Marital status and abortion among young women in Rupandehi, Nepal

Background Despite liberalization of the Nepal abortion law, young women continue to experience barriers to safe abortion services. We hypothesize that marital status may differentially impact such barriers, given the societal context of Nepal. Methods We evaluated differences in reproductive knowledge and attitudes by marital status with a probability-based, cross-sectional survey of young women in Rupandehi district, Nepal. Participants (N = 600) were surveyed in 2012 on demographics, romantic experiences, media habits, reproductive information, and abortion knowledge and attitudes. We used logistic regression to assess differences by marital status, controlling for age. Results Participants, who comprised never-married (54%) and ever-married women (45%), reported good access to basic reproductive health and abortion information. Social desirability bias might have prevented reporting of premarital romantic and sexual activity given that participants reported more premarital activities for their friends than for themselves. Only 45% knew that abortion was legal, and fewer ever-married women were aware of abortion legality. Never-married women expected more negative responses from having an abortion than ever-married women. Conclusions Findings highlight the need for providing sexual and reproductive health care information and services to young women regardless of marital status.


Background
Since legalizing abortion in 2002 [1], Nepal has made striking progress in rolling out induced abortion services, establishing comprehensive abortion care at public-sector facilities, and providing safe abortion care services to >500,000 Nepali women [2]. The relatively liberal law, which allows for legal termination of pregnancy up to 12 weeks of gestation-or up to 18 weeks in cases of result of rape or incest or at any time if medically indicated-appears to have contributed to an overall national decline in maternal mortality [3]. However, unsafe abortion remains a health concern. Abortion accounted for an estimated 14% of maternal deaths at health care facilities in 2008-2009 [4], which underscores that even in settings where abortion is legal, women may resort to unsafe procedures. In general, barriers to accessing safe abortion care can include negative provider attitudes, fear of repercussion, lack of access to comprehensive sexuality education, limited financial resources, cost of care, transportation, thirdparty involvement laws, and concerns over privacy and confidentiality [5].
Adolescents, in particular, often face challenges in accessing abortion. Young women who obtain abortion care tend to access it later in pregnancy than older women [6] and are more likely to delay seeking help for abortion-related complications [7]. These delays likely are attributable, at least in part, to stigma surrounding adolescent sexuality [8][9][10]. Many youth lack the negotiation and decision-making skills necessary for abstaining from unsafe sexual practices [11]. In Nepal, only 4.2% of adolescents report using a modern form of contraception [12], and youth describe embarrassment from discussing sexual health with parents, relatives and senior community members [13]. Adolescents also perceive sexual health service providers as judgmental and raise concerns about lack of confidentiality.
Less research has been conducted on the relationship between marital status and abortion. The patriarchal and patrilineal nature of society in Nepal translates into a strong preference for sons [14,15], which could influence women's decision-making about and access to abortion. Having a male offspring is considered crucial for continuing the family lineage, providing financial support to parents as they age, and carrying out other roles (e.g., lighting funeral pyres) while female children are often viewed as a financial burden (e.g., because of dowry costs). Married women in Nepal perceive great pressure from their husband and parents-in-law to bear a son [16], and abortion for sex selection has emerged as a concern [17,18]. Despite the illegality of abortion for this purpose in Nepal [1], we hypothesize that the intense pressure that married women are under to have a male child could contribute to a greater acceptance of abortion-regardless of the reason in individual cases-for married women compared to unmarried women.
Despite cultural norms against premarital sexual activity in Nepal [19], the practice appears to be common [20,21]. Social proscriptions against sex outside of marriage, though, could limit unmarried women's access to contraception or safe abortion services. For example, government hospitals and some pharmacies distribute condoms only to married individuals [22], and their purchase by unmarried women is considered culturally unacceptable. A study conducted at the largest women's hospital in Nepal found that most induced abortions were performed on married women [23]. This disparity by marital status could reflect a greater need for abortion services among married women or could result from unmarried women being less likely to access safe services. We hypothesized that regardless of their age, unmarried women in a patriarchal society such as Nepal might face different barriers to obtaining safe abortion care than their married counterparts.

