Factors associated with seeking post-abortion care among women in Guangzhou, China

Background In China, the vast majority of induced abortions are performed in public hospitals. However, post-abortion care (PAC) services are provided through the national network of family planning clinics, which are independent of the health care system. The integration of PAC services into abortion clinics in public hospitals is a new concept. This study aimed to assess PAC utilization among abortion patients, and identify the possible factors associated with PAC uptake in Guangzhou, China. Methods A cross-sectional survey was conducted among 431 women aged 15–43 years in Tianhe district of Guangzhou, China from June to September 2018. We estimated multivariate logistic regression model to examine the factors associated with utilization of PAC services. Results Less than half (42%) of the participants used PAC services. Married women were 2.7 times significantly more likely to use PAC services than their unmarried counterparts. Immigrants were 52% significantly less likely to use PAC services than non-immigrants. Women who perceived that their fertility could return later and those who did not know were 45 and 61% significantly less likely to use PAC services compared to those who knew that their fertility could return soon after an abortion. Women with limited decision-making autonomy regarding contraceptive use were 54% significantly less likely to use PAC services than those who made such decisions themselves. Conclusions The findings suggest the need for policies and programs to not only strengthen the provision of PAC services but also promote uptake among disadvantaged sub-groups of women in the study setting.


Background
Unplanned pregnancies and induced abortions (IAs) remain an important public health issue, especially among women from developing countries. IAs result from unplanned pregnancies which occur due to contraceptive failure or non-use of contraception during sexual intercourse. Estimates suggest that the average IA rate is approximately 58 per 1000 women worldwide [1]. In China, IA is a legal procedure in the first trimester of pregnancy. Recent estimates show that between 6 and 9 million IAs occur in the country every year [2]. Nearly half of lowincome urban women in the country had experienced IAs with 31% being repeated abortions [2]. The causes of IAs are complicated. One of the major reasons is lack of quality post-abortion care (PAC) counseling [2].
The World Health Organization recommends counseling on and provision of contraceptive methods to all those women who wish to prevent unintended pregnancies and subsequent IA [3]. The term PAC was first used in 1991 and referred to an approach to break the cycle of unwanted pregnancy and improve women's sexual as well as reproductive health [4]. The integration of family planning (FP) counseling and method provision into abortion services is an essential part of PAC. Multiple studies have demonstrated the effectiveness of PAC in reducing repeat IAs when offered to abortion patients prior to discharge from health facilities [5,6]. However, knowledge of the extent of PAC utilization is limited. Available data regarding PAC utilization rate is varied across countries and regions. How the specific context affects the implementation of PAC and what factors contribute to the accessibility of PAC services continuously attract researcher's interest.
Recent studies suggest that type of health facility, the decision maker on timing to have a child, knowledge of fertility return after abortion and husband's attitudes towards contraceptives may be significantly associated with PAC utilization [7][8][9]. The existing literature mainly focused on abortion patients in African countries which have significant disparities in culture and social environment from China. Report on the status of PAC utilization among Chinese abortion patients is scant and inconclusive. Available research uses qualitative methods including semistructured interviews and focus group discussion were employed to explore the feasibility and acceptability of high quality PAC services from the service providers' perception [10]. This is a limitation given the rapidly changing demographics of abortion patients in the south area of China over the past decade and inability to generalize findings from such studies to the region.
In China, the vast majority of IAs are performed in public hospitals. However, PAC services are provided through the national network of family planning clinics, which are independent of the health care system. The integration of PAC services into abortion clinics in public hospitals is a new concept. In 2010, the National Population and Family Planning Commission launched a pilot program to include the provision of PAC into 486 hospitals [11]. Guangzhou, where this study was conducted, was one of the pilot cities where free PAC services were provided in selected public hospitals. There is, however, limited understanding of the extent to which PAC services are utilized among abortion patients in such settings. Furthermore, factors that may be associated with PAC utilization are not well-understood. Hence, this study aimed to assess post-abortion care (PAC) utilization among abortion patients, and identify the possible factors affecting PAC uptake in Guangzhou, China.

Study setting
Guangzhou is the capital city of Guangdong province, an economically developed province in south China. Tianhe district, where this study was conducted, was the pilot site where free PAC services were offered in public hospitals.

Study design
A cross-sectional survey was conducted in Tianhe district of Guangzhou during the period of June to September 2018. All those women who had experienced at least once IA in the selected public hospitals were included in the study. However, we excluded women who were unable to speak or listen and those with psychiatric disorder.

Sample size determination
The sample size was determined using the single population proportion assuming 57.4% of women utilizing PAC services and the desire to obtain reasonable estimates at 95% confidence level and 5% margin of error [12].
The total sample size was 431 women, taking into account 15% non-response.

