indices for HB vaccination were shown in Table 1. The positive CIs means a greater tendency for the rich to vaccinate and vice versa. Though the CI in this study lies in the interval [𝜇 − 1, 1 − 𝜇] due to the vaccination uptake is a binary variable, 𝑊𝐶𝐼 and 𝐸𝐶𝐼 vary between -1 and +1. The results confirm a strong significant socioeconomic gradient in rural adult HBV vaccinations and indicate that the rich had a greater tendency to take-up vaccinations than the poor. The results of 𝐸𝐶𝐼, which is an absolute inequality index as it is independent of the coverage rate, show that levels of absolute inequality for whether take vaccine and complete vaccination were almost equal.
Determinants of vaccination uptake
The first column of Table 3 and Table 4 shows the determinants of whether to take the vaccine and whether to complete all three vaccination stages. The results in Table 3 and Table 4 were broadly the same for whether to take at least 1 dose or all doses, so we explain them together.
Vaccination was more likely to occur for respondents in higher income quintiles, higher education levels and non-farm occupations. Other non-socioeconomic factors increasing the probability of vaccination were number of identified true routes of transmission and perceived protection from vaccination. The factors decreasing the probability of vaccination were the cost of the vaccine, vaccination service fee, total time spent obtaining the vaccination and travel costs. The age variables show that older age groups were less likely to be vaccinated than younger age groups. Gender was not significant in the analysis.
Decomposition of income inequality in the vaccination
Also displayed in Table 3 and Table 4 is the decomposition of Erreygers’ normalization concentration index, which shows how income inequality impacted the vaccine uptake. The CI column demonstrates each variable’s distribution across wealth. The CIs of total time spent travelling or waiting, travel costs and other occupations were negative, which means these variables were concentrated among people with lower economic status. In contrast, vaccination fee, gender, education level, occupation, true routes of transmission and protection had a positive concentration index, indicating these variables were concentrated among the rich.
The column for contribution shows the absolute contribution of each determinant variable to economic inequality. The results show that income contributed the largest percentage to HBV vaccination inequalities, accounting for 52.17% of at least one dose and 52.03% of all three doses. For at least one and for all doses, the contributions of education level (13.51% and 12.90%), occupation (14.58% and 12.80%), true routes of transmission (13.81% and 11.40%) and protection (20.55% and 17.43%) were positive, which means these variables exacerbated
inequality by contributing to a higher vaccination uptake among the rich. Contributions of vaccination fee (-11.35% and -10.25%) and age (-17.06% and -16.21%) were negative, which means these variables contributed to inequality through lower vaccination uptake by the rich.
Factors such as time (-0.65% and -0.40%) and gender (-0.29% and -0.28%) explained an insignificant percentage of the inequalities.
Discussion
Conducting an 8 provinces individual-level survey, this study shed light on the socioeconomic determinants of HBV vaccine uptake and uptake inequality in rural China. The research used the concentration index and decomposition method to undertake an in-depth analysis of economic inequality on HBV vaccination rates of rural adults. In our sample, 20.2% received at least one dose and 16.0% received three HBV doses, which was higher than a Beijing city25 study hepatitis B vaccination rate of 8.4% (Wang et al., 2010), but consistent with Zhu et al.
(2014) adults’ coverage rate of 23.7% in Jiangsu province10.
For rural China adults, the Erreygers’CI of at least one dose (0.081) and all three doses (0.076) means that the rich had a greater tendency to vaccinate and inequality favouring the rich was almost equal for the complete three doses. Further, high education levels, HBV knowledge and knowledge of HBV transmission were concentrated among the rich, which further contributed to inequality. Also, payment capacity was higher within the higher income group.
Our study found that the vaccine and vaccination service fee, time spent travelling and waiting, and travel costs decreased the probability of vaccination. These effects can be interpreted as the households’ direct and indirect vaccination costs. In Zhu et al. (2014) it is argued that in
particular the effects of direct vaccination costs could represent interactions of demand and supply side factors12. They argued that high user fees increased the supplier’s ability to offer the vaccination to more individuals, but higher travel costs and longer waiting times did not benefit the vaccine suppliers. One approach designed to isolate the effect of price on the demand is the contingent evaluation method. This method has been used to estimate the households’ or individuals’ willingness to pay for vaccines, and the results indicate that vaccination demand is price sensitive36,37. Table 2 and 3 show that the contribution of user fee was negative for both at least one dose and all three doses, which reduced inequality in vaccination. The positive CI of user fee indicates that the poor usually paid a lower price than the rich. This can be explained by the price of domestic versus imported hepatitis B vaccines varying considerably. Since it was assumed that imported vaccines are higher quality than local vaccines, the rich preferred the higher cost imported HB vaccine. In contrast, the poor were price sensitive, selecting cheaper domestic vaccines. Also, the lower user fee increased the demand for vaccinations among the poor. While the contributions were relatively small, indirect time and travel costs were positive, which increased inequality through higher vaccination uptake among the rich.
