Factors Inuencing Unintended Pregnancies among Childbearing Age Women in Indonesia

Background: Unintended pregnancy can cause pregnancy termination, which leads to safety risks. The purpose of the study was to analyze the factors that inuence unintended pregnancies in Indonesia. Methods: The analysis unit was women aged 15-49 years old who gave birth to the last 5 years. The sample size was 36,472 women. In addition to unintended pregnancy as the dependent variable, other variables analyzed were the place of residence, age, education, husband/partner, employment, wealth, parity, pregnancy termination, the person deciding woman's access to health care, heard about family planning messages on radio, television, and newspaper/magazines. The nal stage analysis uses binary logistic regression. Results: Women in urban areas were 1.834 times more likely than women in rural areas to experience an unintended pregnancy. The 20-24 age group was 0.202 times more than the 15-19, while the 45-49 was 1.916 times compared to the 15-19 to experience an unintended pregnancy. Secondary education women were 1.447 times more likely than no education women, while the poorer women were 1.190 times more likely than the poorest women to experience an unintended pregnancy. Parity was found to be a strong determinant of unintended pregnancy. History of pregnancy, decision making by husbandpartner, and heard about family planning messages on radio and television in the last few months are risk factors for unintended pregnancy. Conclusions: Eight variables affect unintended pregnancy, namely age, education, wealth, parity, pregnancy termination, the person deciding woman's access to health care, and heard about family planning messages on radio and television.


Background
Unintended pregnancy can lead to pregnancy termination. Unintended pregnancy increases the risk of safety, which can end in death 1 . It is estimated that every year there are 85 million women in the world facing unintended pregnancy 2 . While the United Nations launched a WHO report stating that the major public health problems of unplanned pregnancies globally in low and middle-income countries are on such a scale that 74 million women have unintended pregnancies each year. This condition causes about 25 million unsafe abortions. It further contributes to 47,000 maternal deaths each year 3 .
The WHO report states that between 2010-2014, every year around the world an average of 56 million induced abortions, both safe and unsafe. Of these, there were 35 abortions induced per 1000 women aged between 15-44 years. 25% of all pregnancies end in induced abortion. Abortion rates are higher in developing countries than in developed countries. An estimated 25 million unsafe abortions take place worldwide each year. WHO states that almost all unsafe abortions exist in developing countries 4 .
In Indonesia, a previous study analyzing the 2012 Indonesian Demographic and Health Survey informed that the prevalence of women experiencing unintended pregnancy in Indonesia was 8%. The highest prevalence is in the province of Central Sulawesi (11.9%) and the lowest in Papua (2.9%). The bivariate analysis found a signi cant relationship between economic level, education level, place of residence, marital status, parity, birth distance, and age, with an unintended pregnancy. In multivariate analysis, the researchers concluded that marital status was the most dominant factor related to unintended pregnancy 5 . Another study conducted in nine major cities in Indonesia informed that there were around 37,000 unintended pregnancies. Of this 27% are unmarried couples. While 12.5% are junior/senior high school students 6 .
Studying unintended pregnancy is felt to be very important for the safety of women in Indonesia.
Information on the results of this study provides clear and targeted goals for policymakers to maximize efforts to prevent unsafe abortion due to unintended pregnancy in women. Based on the background, the study aimed to analyze the factors that in uence unintended pregnancies among childbearing age women in Indonesia.

Data Source
The study utilizes the 2017 Indonesian Demographic Data Survey (IDHS) as an analysis material. The unit of analysis in this study was women aged 15-49 years old who had given birth in the last 5 years.
The sample size of 36,472 respondents.