Methods
We conducted a cross-sectional survey during April and May 2012 using a probability-based household sample of 600 women aged 16-24 years in Rupandehi district, Nepal. The district was selected for its strong community partners (for planning future interventions), its geographic accessibility, and its diverse representation of the nation's castes and ethnicities and major religions, namely, Hinduism, Muslim and Buddhism. One of 75 districts in Nepal, Rupandehi district is located in the Lumbini Zone in the Western Development Region and is rapidly urbanizing, with a population growth rate of 3.05 percent in [24].
Trained female interviewers administered a survey to collect information on participant demographics, romantic and sexual experiences, media habits, sources and types of reproductive information, and abortion knowledge and attitudes. The survey was developed in English and translated into Nepali before being pretested with approximately 20 young women from a nearby geographic area, who were expected to be similar to the target population. All interviews were conducted in a private setting. Only women who provided written consent were eligible for participation. The National Health Research Council (NHRC) in Nepal and Allendale institutional review board in the U.S. approved the protocol.

Sampling
The sample was selected with a two-stage cluster sampling design. In the first stage, the clusters (wards) were identified using the probability proportional to size (PPS) method among all wards within the intervention area (using ward populations from 2001 census data) [25]. The sample was conducted from 37 of Rupandehi's 71 main socio-political units (including 69 village development committees (VDCs) and two municipalities, which are each subdivided by population size into approximately nine wards. After listing all intervention wards in alphabetic order by VDC, municipality and ward name, we used the PPS method to select 30 wards.
In the second stage of sampling, we randomly selected 600 respondents from households within each sampled ward. First, we consulted with key informants from the selected wards to create lists of household heads within each selected ward. Then, we used simple random sampling (using a random start and sampling interval) to select 20 households from each list. Using simple sketched maps of the selected wards to locate households, interviewers administered a screening questionnaire to the head or knowledgeable person of the sampled household to determine the residence of any women 16-24 years of age. If multiple women meeting the age criterion resided in the household, only one was randomly selected for participation using a lottery technique. If the selected respondent was unavailable after three visits or if she declined to consent, the nearest neighboring household was substituted.

Analysis
We report simple descriptive statistics for the factors assessed. For socio-demographic characteristics and reproductive history, we assessed differences between never-married and ever-married women using chisquared tests and t tests. We used logistic regression models adjusted for age (in years) in order to identify potential differences by marital status in media habits, type and sources of information on abortion and contraception, and abortion knowledge and perceptions.

Study sample
We visited a total of 805 households, of which 611 households had at least one woman 16-24 years of age. In 11 of the households, no woman was available to be interviewed due to not being present at the time of the interview (n = 6), being sick or physically weak (n = 3) or declining participation (n = 2). Thus, the analysis is based on interviews with 600 participants.

Romantic experiences
Few participants reported having had romantic experiences (i.e., relationships, time alone with a boy, kissing, or sex) outside of the context of marriage or attempting to end an unwanted pregnancy ( Figure 1). However, when asked about the experiences of their friends, participants gave more accounts of each activity, with 49% of participants reporting having a friend with a romantic relationship outside of marriage, 20% having a friend have sex outside of marriage and 8% having a friend attempt abortion.

Media habits
More than half of respondents (53%) reported viewing television daily, with another 24% watching television at least once per week. Furthermore, 38% and 23% of women listened to the radio either daily or at least once per week, respectively. Print media were relatively popular among participants, with 40% reporting newspaper or magazine use. Fewer women (10%) reported any internet use. Among those using the internet, most did so in their own homes (68%) and many accessed the social website Facebook (86%). Reported attendance at youth or school clubs was low (5%). Notably, a higher proportion of never-married participants engaged with media sources than ever-married participants, particularly with regard to daily television (60% vs. 45%, respectively; p < 0.01), daily radio (43% vs. 32%, respectively; p < 0.01) and print media (52% vs. 25%, respectively; p < 0.01).