Similar to user fees, high indirect costs tend to decrease the demand for vaccinations. But the CI was negative, which reflects the fact that the poor usually live far from vaccination sites with low accessibility.
Socioeconomic and demography variables such as income, age, education level, and occupation were associated with vaccination utilization and consistent with previous studies38,39. Also the number of identified true routes of transmission and perceived protection were associated with vaccination utilization. The influence of income increased the probability of vaccination, due
to the ability to pay for the vaccine and the vaccination service26. Age had a negative impact on taking vaccinations. One reason could be that the young were better educated and had better jobs than the older age groups that increased the probability of vaccination. The high educational level group had greater knowledge of hepatitis B infections, such as the number of identified true routes of transmission27, understood better the perceived protection of vaccinations, and paid more attention to their health, so are more active in attaining vaccinations25.
The contributions of income, education level, true routes of transmission and protection were positive, which means these variables increased inequality in terms of higher vaccination uptake among the rich. These variables were concentrated in the rich (the CIs were positive), and explained the positive proportion in the vaccine uptake. With the effect of these two aspects, the coverage gap between rich and poor was greater. Income had the largest contribution, which means income status was the main barrier for the poor to take-up the HB vaccine. Hepatitis B awareness made the second most important contribution, with the number of identified true routes of transmission and perceived protection together accounting for around 30 percent of the higher rich uptake. The results confirm a strong socioeconomic gradient in hepatitis B awareness, and an important cause of inequality in hepatitis B vaccinations. To reduce these inequalities, HB-related health education targeting individuals with low socioeconomic status should be performed.
Our study have the following limitations: First, information on costs in our data relied on self- reporting and was subject to recall and reporting errors. Second, this was a cross-sectional study.
Future studies should collect longitudinal data to better understand the categories of determinants and measure the changes in inequalities. Third, our sample included the birth cohorts 1992 and 1993 which, in principle, could have been offered HBV vaccination as part of the national child vaccination program. However, we find these birth cohorts comparable to the older cohorts, both because the families had to cover the full costs of HBV vaccination and the coverage rates for the vaccination of newborn were generally low in rural areas, and because these cohorts were not included in the national catch-up vaccination program.
Conclusion
This paper provides strong evidence of a substantial economic gradient in adults’ HBV vaccination in China. The results of the paper suggest that the factors associated with HBV vaccination uptake and inequalities are multifaceted. Factors such as income, user fee, education level, hepatitis B awareness play a large role in explaining inequalities in at least one vaccination and the complete vaccination routine of rural adults. While the factors associated with HB vaccination uptake and inequalities were multifaceted, income status and HB awareness were the main barriers for the poor to take HB vaccine by adults in rural China.
Materials and methods
Sampling method and sample selection
We collected data using a household survey in 48 villages from 8 provinces comprising Hainan, Hebei, Heilongjiang, Henan, Jiangsu, Ningxia, Beijing, and Shandong, with notable regional, economic, and epidemiological diversity. Counties within each province were stratified by level
of economic development (low, medium, high), and villages in a county were stratified based on short, medium, and long travel distance to the vaccination sites. In small villages, all households were invited while in larger villages, households were randomly selected using household size as sampling weights (probability proportionate to size). The sampling strategy ensured that two presumably important factors, income and travel distance, showed substantial variation in the resulting sample. In the planning process, we balanced the expected benefits of including more villages and obtaining more between-village variation (user fee, time, and travel cost vary little within each village) against the managerial costs of surveying more villages. The final sample consisted of 22,282 adults, aged 18-59 years (born in or before 1993), from 8,444 households.
The interviews were conducted by trained staff and took place in 2011 and 2012. The households were interviewed based on questionnaires that included questions about all household members’ vaccination history, individual and household characteristics, and knowledge of HBV. A pilot survey was carried out before starting the field survey. All data were double inputted using Microsoft Access and checked for consistency.