Data Analysis
The dependent variable in this study was unintended pregnancy. Unintended pregnancy is de ned and calculated as a pregnancy that is either unwanted or mistimed. A woman who has an unwanted pregnancy does not want to be pregnant or have any children. A woman who has a mistimed pregnancy does not want to be pregnant at this time but wants the pregnancy later 7 .
Other variables analyzed as independent variables are the type of place of residence, age group, education level, having husband/partner, employment status, wealth status, parity, history of pregnancy termination, the person deciding woman's access to health care, hearing about family planning messages on radio, heard about family planning messages on television, and heard about family planning messages on newspaper/magazines.
The type of place of residence is divided into two criteria, urban and rural-based on the criteria determined by the Central Bureau of Statistics. Age groups are divided into 7 groups on a 5-year basis, namely 15-19, 20-24, 25-29, 30-34, 35-39, 40-44, and 45-49. The education level is the respondents' acknowledgment of her latest education level. This variable is divided into 4 criteria, such as no education, primary, secondary, and higher education. Having a husband/partner is divided into 2 categories, which are not owning and owning. Don't have a husband/partner is never in a union, widowed, divorced, or no longer living together/separated, have a husband/partner is married, or living with a partner. Employment status is divided into 2 categories, not employed and employed.
Wealth status is the respondents' acknowledgment of the wealth quintile in a household. Household wealth was assessed based on types of furniture and their prices. It counts a television, a bicycle or a car, and household goods, such as drinking water sources, bathroom amenities, and main building materials for ooring. The assessment of this variable was calculated with the principal component analysis.
National wealth quintiles were arranged based on household scores for each person in the household and then divided into the same ve categories, each of which contributes 20% of the population based on the distribution 8 .
Parity is the frequency of giving birth to a live baby. Parity is divided into 2 categories, namely primiparous (< 2) and multiparous (≥ 2). History of pregnancy termination is determined based on the respondent's recognition, divided into 2 categories, namely experiencing and not experiencing pregnancy termination. The person deciding woman's access to health care, determined based on the respondent's acknowledgment, is divided into 4 categories, namely respondent alone, respondent and husband/partner, husband/partner only, and someone else. Heard about family planning messages on radio, heard about family planning messages on television, and heard about family planning messages on newspaper/magazines, determined based on the respondent's acknowledgment, which is divided into 2 categories, namely having heard a message and never hearing a message.
All variables used in this study are categorical variables, so the Chi-square test was used to select variables related to unintended pregnancy. Because of the nature of the dependent variable, Binary Logistic Regression was used for multivariate nal determination. The author uses SPSS 22 software to help all stages of statistical analysis.

Source
The map depicted in the image belongs to the author Table 1 displays the results for the collinearity test of unintended pregnancy among childbearing age women in Indonesia. The analysis shows that there is no co-linearity between the dependent and independent variables. Table 1 informs that the tolerance values of all variables are greater than 0.10.
While the VIF value for all variables is less than 10.00. Referring to the basis of decision making in the multicollinearity test, it can be concluded that there are no symptoms of multicollinearity in the regression model.  Table 2 is a statistical display of the description of respondents of unintended pregnancy among childbearing age women in Indonesia. Table 2 informs that unintended pregnancy is more dominant in urban areas. Based on age groups, although all age groups experience an unintended pregnancy, the incidence of unintended pregnancy is dominated by childbearing age women in the 35-39 category. Table 2 shows that based on the education level category, unintended pregnancy is dominated by women with secondary education. Unintended pregnancy is also dominated by women who have husbands/partners, employed women, the poorest women, and multiparous women.
Based on the history of the pregnancy termination category, unintended pregnancy is dominated by women who have never had a pregnancy termination. Based on the person deciding a woman's access to healthcare, unintended pregnancy is dominated by decision-making criteria by respondent and husband/partner. Note: * p < 0.05; * * p < 0.01; * * * p < 0.001. Note: * p < 0.05; * * p < 0.01; * * * p < 0.001.     Table 3 shows that multiparous women are 12.220 times more likely than primiparous women to experience unintended pregnancy (OR 12.220; 95% CI 8.630-17.304). This information shows that parity is a strong determinant of the incidence of unintended pregnancy among childbearing age women in Indonesia. Table 3 informs that women who have a history of pregnancy termination are 1.092 times more likely than women who do not have a history of pregnancy termination to experience unintended pregnancy (OR 1.092; 95% CI 1.011-1.180). This information shows that the history of pregnancy termination is a risk factor for unintended pregnancy among childbearing age women in Indonesia.
Based on the person who decides woman's access to healthcare, women who have a husband/partner only decision-maker are 1.172 times more likely than women who make their own decisions to experience unintended pregnancy (OR 1.172; 95% CI 1.050-1.307). This information shows that decision making by husband/partner only is a risk factor for unintended pregnancy among childbearing age women in Indonesia. Table 3 shows that women who were heard about family planning messages on the radio in the last few months were 1.336 times more likely than women who were not heard about family planning messages on the radio in the last few months to experience unintended pregnancy (OR 1.336; 95% CI 1.191-1.498).
Women who were heard about family planning messages on television in the last few months were 1.120 times more likely than women who were not heard about family planning messages on television in the last few months to experience unintended pregnancy (OR 1.120; 95% CI 1.046-1.200). This information shows that heard about family planning messages on radio or television in the last few months is a risk factor for unintended pregnancy among childbearing age women in Indonesia.

