Stigma and agency: exploring young Kenyan women’s experiences with abortion stigma and individual agency

ABSTRACT Although abortion is now legal in Kenya under expanded circumstances, access is limited and many providers and individuals still believe it is illegal. This study aimed to characterise Kenyan women’s perceptions and experiences with abortion and post-abortion care (PAC) services in Nairobi regarding barriers to care, beliefs about abortion, and perceived stigma. We conducted 15 semi-structured in-depth interviews with Kenyan women aged 18–24 years who recently received abortion and PAC services at four Marie Stopes Kenya clinic sites in Nairobi. The most significant psychosocial barrier respondents faced in promptly seeking abortion and PAC was perceived stigma. In response to stigma, participants developed a sense of agency and self-reliance, which allowed them to prioritise their own healthcare needs over the concerns of others. To adequately address perceived stigma as a barrier to abortion- and PAC-seeking, significant cultural norm shifting is required.


Background
Until 2010, Kenya's abortion law was highly restrictive and abortion was only permitted to save the life of the woman. 1 In 2010, the new constitution was signed into effect, which added the clause "and to protect the woman's health." 1 As a result, abortion is now considered permissible under four key circumstances: if the life of the mother is in danger, if the health of the mother is in danger, if providing emergency care, and if allowed by any other law. 1,2 This clause is interpreted by some to mean that abortion is now legal under the Kenyan constitution. However, many providers are still extremely wary of whether or not they would receive legal protection under this new constitutional provision if they were to begin providing safe abortion services. 3,4 Since the enactment of the new constitution, no new legislation has been put forth to clarify or provide guidance on the issue; therefore, the ambiguity remains and the fierce national debate on abortion rages on. 1,2 This ambiguity and debate translate into confusion about the circumstances under which women can now seek and legally receive abortions.
Furthermore, in the Kenyan context, traditional sociocultural norms and religious beliefs continue to stigmatise and condemn abortion, such that even if it were more legally accessible, use would be limited. 1 Abortion in Kenya, as in many cultures, is seen as a transgression of gender-based archetypes for how a woman should ideally behave, including the idea that female sexuality is for procreation purposes only and that women are born with an innate desire to be a mother. 5 For this reason, abortion stigma is produced and reproduced in two major ways. 6 First, felt or perceived stigma is the result of other people's abortion attitudes and cultural expectations regarding how a woman should respond to news about her pregnancy. 7 Second, internalised stigma is the result of a woman's acceptance of negative cultural attitudes about abortion. 7 Evidence of both types of abortion stigma in Kenya include studies from focus group discussions with men and women from central Kenya, which found that the most common male perception of abortion was that it was "women's strategy for concealing their waywardness, unfaithfulness and promiscuity," so women "can do all their evil and [just] abort when they become pregnant." 8,9 Female study participants responded that because of severely felt stigma "women seeking to terminate an unintended pregnancy enjoyed no respect, sympathy, or support." 10 Furthermore, they internalised this belief and repeatedly stated that abortion was "the worst thing you could do as a woman … if it is found out, you will lose every respect you have" and people will immediately shun you. 10 Consequently, the stigma of abortion plays a significant role in women's decision to access safe or unsafe abortion services. 11 Unsafe abortion persists in Kenya, in part as a result of widespread abortion stigma. 12 Kenya's maternal mortality ratio due to unsafe abortion remains one of the highest in East Africa at 266 deaths/ 100,000 live births. 13 Similarly, unsafe abortion is predicted to account for nearly 35% of maternal deaths in Kenya, compared to 13% globally and 18% in East Africa. 14,15 Further evidence of the growing abortion problem in Kenya is provided by a 2015 mixed methods study. Mohamed et al. found that the number of induced abortions in Kenya in 2012 was approximately 464,700. 1 Of the 464,700 women who had abortions, more than 116,000 (25%) were treated for abortion-related complications. Nearly 87,000 (75%) of those complications were moderate to severe, and nearly 40,000 (45%) of those complications occurred in women aged 19 years and younger. 1 In addition, the age-specific abortion rate showed that the highest incidence of unsafe abortion occurred amongst Kenyan women less than 25 years of age. This is in line with other similar research conducted amongst Kenyan women seeking induced and post-abortion care services. Yegon et al. established in 2016 that age, marital status, type of abortion service (i.e. induced abortion or post-abortion care), and socioeconomic status of respondents are all significantly associated with higher levels of abortion stigma. 11 Young, unmarried women, especially those from lower socioeconomic strata, are much more likely to bear the brunt of abortion stigma in Kenya. 16 Survival from abortion-related complications is most closely associated with prompt post-abortion care-(PAC) seeking. 17,18 PAC is defined as the treatment of obstetric complications from unsafe abortion and incomplete spontaneous abortion. 15 Core components of PAC include emergency treatment of obstetric complications, post-abortion counselling, contraception provision, STI and HIV education, provider referral for other sexual and reproductive health (SRH) concerns, and community partnerships between patients and providers to prevent unintended pregnancy and unsafe abortion. 15 Kenya's national healthcare provision guidelines require that PAC services be made available at all level II health facilities and up. 17 Level II health facilities in Kenya are defined as primary healthcare clinics with dispensaries. These facilities are also the first point of contact with care providers, especially in rural areas. 17 For this reason, level II facilities are often overwhelmed, so women seeking PAC services either get referred to higher level health facilities or are made to wait longer before treatment is initiated. 17 One prospective cohort study of maternal mortality in Libreville, Gabon found that the mean delay between patient admission and initiation of treatment was 1 hour for eclampsia, 1.3 hours for postpartum haemorrhage, and 23.8 hours for abortionrelated complications. 19 Similar studies on PAC in Kenya have shown that young maternal age (less than 25 years old), marital status (unmarried), unintended pregnancies and severe abortion complications all put women at significantly increased risk of delays in care-seeking, and therefore at higher risk of maternal mortality. 17 Given the vulnerability of young Kenyan women and the extensive delays in healthcare delivery at public facilities, Marie Stopes International (MSI), a private charity based in the UK, stepped in to fill the gap in Kenya's healthcare provision. 20 The largest specialised SRH and family planning organisation in Kenya since 1985, MSI provides a wide array of high-quality and client-centred SRH services to men, women and adolescents. 21 Together, the 15 mobile outreach teams, 23 clinics, 400 social franchises around the country, and one specialist obstetric hospital in Nairobi offer a full range of services: family planning options, emergency contraception, health education counselling, pre-and post-natal care, safe labour and delivery, ultrasound, prevention of mother to child transmission (PMTCT) of HIV/AIDS, cervical cancer screening and PAC. 21 In 2014 alone, Marie Stopes Kenya (MSK) helped to prevent 153,000 unsafe abortions through its provision of contraception and PAC services. 21 However, little is known about Kenyan women's qualitative experiences seeking induced abortion and PAC services at MSK facilities. The aim of this study was to understand young Kenyan women's experiences with induced abortion and PAC services in Nairobi, since women under 25 years are known to be at increased risk of delayed abortion care-seeking and face higher levels of abortion stigma. We aimed to characterise the quality, barriers, cultural beliefs and community norms around induced abortion and PAC-seeking in Nairobi using in-depth, qualitative interviews with young women who recently sought abortion services.