Measures
Information collected in the survey included gender, age, education level, occupation, household income and information about vaccination (protection, vaccine fee and so on).
Sample descriptive statistics for the key variables are provided in Table 1. Vaccination status is a categorical variable with three outcomes: unvaccinated, received at least 1 dose, and complete vaccination (receive 3 doses). The income variable divided the households into equally sized
groups by quintiles in per capita annual income. Age was divided into three groups, 18-29, 30- 44 and 45-59. The sex of the individual used female as the reference; the education levels were divided into three groups by years of schooling; the occupations were farmer, migratory workers, student and other occupation. True transmission route index counted the number of actual transmission routes that were identified. The perceived protection is defined by how long the participants believe the protection would remain effective if they took the vaccine. The user fee was measured on the village level and was defined as the village-specific average user fee paid by adults who had been vaccinated. The same value was assigned to all adults who lived in the same village, regardless of individual vaccination status. The relative frequencies of villagers who had been vaccinated at available sites (e.g., village, county, and township clinics) were used as weights. The variables travel cost and time were defined in the same manner; the former represents the cost of a round trip, while the latter is the total time used to travel to the vaccination site and the waiting time. Both user fee and travel cost were measured in 10 CNY (about 1.58 USD in 2012) intervals.
Concentration index
The Concentration Index is a common index for measuring inequality in health service28-32. The CI is defined as twice the area between the concentration curve and the line of equality (the 45- degree line). The general form of CI formula18 is:
𝐶 = 2
𝑛𝜇∑𝑛𝑖=1ℎ𝑖𝑟𝑖− 1 = 2
𝜇𝑐𝑜𝑣(ℎ𝑖, 𝑟𝑖) (1)
in which 𝑟𝑖 is the rank of economic status, ℎ𝑖 is whether the individual take vaccination, 𝜇
is the mean of h in the sample, and cov is the weighted covariance. Economic status was determined by household’s per capita income in this study. In general, the concentration index can vary between –1 and +1, where a positive value means a greater tendency for the rich to take vaccinations and a negative value indicates pro-poor inequality. In our case, in which the health care variable is binary, Wagstaff (2005) demonstrated that the CI is restricted to the narrower interval between (𝜇 − 1 ) and (1-𝜇 ). We followed Wagstaff’s suggestion for normalizing the CI:
𝑊𝐶𝐼 = 𝐶
1−𝜇 (2) We also applied a formula suggested by Erreygers (2009)19, intended for cases with ordinal or binary health care variables19,
𝐸𝐶𝐼 = 4𝜇𝐶
𝑏−𝑎 (3) Here (b -a) is the range of the variable of interest, equal to 1 in our study.
Decomposition analysis
Wagstaff et al. (2003) and Hosseinpoor et al. (2006) have shown that the CI can be decompose into its determinant variables to explain how they contribute to inequality33-35. The formula can be written as:
𝐶 = ∑ (𝛽𝑘𝑋𝑘
𝜇 ) 𝐶𝑘+𝐺𝐶𝜀
𝑘 𝜇 (4) Where βk, 𝑋𝑘, and 𝐶𝑘 are the coefficient, mean, and concentration index of Xk, respectively;
and 𝐺𝐶𝜀 is a generalized concentration index for the error term 𝜀 . The elasticity 𝛽𝑘𝑋𝑘
𝜇 of variable (h) with respect to Xk is the measure of how responsive h is to a change in another Xk, and (𝛽𝑘𝑋𝑘
𝜇 ) 𝐶𝑘 measures the contribution of Xk to CI.
The Erreygers’ concentration index (𝐸𝐶𝐼) can be decomposed as:
𝐸𝐶𝐼 = 4 ∗ [∑ 𝛽𝑘 𝑘𝑋𝑘𝐶𝑘+ 𝐺𝐶𝜀] (5)
Equation (5) reveals that the effect of any variable x on 𝐸𝐶𝐼 depends both on its own generalized concentration index (𝐶𝑘 multiplicated by its mean 𝑋𝑘) and on the marginal effect of the variable on the vaccination. Since the dependent variables in this study are binary variables, we estimate the 𝛽𝑘s from a logistic regression and we calculate partial effects of the logistic estimates of each independent variable at sample means. To obtain standard errors for the concentration indices and for the contributions of each explanatory variable, we use a non- parametric bootstrap method with 1000 replications. All statistical analyses were performed in STATA 14.0.
