Portal de Periódicos da CAPES

Utilisation of maternal health care in western rural China under a new rural health insurance system (New Co-operative Medical System)

2010; Wiley; Volume: 15; Issue: 10 Linguagem: Inglês

10.1111/j.1365-3156.2010.02602.x

1365-3156

Qian Long, Tuohong Zhang, Ling Xu, Shenglan Tang, Elina Hemminki,

Global Health Care Issues

Objective To investigate factors influencing maternal health care utilisation in western rural China and its relation to income before (2002) and after (2007) introducing a new rural health insurance system (NCMS). Methods Data from cross-sectional household-based health surveys carried out in ten western rural provinces of China in 2003 and 2008 were used in the study. The study population comprised women giving birth in 2002 or 2007, with 917 and 809 births, respectively. Correlations between outcomes and explanatory variables were studied by logistic regression models and a log-linear model. Results Between 2002 and 2007, having no any pre-natal visit decreased from 25% to 12% (difference 13%, 95% CI 10–17%); facility-based delivery increased from 45% to 80% (difference 35%, 95% CI 29–37%); and differences in using pre-natal and delivery care between the income groups narrowed. In a logistic regression analysis, women with lower education, from minority groups, or high parity were less likely to use pre-natal and delivery care in 2007. The expenditure for facility-based delivery increased over the period, but the out-of-pocket expenditure for delivery as a percentage of the annual household income decreased. In 2007, it was 14% in the low-income group. NCMS participation was found positively correlated with lower out-of-pocket expenditure for facility-based delivery (coefficient −1.14 P < 0.05) in 2007. Conclusions Facility-based delivery greatly increased between 2002 and 2007, coinciding with the introduction of the NCMS. The rural poor were still facing substantial payment for facility-based delivery, although NCMS participation reduced the out-of-pocket expenditure on average. Utilisation des soins de santé maternelle en milieu rurale dans l'ouest de la Chine sous un nouveau système d'assurance maladie en milieu rural (New Co-operative Medical System: NCMS) Objectif: Investiguer les facteurs influençant l'utilisation des soins de santé maternelle en milieu rural dans l'ouest de la Chine et la relation avec le revenu avant (2002) et après (2007) l'introduction d'un nouveau système d'assurance santé en milieu rural (NCMS). Méthodes: Les données d'études transversales sur la santé des ménages effectuées dans dix provinces en milieu rural dans l'ouest de la Chine en 2003 et 2008 ont été utilisées dans l'étude. La population étudiée comprenait les femmes ayant accouché en 2002 ou 2007, avec 917 et 809 naissances, respectivement. Les corrélations entre les résultats et les variables explicatives ont étéétudiées par des modèles de régression logistique et un modèle log linéaire. Résultats: Entre 2002 et 2007, le fait de n'avoir aucune visite prénatale à diminué de 25%à 12% (différence de 13%, IC95%: 10%-17%); L'accouchement dans un établissement est passé de 45%à 80% (différence de 35%, IC95%: 29%-37%) et les différences dans l'utilisation des soins prénataux et d'accouchement entre les groupes de revenus ont été réduites. Dans une analyse de régression logistique, les femmes moins instruites, issues de groupes minoritaires ou de paritéélevée étaient moins susceptibles d'utiliser des soins prénataux et d'accouchement en 2007. Les dépenses pour l'accouchement dans un établissement ont augmenté au cours du temps, mais les dépenses directes de la poche en tant que pourcentage du revenu annuel des ménages ont diminué. En 2007, elles étaient de 14% dans le groupe à faibles revenus. Une corrélation positive a été trouvée entre la participation au NCMS et des dépenses directes de la poche plus faibles pour l'accouchement dans un l'établissement (Coefficient -1,14; P <0,05) en 2007. Conclusions: L'accouchement dans un établissement a considérablement augmenté entre 2002 et 2007, coïncidant avec l'introduction du NCMS. Les pauvres en zones rurales sont encore confrontés à des paiements substantiels pour l'accouchement dans des établissements, bien que la participation au NCMS ait réduit en moyenne les dépenses directes de la poche. Utilización de servicios de salud materna en el área rural del oeste de China bajo un nuevo sistema de seguro sanitario rural (Nuevo Sistema Médico Cooperativo) Objetivo: Investigar los factores que influyen en el uso de los servicios de salud materna en China rural occidental y su relación con los ingresos, antes (2002) y después (2007) de introducir un nuevo sistema de seguro sanitario rural (NSMC). Métodos: Se utilizaron datos de estudios croseccionales domiciliarios realizados en diez provincias rurales occidentales de China en el 2003 y el 2008. La población de estudio comprendía a mujeres que habían dado a luz en el 2002 o en el 2007, con 917 y 809 nacimientos, respectivamente. Se estudiaron las correlaciones entre resultados y variables explicativas mediante modelos de regresión logística y logarítmico lineal. Resultados: Entre el 2002 y el 2007, el no haber tenido ninguna visita prenatal disminuyó del 25% al 12% (diferencia del 13%, 95%IC 10%-17%); los partos atendidos en centros hospitalarios aumentaron