Participants and procedures
The sample population for this study was composed of young Kenyan women between the ages of 18-24 years who recently received induced or PAC services (either medical or surgical) at one of four MSK clinic sites in Nairobi. For the purposes of this study, "recent" was defined as individuals who received abortion and PAC services that day so that interviewees could easily recall information about the quality of care they received.
The four MSK clinic sites (Table 1) differ in their locations throughout Nairobi, the level of specialty care they provide, and the cost of abortion services. The largest facility (site A) serves as a fulltime maternity hospital and nursing home. For this reason, clients who suffered from complications or who required specialty care were sent to that facility for overnight observation and treatment. In comparison, the three other clinics were considerably smaller and primarily provided outpatient services, such as general wellness visits for expecting mothers, family planning consultations, and induced abortion and services.
For this study, convenience sampling was used to recruit women under 25 years. The initial plan was to purposively sample and then stratify women aged <20 years and those aged 20-25 years. However, the numbers of young Kenyan women who received induced and PAC services at the MSK clinics each day was not large enough to use these sampling methods. Given that young Kenyan women are more likely to experience high levels of abortion stigma, the intention behind purposive sampling and stratification was to explore potential differences in perceived stigma and barriers to care seeking based on age. Recruitment was done in close collaboration with the MSK clinic staff to ensure that participants were all well enough to participate in a 30-60-minute interview after completing counselling and *Respondents were all given the option to pursue either a medical or surgical abortion. Medical abortions were provided through a combination of mifepristone and misoprostol, while surgical abortions were provided through manual vacuum aspiration (MVA). 24 **The respondent who used multiple methods of contraception identified using both condoms and injectables in the past. treatment. To reduce the rate of participant refusal and to compensate participants for their time, non-monetary incentives were provided to interviewees, in the form of Airtime scratch cards (150 Ksh). The exact number of in-depth qualitative interviews to conduct was determined by the point that data collection reached saturation. After conducting 15 interviews, saturation had been reached. The content of each interview was based on the semi-structured interview guide. Themes or constructs addressed during the interviews included: barriers to seeking abortion care, participants' beliefs about abortion and types of felt abortion stigma. The questions about barriers to care asked about both physical and psychosocial barriers, such as concerns regarding distance of the facility or fear of judgment from others. Participants were then asked in more detail about how social norms, religious and cultural beliefs influenced their decision to seek abortion services. The third construct asked participants about different types of perceived abortion stigma, such as individual stigma or judgment from others, and how participants persisted despite the stigma.

