Changes in Determinants and Equity of Family Planning Utilization after the Implementation of a National Health Insurance Policy in Indonesia: A Secondary Analysis of The 2012-2016 National Socio-Economic Survey of Indonesia

The Indonesian government has been implementing the National Health Insurance Policy (Jaminan Kesehatan Nasional-JKN) since 2014. The utilization of family planning service is one of the programs to increase maternal and child health status that is included in the benefit package in JKN. This study aimed to describe determinants and to evaluate JKN based on equity indicators, especially in family planning services. Data were obtained from the 2012-2016 National Socio-Economic Survey (SUSENAS) of Indonesia. Contraceptive Prevalence Rate (CPR) and Long-acting contraceptives (LACs) use were used as indicators to evaluate family planning utilization. Chi-square and logistic regression tests were used to analyze the data. Respondents were married women between 15 and 49 years of age.

viz. every country reaches MMR less than 70 per 100,000 live births and IMR less than 12 per 1000 live births in 2030 [3], Indonesia has been implementing many programs that are related with maternal and child health.
The family planning program is one of the ways to increase maternal and child health. Ensuring access to contraceptive use contributes to the success of the family planning program. Contraceptive use through reduction of fertility reduces maternal and neonatal mortality [4,5]. Contraceptive use, especially Long-acting contraceptives (LACs) methods, prevents unintended pregnancy while reducing abortions, as shown by prior research [6].  [1,[7][8][9][10]. The target of CPR in the 2015-2019 Indonesian national development is 66%, while the target of LACs use is 23.5% [11]. One of the obstacles to accessing family planning programs is finances, especially for the poor. To reduce the financial barrier, the Indonesian government has been implementing a National Health Insurance Policy (Jaminan Kesehatan Nasional-JKN) since 2014 [12]. The family planning program is included in the benefit package in JKN [13]. JKN is a social health insurance and compulsory insurance that was implemented gradually to achieve universal coverage in 2019. In early 2014, 49% of the population had insurance coverage. In 2016, 66.46% of the population had insurance coverage. At the end of 2019, all Indonesian people are expected to be protected by health insurance [12][13][14][15] One of the indicators used to assess the success of social health insurance is equity.
It is consistent with one of the goals in SDGs, specifically to reduce the health gaps inside a country and between countries. Equity is defined as no number and/or difference of frequency of using health services based on socio-economic status [16]. In meeting this challenge, Indonesia with 34 provinces confronts a wide diversity and divergence in maternal health status and maternal health service utilization.
Evaluation is needed to measure the influence of the JKN policy on utilization and equity progress of the family planning program. Data within five periods (2012-2016) enable us to analyze the progress of equity. Based on the National Socio-Economic Survey (Survei Sosial Ekonomi Nasional-SUSENAS) data, we aimed to investigate and describe the progress of utilization and equity in the family planning program and its connections with health insurance ownership, education, household economic status, and geography factor. Findings from this study are expected to provide baseline information for Indonesia policy-makers to improve JKN policy related to the family planning program.

Study design
In this research, a cross-sectional study was performed based on the national secondary database provided by SUSENAS. SUSENAS is one of the regular surveys conducted by the government of Indonesia through the Central Bureau of Statistics (Badan Pusat Statistik-BPS) every year. Data were collected with a questionnaire.
The data contain information on the socio-economic conditions of society, including health conditions, education, fertility, family planning, and housing. SUSENAS surveys include, on average, 300,000 households every year that are spread over 34 provinces and 511 districts/cities in Indonesia. SUSENAS has been conducted since 1979. The sample design of the SUSENAS, which uses probability sampling, allows for the estimation of district-level coverage. Detailed information about the survey and the sampling design with census block allocation are available at http:// microdata.bps.go.id/mikrodata/index.php (in Bahasa) [17].

Study population
The study population included the married woman .

Variables and definitions
Independent variables were health insurance ownership, education, household economic status, and geography factor. Dependent variables were contraceptive use and LACs use by participants. Contraceptive use accumulation obtained CPR and LACs use. Definition of CPR is the percentage of women aged 15-49 years, married, who are currently using, or whose sexual partner is using, at least one method of contraception, regardless of the method used. While, the definition of LACs use is the percentage of women aged 15-49 years, married, who are currently using, or whose sexual partner is using, at least one method of LACs methods, i.e. implant, intrauterine device (IUD), vasectomy, and tubal ligation. We created dichotomous variables for all categories using binary numbers. In the contraceptive use variable, we used 1 for not a contraceptive user and 0 for a contraceptive user. In LACs use, we used 1 for not a LACs user and 0 for LACs user. Health insurance ownership was scored as 1 for not having health insurance and 0 for having health insurance. The geography factor was scored as 1 for household location in a rural area and 0 for household location in an urban area. Education was scored as 1 for illiterate-middle and 0 for high secondary or above. Household socio-economic status was measured using household expenditures, with 1 representing 50% of the lowest and 0 representing >50% of the highest.

Statistical analysis
The analysis was done using Stata version 13.1. The analysis was performed with chi-square (X 2 ) test for categorical variables. The association between dependent and independent variables was measured using the odds ratio (OR), for which the 95% confidence interval (CI) was calculated. Variables were determined to be statistically significant if the result indicated p < 0.05. Variables that showed a statistically significant association (p < 0.25) at the bivariable level were further analyzed at the multivariable level by logistic regression.