del 45% al 80% (diferencia del 35%, 95%IC 29%-37%); y disminuyeron las diferencias en el uso de servicios prenatales y de parto entre los grupos con diferentes ingresos. En un análisis de regresión logística, las mujeres con una menor educación, de grupos minoritarios, o una alta paridad tenían una menor probabilidad de utilizar los servicios prenatales y de parto en el 2007. El gasto por parto hospitalario aumentó en el período, pero el gasto de bolsillo para el parto como un porcentaje del ingreso anual del hogar disminuyó. En el 2007, era del 14% en el grupo de menores ingresos. La participación en NSMC tenía una correlación positiva con un menor gasto de bolsillo para el parto hospitalario (Coeficiente -1.14 P<0.05) en el 2007. Conclusiones: El parto hospitalario aumentó considerablemente entre el 2002 y 2007, coincidiendo con la introducción del NSMC. Los campesinos pobres seguían haciéndose cargo de una parte sustancial del parto hospitalario, aunque la participación en NSMC reducía el promedio del gasto de bolsillo. China is on track to achieve the Millennium Development Goal of a reduction in both maternal and child mortality. The maternal mortality ratio (MMR) has declined from 89 per 100 000 live births in 1990 to 47 per 100 000 live births in 2005. However, the achievements show significant regional disparity. In poor western rural areas, the MMR was four times that of urban areas and twice that of average rural areas (Jing & Kaufman 2008). Studies in many developing countries show a positive association between reductions in maternal mortality and utilisation of maternal health care (Adam et al. 2005; Ekman et al. 2008). Utilisation is defined as attendance at pre-natal and post-natal care and delivery at a health facility. In China, inadequate use of maternal health care, especially delivery care, has usually been considered a major reason for persistently high rural maternal mortality (Anson 2004; Short & Zhang 2004). In 2003, in the poorest rural areas, only 16% of women gave birth at health facilities compared to 93% of women in urban areas (Ministry of Health 2004). Financial difficulty and a lack of perceived benefits from seeking care are seen as the main causes for the underuse of maternal health care in China (Bogg et al. 2002; Kaufman & Jing 2002; Wu et al. 2008). In 2003, the most important reason given for home delivery was that a facility-based delivery was too expensive (Ministry of Health 2004). The willingness to meet this cost is further undermined by the perceived lack of benefit from seeking care, with perceptions influenced by age, education, history of pregnancy-related activities, cultural beliefs and family support (Li 2004; Short & Zhang 2004; Wu et al. 2008). In the 1970s, China had a nationwide rural health insurance system characterised by its collective financing, called the Co-operative Medical System (CMS). It covered 90% of the rural population, ensuring universal access to essential medical care. However, this system collapsed in the 1980s with the economic transition (Feng et al. 1995), meaning health care became less affordable in rural China (Liu et al. 2008). In the early 1990s, the Chinese government tried to re-establish CMS, but the scheme could not be sustained in poor areas because of insufficient financial support (Carrin et al. 1999). Since 2003, a new health insurance system for rural people, the New Co-operative Medical System (NCMS), has been implemented that aims to provide financial risk protection for the rural population in regard to catastrophic disease. It is financed by individual contributions and by central and local government. The NCMS operates with a voluntary enrolment; it is administered at county level, although rules and policy guidelines are set by central government (Wang 2007; You & Kobayashi 2008). In the initial stage of the NCMS, rural residents had little willingness to participate in the schemes. Participation was associated with relatively low household income, higher education, number of sick family members, potential benefits from premium and co-payment costs (Zhang et al. 2006; Wang et al. 2008). The participation rate has consistently increased, however, reaching 86% in 2007 (You & Kobayashi 2008). The NCMS has a maternal health care benefit package. Its design and implementation vary across counties. The package usually provides reimbursement either as a fixed proportion of expenditures or a fixed amount for facility-based delivery, both for normal delivery and caesarean section. Women pay for pre-natal care, including physical check up by a doctor, laboratory tests, drugs and treatments. To our knowledge, only a few counties have included part of pre-natal care into the NCMS maternal health care benefit package. This study focuses on the demand side of maternal health care, aiming to investigate factors influencing the use of maternal health care in western rural China and its relation to income before (2002) and after (2007) introducing the NCMS. We analysed regionally representative data from ten western provinces gathered in a health survey carried out in 2003 and 2008. The survey was developed by the Centre for Health Statistics and Information (CHSI), which operates under the Chinese Ministry of Health. In both surveys, multistage sampling was used in each of the ten provinces to randomly select