Our findings indicate that in both 2006 and 2008–2009, there was inequity in deliveries occurring at a health care facility of county- or higher-level hospitals. However, the proportion of home delivery was not equal among low- and high-income families, with higher income families representing the larger portion of births that occur in health care facilities in both years. The obvious inequity between the poor group and the wealthy group reflects the fact that the costs associated with delivery at a healthcare facility, such as non-delivery related hospital fees and travel or time away from the home and work for other family members or caretakers, all of which remain a burden for low-income families.
In addition, the same condition was observed for use of attending ≥5 antenatal visits and partaking in postnatal care, as well as the proportion of ≥3 postnatal visits attended. The poor women remained less likely to attend ≥5 antenatal visits, to receive antenatal care within the recommended 12 weeks, or to carry out the recommended ≥3 postnatal visits, as compared to their non-poor counterparts. Differences in carrying out postnatal visits were also remarkable among the different economic groups. This may reflect an inhibition produced by out-of pocket expenses. In general, postnatal visits are not very common in China [18], and this was also observed in our study. Usually, poorer women live further away from hospitals, which represents an additional challenge in traveling to the facility and results in less postnatal visits. Moreover, the poorer population often has a less robust consciousness of self-care, due to the limited education compared to the wealthier population; therefore, they tend not to take the initiative to seek health care services. Similar results have been found in another study on basic health service utilization in China [19].
An inequity in the use of C-sections appeared in 2008–2009, which indicated that poor women were inclined to take C-section than their non-poor counterparts. This outcome was also supported by other research carried out in rural China by another team. In that study it was shown that from 1993 to 2010, 129,219 of the 49,6054 cesarean deliveries were carried out upon maternal request (CDMR). The prevalence rates of cesarean delivery and CDMR were reported as 37.6% and 10.0%, respectively; moreover, the proportions of CDMR for all cesarean deliveries were found to be significantly increased in both the north and south regions [20]. A subsequent study also indicated that the proportion of CDMR continues to rise [21].
Among all of the indicators examined in our study, the rate of home delivery showed the greatest inequity. In China, the home delivery rate is generally very low (5% in 2006 and 1.5% in 2008 from our study), and the main factors associated with home delivery have been reported to be the mother’s economic conditions, level of education, and location [22]. Pregnant women who are poor generally live farther away from health facilities than their wealthier counterparts, and have lower levels of education. Comparison to other indicators in our study showed that home delivery is the strongest correlated factor to a mother’s economic situation. Moreover, home delivery was mostly concentrated within the poorer population sample and showed the most robust inequity in our analysis.
Another intriguing result of our study was that CIs of gravidity history and outcome of delivery, including number of pregnancies, showed no significant, correlation with proportion of low birth weight and gestational weeks ≥37. This finding agreed with previous reports, in which most of the indicators were associated with mother’s age and inter-pregnancy intervals [23, 24]. The mother’s economic situation has not been reported as an associated factor.
The improvement on utilization of maternal and child health we observed is in accordance with the Chinese government’s goals of increasing access to and utilization of healthcare by all citizens, regardless of economic status. It is also in line with trends observed in previous studies of western rural China, especially in undeveloped areas [25–27]. The Chinese national initiatives of NCMS and MCH were established with the principal aim of lessening inequities in human health care that disfavor the poor. NCMS was introduced in counties of the Shaanxi province in 2004, and was gradually expanded throughout the region. One feature of NCMS was to cover the costs of delivery at health care facilities, and this may have encouraged a large amount of women to eschew home delivery. In addition, the CHIMACA project has also helped to improve the quality of health service and health education in local areas, to a certain degree. In 2008, some local programs were introduced to cover the entire cost of hospital delivery at a health facility. This more comprehensive initiative further promoted healthcare utilization and effectively improved the rates of utilization of hospital delivery among poor women while reducing the inequity in maternal health care utilization in rural China [28].
In recent years, a substantial amount of national projects have been launched to ameliorate maternity services. These public health initiatives have improved the access to and utilization of health services in western rural China. In addition, it has also become evident that regular use of antenatal services can improve delivery outcomes, both for the mother and the child. As such, the Chinese government might find it beneficial to devote more efforts towards promoting maternity services, especially targeting women of lower economic status at fertile ages. The government needs to develop and implement a reasoned strategy to ensure stable and long-term funding of health services and to increase their availability to all socio-economic layers of society.
However, there are some inherent limitations to our study design that should be considered when interpreting the results. First, all the data were collected by a self-report approach, and there may be recall-bias. However, in both surveys, the data was collected within one year of the childbirth under investigation. Pregnancy and childbirth are events that women remember for years, thus the recall bias is assumed to be small. Second, we relied solely on the numerical value of family income to indicate the family’s economic index. This feature may have been insufficient to accurately measure a family’s wealth. For example, a family may have had other sources of wealth, such as property, that would not be reflected by yearly income. Thus, during the portion of the interview (using the structured questionnaire) that was related to income and expenditures, the interviewers asked: “compared with other families in your village, how is your economic status rated on a 5-degree scale?” We then verified the economic levels by the reported income and expenditures, respectively, and found that the relative levels were similar. Third, the time for giving a survey must be considered in order to lessen the rate of non-response. In 2007, the survey was conducted after the national Spring Festival, and the higher non-responder rate (vs. in 2008 – 2009) was reported by the local doctors to be due to the general practice of travel during this time. Then in 2008–2009, the part of the survey that was conducted before the Spring Festival had completed. In general, the incomes of families with an employed mother is better than families with a mother who is unemployed, and an inaccurate response rate for this parameter may impact the study’s results (i.e. underestimating the inequity). Finally, the maternal mortality rate is generally considered a good index of a women’s health. Unfortunately, the sample sizes of our surveys were too small to analyze that particular index and larger study populations are needed. In the current study, the CI of combined data from both counties was similar to the CIs that were calculated for each county in isolation. For example, the inequity in deliveries occurring at a healthcare facility showed the same statistical significant in each year and in each county. Similar results were also obtained for the analyses of antenatal care within 12 weeks and proportion of home deliveries. Certainly, future equity analyses need to be nuanced enough to identify the remaining disparities between different types of rural and urban areas, in addition to the broad rural–urban inequities.