Types and sources of reproductive information
Most participants reported receiving information about contraceptive methods (89%) and sources of services (85%) in the past year (Table 2). Fewer were informed on their costs (28%), directions on use (11%) and side effects (11%). Regarding abortion, while most women reported receiving information in the past year on where to obtain a safe abortion (76%), fewer were informed about abortion methods (52%), costs (31%), and possible complications (43%). Most participants were able to identify medical abortion (82%) and dilation and curettage (69%) as methods of abortion while few (8%) had heard of manual vacuum aspiration (MVA). Marital status did not appear to be related to the type of contraceptive information received in the past year. In contrast, married women were more likely to report receiving information on abortion costs (p = <0.01), possible complications (p = 0.02), legality (p = 0.02) and awareness of abortion methods (p = 0.02) than never-married women after controlling for age.
Women received reproductive-health information from a variety of sources (Figure 2a and b). Not surprisingly, the most popular source of information on contraception and abortion was an interpersonal one: a friend or neighbor (67% and 56%, respectively). Television and radio also were common sources of information on contraception (52% and 50%, respectively) and, to a lesser extent, on abortion (45% and 27%, respectively). Female community health volunteers (FCHVs) were the most frequently cited among the professional and community sources for information on contraception (33%) and abortion (24%). Sources of information differed by marital status. For example, never-married participants were more likely to rely on television or radio for both contraception and abortion information whereas married women were more likely to receive information from FCHVs and health care providers.

Abortion knowledge and attitudes
Participants had low levels of knowledge about the legal status of abortion in Nepal: only 45% (N = 271) of respondents knew that termination of pregnancy was legal. Furthermore, the details of the law were not well known. For example, only 28% of women knew that 16 was the legal age of consent for abortion (i.e., younger women are required to be accompanied by a legal guardian). Only 33% of participants knew that being unmarried did not automatically make abortion illegal and only 37% knew that having an unintended pregnancies as the result of contraceptive failure does not preclude women from accessing legal abortions. Ever-married women consistently were less informed than never-married women even after controlling for age ( Figure 3).
Respondents were asked whether they agreed or disagreed with a series of statements related to abortion practices (Table 3). Opinions were favorable for a number of important items such as support if friends needed an abortion (86%); feeling comfortable in talking with health care providers about abortion (74%); and feeling comfortable initiating talks with my friends about abortion-related issues (70%). However, participants believed that abortion costs and acceptability differ by the marital status of the woman involved. Most respondents believed that doctors usually charge more for an abortion for unmarried women (68%) and that abortion is more acceptable for married women (47%) than for unmarried women (16%) in their communities. Furthermore, never-married women appeared more comfortable with the idea of abortion than ever-married women. More never-married women than their evermarried counterparts reported commonly talking about abortion-related issues with their friends (70% vs. 65%, respectively; p = 0.01); being comfortable initiating talks with their friends about abortion-related issues (73% vs. 67%, respectively; p < 0.01); confident about their ability to help a friend find abortion services if needed (81% vs. 74%; p < 0.01); saying the words "vagina" and "penis" without embarrassment (18% vs. 11%; p < 0.01) and supporting a friend if she needed an abortion (88% vs. 83%; p = 0.02). On the other hand, more never-married women than ever-married women expected to be treated badly if they had an abortion (70% vs. 55%; p <0.01) and fewer imagined that health care workers would treat them kindly if they needed an abortion (37% vs. 54%, respectively; p < 0.01).