Sampling procedure
Three public hospitals in Tianhe district that provided PAC services on a pilot basis were included in the study. The average number of PAC users was estimated according to the recent quarterly report of patients flow in each hospital. We used a systematic random sampling approach to select participants. The number of participants from each hospital was determined based on population proportion to size. Thus the number of participants was 156, 127 and 148 from the first, second and third hospital respectively.

Data collection
Data were collected using an anonymous structured questionnaire. The development of the questionnaire was informed by existing literature on PAC utilization. The questionnaire was reviewed by experts in reproductive medicine and clinical epidemiology from China. After the pilot test, the questionnaire was revised accordingly, covering the following information: 1) sociodemographic characteristics, such as age, marital status, education, employment status, income and migrant characteristics; 2) reproductive history such as parity, previous IAs, number of living children; 3) contraceptive and reproductive health knowledge such as knowing how soon fertility returns and the time of getting pregnant again, uptake and method of contraception; 4) other related variables such as husband's/partner's attitude towards contraceptive use and the person responsible for making contraceptive decisions.
All participants were interviewed face-to-face after obtaining written permission. They were assured of the confidentiality of their identity and responses. The data collection phase was completed with the help of seven post-graduate female nurses. They were trained for 2 days by the principal investigator covering interview techniques, quality control, completeness of information and research ethics. All completed questionnaires were checked for completeness and consistency.

Statistical analysis
The primary data was entered into Epidata 3.1 before being exported to SPSS 20.0.
Women with missing information on key attributes were excluded from the analysis.
Categorical variables were presented as counts and proportions. Cross-tabulations with Chi-square test were used to assess significance of differences in sociodemographic characteristics, reproductive history, reproductive health knowledge and other related factors between abortion patients who used PAC services and those who did not. The independent variables that were significantly associated (p<0.05) with PAC services utilization in cross-tabulations were considered as possible contributing factors and entered into a multivariate logistic regression model. The model examined factors that were significantly associated with use of PAC services after controlling for possible confounders. The outcome variable in the regression analysis, use of PAC services, was dichotomous (whether the participants had used the services or not). The model controlled for marital status at the time of interview (married and unmarried), place of household registration (whether the participant's household was registered in Guangzhou or not), knowledge of return to fertility (within 10-14 days, after 3-4 weeks, and don't know), and the person responsible for making contraceptive decisions (respondent herself, respondent's husband/partner, or both). Odds Ratios (ORs) with 95% confidence intervals (95% CIs) were calculated to measure the strength of association. Estimates with p-values less than 0.05 were considered statistically significant.

Ethics
The study protocol was approved by the Research Ethics Board of Southern Medical University, China. All participants provided written informed consent before being interviewed.

Socio-demographic characteristics of study participants
Of the 431 eligible women, 425 (98.6%) consented to participate in the survey. Among those who consented to participate in the survey, 413 (97%) completed the interviews. Of the 413 abortion patients who completed the interviews, 174 (42%) utilized PAC services.
Slightly more than a quarter (28%) of the participants were aged between 25 and 29 years and a similar proportion (27%) were between the ages of 20 and 24 years. More than half (55%) of the participants were unmarried. More than three-quarters (79%) of the participants had senior high school or college level education. Regarding employment status, 231 (56%) were employed and 125 (30%) were students. The highest proportion (42%) of participants earned between 2000 and 4999 Chinese Yuan per month, followed by those who earned between 5000 and 7999 Chinese Yuan (25%). Distribution by place of household registration shows that 62% of the participants were immigrants (Table 1).

Variations in the use of PAC services by background characteristics
The proportion of married women who used PAC services was significantly higher than that of unmarried women (52 and 31%, respectively; p < 0.01; Table 2). The proportion of immigrants that used the services was almost four times lower than that of non-immigrants (19 and 80%, respectively; p < 0.01; Table 2). In addition, the proportion that used the services was significantly higher among women who knew that their fertility would CNY Chinese Yuan return soon following an abortion (77%) than among those who knew that their fertility would return later (50%) or those who did not know when their fertility would return (10%; p < 0.01; Table 4). The proportion was also significantly greater among women who made contraceptive decisions themselves (56%) than among those who made the decisions jointly with their partners (39%) or those whose partners solely made such decisions (30%; p < 0.01; Table 4). There were, however, no statistically significant variations in the proportions using the services by the other factors considered (Tables 2, 3 and 4). Similarly, participants who knew that their fertility could return 3~4 weeks post-abortion were about 45% significantly less likely to utilize PAC services compared to women who knew that their fertility could return within 10~14 days (OR = 0.545, 95%CI: 0.3080 .802). Those who were uncertain about when their fertility could return were also 61% significantly less likely to receive PAC services compared to those who knew that their fertility could return within 10~14 days (OR = 0.391, 95%CI: 0.294~0.617). The likelihood of using PAC services also varied by the person responsible for making contraceptive decisions. Women whose husbands were the main decision-makers regarding contraceptive use were 54% significantly less likely to use PAC services compared to those who made such decisions themselves (OR = 0.460, 95% CI: 0.272~0.809).