Ethics and institutional approval
Participation was voluntary and potentially sensitive questions were not included in the questionnaire. All study participants were informed about their right to refuse to answer any question. The project was approved by the Medical Ethics Committee at the Shandong University School of Medicine (Grant No. 201001052).
Abbreviations
HB: Hepatitis B;
CI: concentration index;
𝑊𝐶𝐼: Wagstaff’s normalization CI;
𝐸𝐶𝐼: Erreygers’ normalization CI;
HBV: Hepatitis B virus;
Disclosure of potential conflicts of interest
None of the authors has any conflicts of interest.
Funding
This study was partly funded by the Norwegian Research Council under Grant no. 196400/S50.
Reference
1. Kao JH, Chen DS. Global control of hepatitis B virus infection. Lancet Infect Dis 2002;
2:395-403.
2. Liang X, Bi S, Yang W, Wang L, Cui G, Cui F, Zhang Y, Liu J, Gong X, Chen Y, et al.
Epidemiological serosurvey of hepatitis B in China-declining HBV prevalence due to hepatitis B vaccination. Vaccine 2009; 27:6550-6557.
3. Arístegui J, Usonis V, Coovadia H, Riedemann S, Maung Win K, Gatchalian S, Bock HL.
Facilitating the WHO expanded program of immunization: the clinical profile of a combined diphtheria, tetanus, pertussis, hepatitis B and Haemophilus influenzae type b vaccine. Int J Infect Dis 2003; 7:143-151.
4. Jia Y, Li L, Cui F, Zhang D, Zhang G, Wang F, Gong X, Zheng H, Wu Z, Miao N, et al.
Cost-effectiveness analysis of a hepatitis B vaccination catch-up program among children in Shandong Province, China. Hum Vaccin Immunother 2014; 10:2983-2991.
5. Zhang M, Yuan Y, Mao P, Zhuang Y. Analysis on morbidity and mortality of viral hepatitis in China, 2004-2013. Zhonghua liu xing bing xue za zhi [Chinese journal of epidemiology]
2015; 36:144-147. [In Chinese]
6. Chen P, Yu C, Ruan B, Yang S, Ren J, Xu W, Luo Z, Li L. Prevalence of hepatitis B in insular regions of southeast China: a community-based study. PloS one 2013; 8:e56444.
7. Hu Y, Yu C, Chen B, Wang L. Implication of reported viral hepatitis incidence rate change in Hubei Province, China, between 2004-2010. J Huazhong Univ Sci Technolog Med Sci 2012;
32:428-433.
8. Luo Z, Xie Y, Deng M, Zhou X, Ruan B. Prevalence of hepatitis B in the southeast of China:
a population-based study with a large sample size. Eur J Gastroenterol Hepatol 2011; 23:695- 700.
9. Zhang L, Yan BY, Li MS, Song LZ, Lu JJ, Xu Q, Xu A. Preliminary analysis on the prevalence and causes of breakthrough hepatitis B virus infection among children in Shandong province, China. Zhonghua yu fang yi xue za zhi [Chinese journal of preventive medicine] 2013;
47:933-939. [In Chinese].
10. Zhu L, Zhai X, Zhu Y, Xu W, Bao C, Peng H, Bian Q, Yang H, Wang H, Hu Z, et al.
Evaluation of the impact of hepatitis B vaccination in adults in Jiangsu province, China. PloS one 2014; 9:e101501.
11. Hu R, Cao W, Zhang X. Cost-effectiveness analysis of hepatitis B vaccination in People's Liberation Army. Zhonghua liu xing bing xue za zhi [Journal of epidemiology] 2001; 22:142- 145. [In Chinese].
12. Zhu D, Wang J, Wangen KR. Hepatitis B vaccination coverage rates among adults in rural China: are economic barriers relevant? Vaccine 2014; 32:6705-6710.
13. McGregor J, Marks PJ, Hayward A, Bell Y, Slack RC. Factors influencing hepatitis B vaccine uptake in injecting drug users. J Public Health Med 2003; 25:165-170.
14. Lu PJ, Byrd KK, Murphy TV, Weinbaum C. Hepatitis B vaccination coverage among high- risk adults 18-49 years, U.S., 2009. Vaccine 2011; 29:7049-7057.
15. Stuver SO, Boschi-Pinto C, Trichopoulos D. Infection with hepatitis B and C viruses, social class and cancer. IARC scientific publications 1997:319-324.