Discussion
In this household-based survey conducted in Rupandehi district in Nepal, young (16-24 years of age) women reported having good access to information on reproductive health: most women had received information on sources and methods of contraception and abortion care in the past year. Furthermore, most women (88%) could name a method of abortion, most often medical abortion and, to a lesser extent, dilation and curettage. Few participants (7%) mentioned MVA as a possible method. Programmatic efforts should focus on increasing young women's knowledge of MVA and equipping providers with the proper training and capacity to perform this safer alternative to dilation and curettage [26]. In Nepal, medication abortion is available at primary health centers while MVA services are accessible at higher levels of the health system. Improving women's understanding of medication abortion regimens, and the optimal gestational age for accessing these services, could further improve reproductive health outcomes. Participants denied having engaged in romantic activity, including sexual activity, outside of marriage. However, they were much more likely to report that their friends engaged in romantic relationships, spending time alone with a boy, kissing, and sexual contact outside of marriage. These findings suggest that social desirability bias might prevent women from admitting to having premarital experiences. This interpretation is consistent with a survey of students conducted at three colleges in Nepal, in which premarital sex was reported by 45% of men but only by 4% of women [27]. Similarly, a study of young factory workers in Kathmandu Valley found that 35% and 16% of unmarried males and females, respectively, reported being sexually experienced [20]. Finally, in a survey of male college students, 39% reported having had premarital sex [21]. Together with the literature, the present findings suggest that despite prevailing social norms, romantic and sexual activity among unmarried youth in Nepal, particularly in urban areas, is a reality.
Knowledge of Nepal's abortion law was low: only 45% of women were even aware that abortion is legal in  Nepal, and substantial proportions of women erroneously believed that a range of conditions (e.g., marital status, having a pregnancy result from a contraceptive failure or not wanting or not being able to afford a child) would affect the legal right to an abortion. This finding is consistent with previous research documenting low levels of awareness of the legal status of abortion [28][29][30], including an earlier assessment in the same district, which included reports of "medical halls" where women go for unsafe abortion [31]. The lack of awareness of legal status appears to result in women seeking unsafe abortion [32]. We found that ever-married women, regardless of their age, were consistently less informed than never-married women about the legality of abortion. Women's sources for reproductive health information also differed by marital status. Never-married women were more likely than ever-married women to report receiving contraception or abortion information in the past year from television, radio, or a friend or neighbor. In contrast, ever-married women were more likely than never-married to have obtained contraceptive or abortion information from a FCHV, health care provider or their husband. The important role that health care providers hold in disseminating information on contraception and abortion to all women should be emphasized and strengthened.
Although most (74%) of the young women in this survey reported feeling comfortable in talking to a health care provider about abortion-related issues, about 23% (with no difference found by marital status) did not think that they could decline sex if they wanted to. Furthermore, women reported being embarrassed by references to male and female genitals. This lack of selfefficacy regarding sex and fertility is a potential barrier to seeking proper reproductive health care.
Women's perceptions of abortion-related factors also differed by their marital status. In general, never-married women tended to be more comfortable and supportive of abortion. They were more likely to talk about abortion topics with their friends and were more comfortable with these discussions than were married-women. They also were more likely to indicate that they would support a friend who needed an abortion or feel confident that they could help a friend obtain the procedure. However, never-married women expected to experience more negative responses from health care workers, family and friends if they were to have an abortion than were married-women. Furthermore, both never and evermarried women agreed that abortion among unmarried women was not accepted in their community (88% and 80%, respectively). Abortion among unmarried women in Nepal is strongly condemned, and providers have voiced concerns that the legalization of abortion is leading to an increase in utilization of services among unmarried women [33]. Both society, in general, and health care providers are more accepting of abortion among married women. For example, a qualitative study of abortion providers in Kathmandu Valley revealed that even though providers perceived abortion for sex selection to be unethical, they still empathized with married women who faced pressure to have a male offspring and thought that abortion for this purpose should be available for some women [34].

Conclusions
Given the early mean age of childbearing among women in Nepal [35], sexual activity among adolescents clearly is common and underscores the importance of access to contraceptive and abortion care among youth. Less attention, though, has been directed to differences in the provision of information and services to unmarried women. This could be especially important in patriarchal societies, which could be expected to have strong norms against sexual agency of unmarried women. The present study demonstrated substantial differences in abortion information, experiences and perception by marital status and highlights the need for the provision of sexual and reproductive health care knowledge and services to all women regardless of marital status.