Discussion
The study aimed to determine the proportion of women who utilized PAC services among abortion patients and the factors associated with use of such services in public health facilities piloting the provision of services in Guangzhou, China. The findings show that less than half (42%) of women seeking abortion services in the pilot facilities used PAC services. Uptake of PAC services in the study setting was comparable to the level observed in east region hospitals (43%) but lower than the levels in middle and west region hospitals in the country (72 and 58%, respectively) [13]. This disparity could be due to the variations in the provision of PAC in public hospitals. In China, FP counseling, an integral component of PAC, is provided through the national network of FP clinics, which are almost independent of the health care system. However, the vast majority of IAs are performed in public hospitals. The integration of PAC services within public hospitals is a new concept. The pace with which hospitals in different settings take up the provision of the services may vary depending on available financial, human and infrastructural resources. Policy support in the form of regulations and guidelines, human resources and financial resources are the main considerations regarding the implementation of PAC services. Low uptake of PAC services in the study could also be due to the fact that most (60%) of the participants were internal immigrants. In particular, allocation of healthcare resources is based on household registration, and immigrants do not have the same rights and benefits to services as local registered residents.
The findings of this study show that married women were significantly more likely to use PAC services compared to their unmarried counterparts, which is consistent with the findings of another study conducted in China [8]. However, another study in Ethiopia found that married women were less likely to use PAC services compared to unmarried women [7]. Variations in uptake of PAC services by marital status may depend on the extent to which partners have influence over women's access to services. Married women may be less likely to seek services in settings where men wield great control over decision-making and resources for seeking care. In contrast, they are likely to seek care in settings where they have control over such resources. Greater use of PAC services among married compared to unmarried women suggests the need for policies and programs targeting unmarried women who often face unmet family planning needs. This could include information, education and communications interventions on sexuality and prevention of unintended pregnancy.
We also found a significant association between household registration place and PAC utilization. Immigrant women were significantly less likely to utilize PAC services compared to their native counterparts. Because medical insurance system in China is based on household registration, immigrants have much less reimbursement in medical expenses such as PAC counseling than the natives in their host cities. In addition, in China, this group tends to be relatively uneducated, in low-income groups, insecure jobs and are excluded from municipal welfare structures in their host cities [14]. The finding suggests that local governments need to promote social integration of immigrants in order to improve access to health services across all segments of the population.
Another key finding of the paper is that women who perceived that their fertility would return later and those who did not know when their fertility would return were significantly less likely to use PAC services compared to those who knew that their fertility would return soon. This is consistent with findings from a study conducted in Ethiopia [7]. Women who knew that their fertility could return soon are likely to take steps to avoid unintended pregnancy, including seeking FP services, compared to those who did not think or know that they could get pregnant soon after having an abortion. The finding suggests the need for improved counseling of clients on the risks of pregnancy following an abortion.
The findings of the study further show that women whose husbands were the main decision makers regarding contraceptive use were significantly less likely to use PAC services compared to those who made such decisions themselves. A study in Nigeria concluded that reluctance to use PAC services was due to husband disapproval [15]. Opposition by the husband could also be due to limited knowledge of women's reproductive health needs or gendered power differences in control of household resources and decision-making process that favor men. In most hospitals of China, male partners are not allowed to access the department of gynecology and obstetrics. Although they are eventually allowed into FP counseling room in some cases, they are just given some materials regarding reproductive health. Future research could explore whether targeting and including male partners in the provision of PAC could have an impact on increasing PAC utilization among abortion patients. This study has some limitations. First, the crosssectional nature of the study did not allow for establishing causal relationships between PAC use and the factors associated with it. Second, the findings may be affected by response bias. Given the sensitivity of the topic, respondents might feel shy to reveal some personal details or may provide responses they feel are socially desirable. Third, the findings might not be representative of abortion patients in China, since participants were recruited from Guangzhou, China.

Conclusions
Uptake of PAC services still remains low in the study setting where the provision of the services in public hospitals was being piloted. In addition, uptake of the services remains low among disadvantaged sub-groups, including unmarried women, immigrants, those with limited knowledge about bodily functions, and women with limited decision-making autonomy. The findings suggest the need for policies and programs to not only strengthen the provision of the services but also promote uptake among disadvantaged sub-groups of women in the study setting.