16. Danis K, Georgakopoulou T, Stavrou T, Laggas D, Panagiotopoulos T. Socioeconomic factors play a more important role in childhood vaccination coverage than parental perceptions:
a cross-sectional study in Greece. Vaccine 2010; 28:1861-1869.
17. Stuver SO, Boschi-Pinto C, Trichopoulos D. Infection with hepatitis B and C viruses, social class and cancer. In: Kogevinas M, Pearce N, Susser M, Boffetta P, editors: Social inequalities and cancer. Lyon (France): International Agency for Research on Cancer, IARC scientific publications 1997; 138: p. 319-324.
18. Eikemo TA, Bambra C, Joyce K, Dahl E. Welfare state regimes and income-related health inequalities: a comparison of 23 European countries. Eur J Public Health 2008; 18:593-599.
19. Erreygers G. Correcting the concentration index. J Health Econ 2009; 28:504-515.
20. Clarke PM, Gerdtham U-G, Johannesson M, Bingefors K, Smith L. On the measurement of relative and absolute income-related health inequality. Soc Sci Med 2002; 55:1923-1928.
21. Wagstaff A. The bounds of the concentration index when the variable of interest is binary, with an application to immunization inequality. Health Econ 2005; 14:429-432.
22. Doherty E, Walsh B, O'Neill C. Decomposing socioeconomic inequality in child vaccination: results from Ireland. Vaccine 2014; 32:3438-3444.
23. Lauridsen J, Pradhan J. Socio-economic inequality of immunization coverage in India.
Health Econ Rev 2011; 1:1-6.
24. Zheng J, Li Q, Wang J, Zhang G, Wangen KR. Inequality in the hepatitis B awareness level in rural residents from 7 provinces in China. Hum Vaccin Immunother 2017:1-9.
25. Wang H, Zhang W, Jiang WU. Hepatitis B vaccination rate in adults of Beijing:a cross- sectional study. Chinese Journal of Public Health 2010; 26:612-614. [In Chinese].
26. Johri M, Perez MC, Arsenault C, Sharma JK, Pai NP, Pahwa S, Sylvestre MP. Strategies to increase the demand for childhood vaccination in low- and middle-income countries: a systematic review and meta-analysis. Bull World Health Organ 2015; 93:339-346c.
27. Yang W, Xin W, Mujun W, You-wei L, Guojie Z, Jian W, Wangen KR. Analysis on contribution of rural residents' Hepatitis B cognition's factors in the perspective of equity: An empirical research on 3 provinces. Chinese Journal of Health Policy 2012; 05:53-58. [In Chinese].
28. van Doorslaer E, Wagstaff A, Bleichrodt H, Calonge S, Gerdtham UG, Gerfin M, Geurts J, Gross L, Häkkinen U, Leu RE, et al. Income-related inequalities in health: some international comparisons. J Health Econ 1997; 16:93-112.
29. van Doorslaer E, Masseria C, Koolman X. Inequalities in access to medical care by income in developed countries. CMAJ 2006; 174:177-183.
30. Yang W. China's new cooperative medical scheme and equity in access to health care:
evidence from a longitudinal household survey. Int J Equity Health 2013; 12:20.
31. Chovar Vera A, Vasquez Lavin F, Paraje G. Inequality and inequity in the use of medical services in Chile, by age group, 2000-2011. Rev Panam Salud Publica 2014; 36:171-178.
32. Kavosi Z, Mohammadbeigi A, Ramezani-Doroh V, Hatam N, Jafari A, Firoozjahantighi A.
Horizontal inequity in access to outpatient services among Shiraz City residents, Iran. J Res Health Sci 2015; 15:37-41.
33. Hosseinpoor AR, Van Doorslaer E, Speybroeck N, Naghavi M, Mohammad K, Majdzadeh R, Delavar B, Jamshidi H, Vega J. Decomposing socioeconomic inequality in infant mortality in Iran. Int J Epidemiol 2006; 35:1211-1219.
34. Wagstaff A, van Doorslaer E, Watanabe N. On decomposing the causes of health sector inequalities with an application to malnutrition inequalities in Vietnam. J Econom 2003;
112:207-223.
35. van Doorslaer E, Jones AM. Inequalities in self-reported health: validation of a new approach to measurement. J Health Econ 2003; 22:61-87.