Ethics
The raw data of SUSENAS 2012-2016 were used for this study with permission from the Central Bureau of Statistics (BPS). Informed consent was obtained from all study participants by the BPS.

Descriptive statistics
The achievement in CPR during 2012-2016 has shown variation in every province.  Table 1).
The data showed that CPR declined every year (2012-2016) in all of the regions. The decline in urban areas was more than in rural areas between 2012-2016. The decline in the group that did not have health insurance was more than in the group that did. The decline in the non-subsidized health insurance group or the rich group was more than in the subsidized health insurance group or the poor group. In 2013, there was an increase in contraceptive use in the subsidized health insurance group. Preference of non-LACs methods was more common than LACs methods.
While there was only a small increase in LACs use during 2012-2016, preference of contraceptive methods was still dominated by the pill and injection. (See Table 2 Table   3&6) Chi-square analysis indicated some factors that significantly correlated with contraceptive use. Only in 2013, health insurance ownership exhibited a positive influence on contraceptive use. In 2014, the first year of the JKN policy, there was no influence between health insurance ownership and contraceptive use.
Furthermore, there was an inversely proportional correlation in 2015 and 2016.
Participants who did not have health insurance were more likely to use contraceptive than those who did have health insurance. The education factor and household economic status exhibited influence on contraceptive use within 2012-2016, but it was inversely proportional. Participants with illiterate or middle education had a greater probability of using contraceptives than those with high secondary education levels or higher . More than half of the households that had the lowest expenses had a higher likelihood of using contraceptives than 50% of households that had the highest expenses. The geography factor influenced contraceptive use after the implementation of JKN, but it was inversely proportional.
Households located in rural areas had an even higher probability of using contraceptives than did households located in urban areas. (See Table 7) Chi-square analysis indicated some factors that were significantly correlated to LACs use. Health insurance ownership, education, household economy, and geography factor exhibited a positive influence on LACs use within 2012-2016. This influence was stable during 2012-2016. Participants who have health insurance were more likely to use LACs methods than the group that did not have health insurance. Participants with high secondary or above education were more likely to use LACs methods than those in the illiterate or middle education level. More than half of participants who had the highest expense were more likely to use LACs methods than 50% of households that had the lowest expenses. Households located in urban areas were more likely to use LACs methods than households located in rural areas. (See Table 8 Table 10)

Discussion
The influence of health insurance ownership, socio-economic status, and geography factor, on contraceptive use and LACs use, indicated some contrary results. JKN policy did not increase contraceptive use in married women between 15 and 49 years of age, but JKN policy could increase LACs use. An inversely proportional correlation between socio-economic status and contraceptive use was found, while in LACs use was directly proportional. There was a tendency to increase contraceptive use by married women between 15 and 49 years of age in rural areas after the implementation of JKN. On the contrary, LACs use increased more in urban areas than in rural areas after the implementation of JKN. The results of this research showed that JKN did not increase CPR at the national level, and it also did not increase equity in family planning use among regions, especially the eastern areas of Indonesia.
This research had some limitations, which need to be considered when interpreting the results. First, not all of the variables that should be considered exist in the secondary data. Some variables that could have influenced CPR and LACs use are not available in SUSENAS data, such as availability of health facilities and contraceptives, and the distance of health facilities from the locality of residence.
We only identified four determinants that contribute to contraceptive use or CPR and LACs use. Second, we only used a before-after approach with a cross-sectional design to measure equity of family planning utilization. This is one of the standard methods used in health impact assessment. Third, this study only measured equity in utilization of contraceptive use and LACs use. Equity in health financing could not be measured because of data limitations.
Even though our study has some limitations, quantitative analysis was able to be  [12,13,[19][20][21]. Besides the lack of coordination among multiple agencies, the reducing trend of CPR after the implementation of JKN can occur because of the weak regulations. Some regulations related to family planning programs in JKN must be evaluated, viz. non-capitation tariff for IUD and sterilization is assumed too low, no limitation to births, no claim for IUD in hospital, and no guarantee for sterilization after delivery [22]. When the government implemented childbirth insurance, called Jaminan Persalinan-Jampersal, in 2011, there were some provisions to encourage the participants enrolled in Jampersal to join in post-partum family planning. This law did not continue in the JKN era [23]. To ensure that the married women between 15 and 49 years of age who are covered by National Health Insurance (NHI) get the optimal services of family planning, the supply-side of family planning programs, ability to manage by multiple agencies in family planning programs, and the regulations related to family planning in NHI must be properly prepared before and during the implementation of NHI. Every kind of health service that is included in the NHI package has a unique problem in its implementation, and as a result, every health service must have appropriate regulations that are suitable to overcome the problem.

Availability of data and material
The data that support the findings of this study are available from the Indonesian Central Bureau of Statistics (BPS) but restrictions apply to the availability of these data, which were used under license for the current study, and so are not publicly available. Data are however available from the authors upon reasonable request and with permission of the Central Bureau of Statistics (BPS).

Competing interests
The authors declare that they have no competing interests.

Funding
The source of funding for this research is from authors.

Authors' contributions
SKN and LT conceptualized the study. SKN prepared the first draft of the manuscript. SKN acquired the raw data for analysis. YM played a major role in structuring arguments and smoothing out the text. LT contributed to conceptualizing and conceived the idea for the paper. All authors read and approved the final manuscript.