thirty rural townships, two villages in each township and thirty households from each village, using residence register data (HuKou), yielding 18 000 households. The survey was carried out by trained township health workers using a structured questionnaire. Where households refused to be interviewed or no-one was home at the time of survey (10%), alternative households were selected. The questionnaire included questions on the general demographic and socioeconomic background of the sampled households and family members, the perceived need and demand for health care, and the utilisation and expenditures of health services. One section on pregnancy-related activities was exclusively for women of reproductive age (15–49). Only data for rural women who gave birth in 2002 or 2007 are reported here. Four indicators of the use of maternal health care as recommended in the 'Maternal and Infant Health Care Law of the People's Republic of China' were examined: (i) use of pre-natal care, that is, having at least one pre-natal visit; (ii) initiation of pre-natal care within the first trimester of pregnancy (≤12 gestation weeks); (iii) five or more pre-natal visits; and (iv) delivery at a health facility, defined as giving birth at a township or higher level health facility. In addition, out-of-pocket expenditure for facility-based delivery was used as an indicator to evaluate the financial consequence for rural households. The explanatory variables were determinants of maternal health care utilisation identified in previous studies: age, women's education attainment (illiterate, primary school, secondary or higher), ethnicity (Han or minorities), parity (number of live births), history of abortions (miscarriage or induced abortion) or stillbirths, distance to the nearest health facility (e.g. village clinic, township or higher level health facility), income level and NCMS participation. Income level was estimated using annual per capita income. The reported gross household income was divided by the number of individuals in the households and grouped into three income categories (low, medium and high), each containing a third of the households interviewed. Centre for Health Statistics and Information (CHSI) allowed data analysis (made by QL) only in CHSI premises to protect data confidentiality. Cross-tabulation was used to compare utilisation of pre-natal and delivery care between 2002 and 2007. The 95% confidence intervals for the differences in maternal health care use between the time periods were tested. The range method and chi-square trend test were used to examine the variation in trends for maternal health care use by increasing income. Logistic regression models were used to analyse associations between the use of pre-natal and delivery care and explanatory variables separately in 2002 and 2007. Furthermore, we did the analysis combining the datasets of 2002 and 2007 adding time (year) as a dummy variable. Based on women's self-report data, we calculated the mean delivery expenditure at the health facility, the mean out-of-pocket expenditure, and the mean annual household income among the different income groups in 2002 and 2007. In addition, in 2007, the out-of-pocket expenditure of facility-based delivery was transformed into natural logarithms of the observation value to address the positive skew of the expenditure data. Finally, a linear regression model was used to study the correlation between the logarithmic value of the out-of-pocket expenditure of delivery care and the NCMS, adjusting for covariates. If the estimated coefficient of the variables was negative, it indicates a correlation with reduced out-of-pocket expenditure; a positive coefficient indicates a correlation with increased expenditure. The number of respondents included in the study who gave birth in the calendar year prior to the survey was 917 in 2002 and 809 in 2007. Table 1 shows that the two groups were relatively similar in regard to age, ethnicity and parity. The most significant change between 2002 and 2007 was the increase in the proportion of women participating in the NCMS: from only a small percentage to almost everyone. Illiteracy among women decreased from 2002 to 2007, and over half of the women in 2007 had received at least a secondary education or more. The numbers of those reporting a history of abortion or stillbirth were almost twice as high in 2007 than in 2002. Utilisation of pre-natal and delivery care increased from 2002 to 2007 (Table 2). The proportion of women not using any pre-natal care decreased, and the proportion making many pre-natal visits increased, with more women starting care early. Home deliveries declined from over half in 2002 to a fifth in 2007, with 46% of women choosing to deliver at either county or higher level health facilities. Most women had a normal vaginal delivery without the use of instruments, but there was a notable increase in the use of caesarean section (from 6% to 17%). In both study years, women with higher income used more pre-natal care and gave birth more often in health facility than women with lower income (Table 3). The ratio (RR) and difference (RD) between the high- and low-income groups for use of pre-natal and delivery care was smaller in 2007 than in 2002. However, with the exception of the start time for pre-natal care, the differences between