Discussion
The results of the analysis found that women who live in urban areas are more likely than women who live in rural areas to experience an unintended pregnancy. This information contradicts previous research in Pakistan, Ethiopia and several countries in sub-Saharan Africa that informs that living in rural areas is a risk factor for unintended pregnancy in women compared to women living in urban areas [9][10][11] . In the Indonesian context, the availability of health facilities is better in urban areas than rural areas 12,13 , including the availability of contraceptives.
The results found that the age group is partially a determinant of the unintended pregnancy among childbearing age women in Indonesia. Age as a determinant unintended pregnancy was also found in previous studies 9,14,15 . Age other than related to the unpreparedness of women when they were teenagers 16,17 , also related to the number of children desired and the reproductive ability of women 18 .
The results found that education level and wealth status are partially determinants of the unintended pregnancy among childbearing age women in Indonesia. This nding strengthens the results of previous studies that found the information to be in line 10,11,19 . Education and wealth status has always been found to be in uential in many ways related to health. A person with good wealth status is often well educated so that he has good health output 20,21 . A systematic review found in several studies that nancial incentives also seem to effectively increase women's contraception intake 22 . This kind of nancing intervention is needed to overcome the barrier of contraceptive use among poor women to minimize the occurrence of unintended pregnancy.
Information on the results of the study shows that parity is a strong determinant of the incidence of unintended pregnancy among childbearing age women in Indonesia. Multiparous women who feel that the number of children they have is more vulnerable to unintended pregnancy. This condition may be related to plans to not become pregnant or use contraception that failed 10,23,24 . The same reason also applies to the history of the pregnancy termination category. Women who have experienced pregnancy termination tend to have a higher chance of experiencing an unintended pregnancy compared to those who have never experienced pregnancy termination 14 .
The analysis found that husband/partner only decision making was a risk factor for unintended pregnancy among childbearing age women in Indonesia. This condition is related to the lack or deadlock of communication between women and their partners 14,18,19 . Several previous studies have found the same information, unintended pregnancy is closely related to women's independence in making decisions about their reproductive health 25,26 .
The results found that hearing about family planning messages on radio or television in the last few months is a risk factor for unintended pregnancy among childbearing age women in Indonesia. This information is in line with the ndings of previous studies in Mozambique and Ghana 1 . While con icting results were reported in a previous study in Northwest Ethiopia. The study found that women with no access/exposure to mass-media have a higher risk than those who have access/exposure to mass-media to experience unintended pregnancy 27 .
This information also indicates the type of media most frequently accessed by women for family planning. Women prefer media with messages based on audio (radio) or a combination of visual and audio (television) [28][29][30] . This information is useful for policymakers or family planning activists to better utilize the two media categories as a medium for delivering messages about family planning.

Conclusions
Based on the results of the analysis it could be concluded that eight variables affect unintended pregnancy. The eight variables were age group, education level, wealth status, parity, history of pregnancy termination, the person deciding woman's access to health care, heard about family planning messages on radio and television.

Declarations
Ethics approval and consent to participate The 2017 IDHS has passed ethical clearance from the National Ethics Committee. The respondents' identities have all been deleted from the dataset. Respondents have provided written approval for their involvement in the study. Informed consent for participation in the study was obtained where participants are children (under 16 years old) from their parent or guardian. The author has obtained permission for the use of data for this study through the website: https://dhsprogram.com/data/new-userregistration.cfm.

Consent for publication
Not applicable

Competing interests
The authors declare that they have no competing interests Funding Not applicable Availability of data and materials Data cannot be shared publicly because of the data are owned by a third party and authors do not have permission to share the data. The 2017 IDHS data set name requested from the ICF ('data set of childbearing age women') are available from the ICF contact via https://dhsprogram.com/data/newuserregistration.cfm) for researchers who meet the criteria for access to con dential data. Figure 1 Distribution map of unintended pregnancy among childbearing age women by the province in Indonesia, 2017 (36,472). Source: The map depicted in the image belongs to the author