36. Islam Z, Maskery B, Nyamete A, Horowitz MS, Yunus M, Whittington D. Private demand for cholera vaccines in rural Matlab, Bangladesh. Health policy 2008;85:184-195.
37. Hou Z, Chang J, Yue D, Fang H, Meng Q, Zhang Y. Determinants of willingness to pay for self-paid vaccines in China. Vaccine 2014;32:4471–4477.
38. Park B, Choi KS, Lee HY, Jun JK, Park EC. Socioeconomic inequalities in completion of hepatitis B vaccine series among Korean women: results from anationwide interview survey.
Vaccine 2012;30:5844–5848.
39. Luna EJA, Veras MASM, Flannery B, De Moraes JC, Group tVCS. Household survey of hepatitis B vaccine coverage among Brazilian children. Vaccine 2009;27:5326–5331.
Table 1 Variable definitions and descriptive statistics (N=22282)
variable variable definition Mean SD
user fee Vaccine fee and service fee paid, per vaccine dose (10 Yuan)
3.310 1.834 Time Total time spent traveling or waiting, per
vaccine dose (hour)
0.462 0.160 Travel cost Round-trip cost of travel to health facility, per
dose (10 Yuan)
0.431 0.417
Age 18-29 Aged 18-29 0.362 -
Age 30-44 Aged 30-44 0.341 -
Age 45-59 Aged 45-59 0.297 -
Gender female Female 0.496 -
Gender male Male 0.504 -
Low education Years of schooling (y.o.s) are less or equal to 6 years
0.314 - Medium education Y.o.s is higher than 6 and less or equal to 9 years 0.496 - High education Y.o.s is higher than 9 years 0.190 -
Farmer Farmer 0.526 -
Migratory workers Migrant worker 0.261 -
Student Student 0.161 -
Other occupation If not Farmer, Migratory worker or Student 0.052 - Income group 1 Income in the bottom quintile 0.175 - Income group 2 Income in the second lowest quintile 0.184 - Income group 3 Income in the middle quintile 0.199 - Income group 4 Income in the second highest quintile 0.248 - Income group 5 Income in the top quintile 0.193 - True transm. route
ind.
No. of identified true routes of transmission 2.065 1.790 Perceived
protection 1
Vaccine is unknown or perceived protection last 0–1 year
0.430 - Perceived
protection 2
Perceived protection last 1–5 year 0.335 - Perceived
protection 3
Perceived protection last 5–10 year 0.099 - Perceived
protection 4
Perceived protection last 10–20 year 0.041 - Perceived
protection 5
Perceived protection last more than 20 years 0.095 -
Note: Dichotomous variables took the value 1 if the condition in the column “variable definition”
was satisfied and the value 0 otherwise. Thus, for these variables the reported mean value represents the proportions satisfying the mentioned conditions.
Table 2 Concentration indices of HB vaccination
CI 𝑊𝐶𝐼 𝐸𝐶𝐼
vaccination (at least 1 dose) 0.100(0.007)*** 0.125(0.010)*** 0.081(0.006)***
complete vaccination 0.119(0.009)*** 0.142(0.011)*** 0.076(0.006)***
Table 3 Decomposition analysis of Erreygers’ normalization concentration index of the vaccination (at least 1 dose) Coefficients Concentration indexes contribution to CI contribution rate
User fee -0.046(0.011)*** 0.114(0.002)*** -0.009(0.002)*** -11.346(2.503)***
Time -0.449(0.131)*** 0.005(0.001)*** -0.001(0.000)** -0.646(0.279)**
Travel cost -0.358(0.055)*** -0.005(0.003) 0.000(0.000) 0.537(0.371)
Age -17.059
Age 30-44 -1.087(0.044)*** 0.015(0.005)*** -0.003(0.001)*** -3.167(1.231)**
Age 45-59 -1.775(0.058)*** 0.061(0.006)*** -0.014(0.001)*** -16.771(2.480)***
Gender male -0.038(0.038) 0.023(0.004)*** 0.000(0.000) -0.288(0.305)