the income groups were statistically significant even in 2007. After adjusting for other factors, in 2007 age was positively related to any pre-natal visit and facility-based delivery, but the positive association with five or more pre-natal visits was not statistically significant (Table 4). Education was positively associated with the use of pre-natal and delivery care. Minority women were less likely to have early or adequate pre-natal visits, but there was no difference in delivery care utilisation. Having existing children was negatively associated with the pre-natal and delivery care utilisation. Women who reported a history of abortion or stillbirth were less likely than those without to start pre-natal visits early. High income was related to the use of any pre-natal care and many visits, but not to early start. Women with medium and high income more often had facility-based delivery than low-income women, but the difference was not statistically significant. NCMS participation was positively related to using any pre-natal care. It was also related to making early or adequate pre-natal visits or having facility-based delivery, but the odds ratios were not statistically significant. Finally, the number of pre-natal visits was positively associated with giving birth at a health facility. When we combined the datasets of 2002 and 2007 adding time (year) as a dummy variable, we found very similar results to those for 2007 alone. However, shorter distance to health facility was related to five or more pre-natal visits (data not shown). In 2002, out-of-pocket expenditure for facility-based delivery was not asked. Given that most rural people did not have any health insurance before 2003, we assumed that total delivery expenditure was equal to the out-of-pocket expenditure. From 2002 to 2007, total expenditure of facility-based delivery increased by 48% and 53% among the low- and medium-income groups, 76% among the high-income group (Table 5). There was almost no increase in out-of-pocket expenditure among the low-income group, a modest increase among the medium-income group, and a bigger increase among the high-income group. In 2007, in all income groups, the annual household income almost doubled compared to 2002. The average out-of-pocket expenditure as a percentage of the annual household income decreased from 2002 to 2007. In both years, it was higher among the low-income group than among the medium- and high-income groups, even though the difference decreased. The logarithmic value of the out-of-pocket expenditure of facility-based delivery in 2007 was analysed by linear regression (Table 6). After adjusting for age, education, ethnicity, income, delivery mode and level of health facility, NCMS participation correlated with reduced out-of-pocket expenditure (coefficient −1.14 P < 0.05). In addition, older age and higher education was correlated with higher out-of-pocket expenditure. The out-of-pocket expenditure increased as the income increased. The out-of-pocket expenditure for caesarean section was higher than for vaginal delivery. The expenditure in county level was higher than in township-level health facility. In China's less developed western rural areas, the use of facility-based delivery greatly improved, while the use of pre-natal care showed a modest increase. There was income-related inequality in using pre-natal and delivery care, but differences between the income groups narrowed in the period under study. The out-of-pocket expenditure for facility-based delivery as a percentage of the annual household income remained high in the low-income group, although NCMS participation reduced the out-of-pocket expenditure for facility-based delivery on average. The data for this study are regionally representative. The sample size is large, giving the analyses good statistical power. However, there are some limitations. First, data on the frequency and timing of pre-natal visits and the expenditure of delivery care are subject to recall bias, although a 1-year recall is unlikely to cause serious bias. Secondly, the NCMS schemes varied across counties in terms of the level of reimbursement and services covered, so without data on these variations, the results should be viewed as preliminary. Maternal health care has been advocated as a main strategy to prevent women from death and morbidity from pregnancy-related causes (Filippi et al. 2006). Our findings show a notable increase in facility-based delivery and a modest increase in utilising pre-natal care in western rural areas. Consistent with other studies in rural China and other developing countries, we found that lower education, minority status, and higher parity were negatively related to pre-natal and delivery care utilisation (Anson 2004; Short & Zhang 2004; Simkhada et al. 2007; Sepehri et al. 2008). In addition, economic factors had a significant association with maternal health care utilisation. The impressive increase in facility-based delivery may be partly because of the reduced financial burden on households, while contrastingly, the willingness to pay for pre-natal care is still low. Financial difficulty is usually considered one of the main barriers to maternal health care utilisation in many low- and middle-income countries (Smith & Sulzbach 2006; Witter et al. 2007). Our study found that low-income women were less likely to utilise pre-natal care and facility-based delivery compared to medium- and high-income women. From the mid-1980s, Chinese health sector reforms shrunk the government allocation for health care. Thereafter, the health care system became heavily dependent on fee-for-service financing. This combined with profit-related bonus payments for health providers resulted in a rapid increase in medical expenditures (Gao et al. 2002). High expenditure may have created a barrier to poor women using maternal health care, particularly in rural areas where demand is low. Bogg and colleagues have reported similar findings in other areas of rural China (Bogg et al. 2002). Nevertheless, in our study, the differences by income in utilisation, especially of facility-based delivery, lessened between 2002 and 2007. A reason to this reduced difference can be the smaller increase in total and out-of-pocket delivery expenditures in the low-income group than the high-income group. The NCMS covered a modest part of the expenditures for facility-based deliveries. To our knowledge, no previous studies have reported the impact of the NCMS on the expenditure of maternal health care. The NCMS may be a useful way to share the financial risk related to delivery and to protect poor women from potential catastrophic payments. A study that estimated the impact of the NCMS on health care in general using panel data collected in fifteen rural counties in 2003 and 2005 showed that on average the NCMS increased the out-of-pocket health expenditures among all income groups, but reduced the incidence of catastrophic spending among the poorest group (Wagstaff et al. 2007). During the study period, the Chinese government funded a large maternal project: 'Reducing maternal mortality and eliminating tetanus infection of newborn babies'. It had a wide coverage in western rural China and implemented comprehensive intervention measures, such as giving health education to rural women and a financial subsidy to cover part of facility-based delivery expenditures (Guo et al. 2008). Hence, it is difficult to separate the contributions of the NCMS from those of that other project. Nevertheless, this project was only for a certain period of time, and it, like other short-term project, cannot sustainably solve the financing problems of maternal health care. In this context, expanding NCMS can be a good opportunity to continue to improve maternal health in rural China. In 2007, the out-of-pocket expenditure of facility-based delivery was about 14% of the annual household income among the low-income group, which remains a large payment at one time referring to the definition of catastrophic spending that threshold varied from 10% to 40% of annual household income (Ranson 2002; Xu et al. 2003). No studies have reported how poor households in rural China manage such high delivery expenditure. Borrowing money and selling assets to cover delivery-related expenditures have been reported in some African and South Asian countries (Nahar & Costello 1998; Storeng et al. 2008). In addition, in facility-based delivery, there are much indirect expenditures, such as transportation and lodging for the accompanying family members. In Tanzania and Nepal, these indirect expenditures have been estimated to be 50% of the total normal delivery expenditure (Borghi et al. 2006). In our study, income was positively related to having pre-natal care and to the number of visits. Having pre-natal care and the number of visits were associated with giving birth at health facility. Pre-natal contacts are a good way to give health education and psychosocial services to women, and to raise women's awareness of safe delivery care (Simkhada et al. 2007). However, ability and willingness to pay can limit the use of pre-natal care. Studies in the Philippines, Vietnam and some African countries showed that health insurance coverage increased the probability of having the recommended number of pre-natal visits (Smith & Sulzbach 2006; Sepehri et al. 2008; Kozhimannil et al. 2009). The new round of health-system innovations set up by the Chinese government in 2009 set a goal of gradually achieving an equalisation in basic public health care, with maternal health care as one of the targets (National Development and Reform Commission, 2009). We suggest increasing the NCMS reimbursement levels for facility-based delivery to further reduce the financial burden for the rural poor households. In addition, pre-natal care should be included into the benefit package of the NCMS to encourage women to use it. However, the norms for the adequate number of visits and care contents should be revisited to correspond to international recommendations (Villar et al. 2001). The CHSI of Ministry of Health funded data collection and management. Data analysis and interpretation was supported by a grant from the China Medical Board (CMB). We thank the CMB for funding a training workshop in data analysis. This study is part of the output of the CHIMACA project (015396) funded by the European Commission INCO Programme and co-ordinated by the National Institute for Health and Welfare, Helsinki. We thank Mark Phillips for polishing the English language of the text. We appreciate the reviewers giving valuable comments. QL appreciates the China Scholarship Council for supporting her study abroad.