Education 13.507
Medium education 0.127(0.048)*** 0.031(0.004)*** 0.001(0.000)** 1.263(0.536)**
High education 0.570(0.059)*** 0.154(0.008)*** 0.010(0.001)*** 12.244(1.779)***
Occupation 14.577
Migratory workers 0.156(0.047)*** 0.064(0.007)*** 0.001(0.001)* 1.732(0.614)***
Student 0.414(0.054)*** 0.331(0.008)*** 0.013(0.002)*** 15.741(2.521)***
Other occupation 1.156(0.078)*** -0.054(0.016)*** -0.002(0.001)** -2.896(1.043)***
Income 52.166
Income group 2 0.137(0.065)** -0.466(0.005)*** -0.006(0.003)* -7.869(3.864)**
Income group 3 0.184(0.063)*** -0.082(0.006)*** -0.002(0.001)*** -2.030(0.745)***
Income group 4 0.264(0.062)*** 0.366(0.005)*** 0.013(0.003)*** 16.276(3.296)***
Income group 5 0.415(0.066)*** 0.807(0.003)*** 0.037(0.007)*** 45.789(5.940)***
True transm. route ind. 0.137(0.011)*** 0.075(0.003)*** 0.011(0.001)*** 13.805(1.658)***
Perceived protection 20.548
Protection 2 0.644(0.045)*** 0.048(0.005)*** 0.006(0.001)*** 7.195(1.104)***
Protection 3 0.543(0.065)*** 0.144(0.011)*** 0.005(0.001)*** 5.830(1.022)***
Protection 4 0.863(0.088)*** 0.073(0.019)*** 0.002(0.001)*** 2.143(0.654)***
Protection 5 0.598(0.066)*** 0.125(0.012)*** 0.004(0.001)*** 5.380(0.951)***
Note: *Two-sided p-value <0.1; **Two-sided p-value <0.05; ***Two-sided p-value <0.01. Standard errors (in parentheses) were obtained by non-parametric bootstrap.
Table 4 Decomposition analysis of Erreygers’ normalization concentration index of complete vaccination
Coefficients Concentration indexes contribution to CI contribution rate
User fee -0.050(0.012)*** 0.114(0.002)*** -0.008(0.002)*** -10.246(2.302)***
Time -0.332(0.141)** 0.005(0.001)*** 0.000(0.000) -0.399(0.216)*
Travel cost -0.352(0.060)*** -0.005(0.003) 0.000(0.000) 0.442(0.305)
Age -16.205
Age 30-44 -0.982(0.048)*** 0.015(0.005)*** -0.002(0.001)*** -2.398(0.926)***
Age 45-59 -1.718(0.065)*** 0.061(0.006)*** -0.010(0.001)*** -13.529(1.963)***
Gender male -0.044(0.041) 0.023(0.004)*** 0.000(0.000) -0.278(0.275)
Education 12.899
Medium education 0.171(0.054)*** 0.031(0.004)*** 0.001(0.000)** 1.418(0.515)***
High education 0.622(0.064)*** 0.154(0.008)*** 0.009(0.001)*** 11.481(1.686)***
Occupation 12.796
Migratory workers 0.134(0.052)** 0.064(0.007)*** 0.001(0.000)** 1.235(0.530)**
Student 0.435(0.059)*** 0.331(0.008)*** 0.011(0.002)*** 14.059(2.339)***
Other occupation 1.135(0.079)*** -0.054(0.016)*** -0.002(0.001)*** -2.498(0.898)***
Income 52.025
Income group 2 0.079(0.071) -0.466(0.005)*** -0.003(0.003) -3.759(3.499)
Income group 3 0.173(0.069)** -0.082(0.006)*** -0.001(0.001) -1.597(0.683)**
Income group 4 0.297(0.067)*** 0.366(0.005)*** 0.012(0.003)*** 15.449(3.133)***
Income group 5 0.447(0.071)*** 0.807(0.003)*** 0.032(0.006)*** 41.932(5.679)***
True transm. route ind. 0.136(0.012)*** 0.075(0.003)*** 0.009(0.001)*** 11.401(1.395)***
Perceived protection 17.431
Protection 2 0.607(0.049)*** 0.048(0.005)*** 0.004(0.001)*** 5.690(0.921)***
Protection 3 0.553(0.070)*** 0.144(0.011)*** 0.004(0.001)*** 5.069(0.941)***
Protection 4 0.787(0.095)*** 0.073(0.019)*** 0.001(0.000)** 1.659(0.528)***
Protection 5 0.644(0.070)*** 0.125(0.012)*** 0.004(0.001)*** 5.013(0.892)***
Note: *Two-sided p-value <0.1; **Two-sided p-value <0.05; ***Two-sided p-value <0.01. Standard errors (in parentheses) were obtained by non-parametric bootstrap.