Data collection and analysis
In-depth semi-structured interviews were conducted in English or Kiswahili with 15 study participants. Individuals who were eligible for the study were referred by MSK reception staff prior to receiving treatment. After completing treatment, individuals who were interested in participating were brought to interview staff by their provider. From there, participants were reassured that should they choose to participate, their responses would not in any way influence their relationship with MSK, or their ability to seek future treatment at the facility. Interviews were conducted in private rooms within each clinic, to maintain privacy for each of the participants.
Qualitative data from this study were analysed using the grounded theory approach. 22,23 The data analysis procedure involved: translation, transcription, back-checking, coding and refinement. First, the interview recordings were translated and transcribed by a transcriptionist. Transcripts were then back-checked for coherency, consistency and quality by fieldwork staff. Afterwards, transcripts were read through multiple times in their entirety by the lead researcher and the co-investigator to allow for familiarisation with the text. Transcripts were subsequently coded based on important themes and the codes were then organised into different groups of related themes. This was followed with a line-by-line microanalysis using open coding, so that codes were reflective of each topic and concept discussed. To help organise, manage and retrieve codes, the data analysis software Atlas-ti was used.
After carefully studying the interview transcripts and field notes, data analysis was conducted by the lead researcher and the co-investigator. Each researcher coded the same three interviews to start, and then discussed discrepancies until an agreement was reached and the codebook was finalised. After that, a single codebook was created to use on the remaining interviews. The remaining transcripts were then divided equally between the two researchers. This was followed by discussions on how to group the codes together and analyse the interview findings.

Participant characteristics
Study participants received abortion services at one of four MSK clinic sites in Nairobi. Characteristics of the study population are summarised in Table 1.
Nearly all (n = 12) respondents found out about the clinic through word of mouth, either from a friend or family member who had been to the clinic themselves, or who knew someone who had used MSK services in the past and had positive experiences there. A few (n = 3) found out about MSK through the internet.

Women's perceptions of quality of abortion care
When asked what prompted them to choose MSK, most participants cited the primary reason as perceptions of high-quality care and assurances of confidentiality. Nearly all participants (n = 14) reported feeling satisfied with the quality of care they received. Many attributed the high-quality care to the kindness and professionalism of the MSK staff, the stigma-free environment, and the active role they played in decisions about their care.
"On a scale of 1-10 I think I'll put them at [ Privacy and confidentiality also played a large role in subjects' decision-making.
"First of all when I tried to get information on the internet, I couldn't get any patient details, which by itself shows that all the records are kept inhouse. There's no slandering of patients on the internet or public records or anything of that sort. [Actually,] not getting information [about] patients outside [the facility] was the first hint for me that this might actually work. Then, coming and signing consent forms before you do anything, that confirmed that they will try and keep it private, which was very important to me when I decided to come here." [Respondent aged 23 years, interviewed at KenCom] High-quality abortion care was frequently cited as one of the most common reasons for choosing MSK facilities for induced abortion and PAC services. All participants reported feeling that abortion services were equally available, and treatment was equally provided, regardless of age, marital status, or other personal attributes.
Women's experiences with barriers to seeking healthcare services Half the respondents (n = 7) stated that physical barriers (e.g. lack of finances or information) played a significant role in how women access, finance and receive high-quality healthcare services. The other half (n = 8) attributed psychosocial and cultural barriers (e.g. worries about judgment and fear of stigma) as to why women often did not present at healthcare facilities for services.
"Maybe they feel people will judge them, I think that is the biggest reason, because most of the people will judge her. When someone is pregnant, and she is not ready for that pregnancy, she goes to [the] facility, but she cannot do something because she is afraid of what people will say about her. So judgment is a big fact[o]r]." [Respondent aged 22 years, interviewed at site C] Other barriers to care included: instability at home (e.g. problems with one's spouse), stigma related to one's disease state (i.e. having HIV/AIDS or cancer), the role women are expected to play in the household (i.e. women are often too occupied with domestic responsibilities to have time to seek out healthcare services for themselves), and fear of Western medicine. Regardless of the barrier in question, respondents shared that all have the potential to result in the same effect: women being denied or denying themselves potentially lifesaving medical care.

Beliefs about abortion and women who have abortions
When participants were asked further about their decision to seek out abortion services, they mentioned that religious beliefs, community norms and cultural traditions directly and indirectly shaped their decisions around abortion and PACseeking. For example, many rural communities and religious groups have a preference for traditional birth attendants (TBAs) and healers rather than hospitals. As a result, women are strongly discouraged from seeking abortion services at healthcare facilities.
"My community is Maasai-Kikuyu. So my Maasai side believes in getting wakunga (Swahili name referring to mid-wives) when you give birth. Even when you want to terminate a pregnancy, they call all the elderly women and they sit on your stomach repeatedly. Like they bounce on it till it's dead … It's really scary." [Respondent aged 24 years, interviewed at site B] When asked why religious beliefs and cultural traditions such as these heavily influenced women's ability to access abortion-related care, participants explained that there is a near ubiquitous belief in Nairobi that abortion is morally wrong, a sin, and illegal. Furthermore, respondents traced this belief back to a firm faith in God and the idea that life is sacred.
"In my religion, we are not allowed to abort a child and apart from religion, it's something that is not right, that's why it was made illegal, right? It's something not good to kill a child." [Respondent aged 20 years, interviewed at site A] In addition to strong religious and cultural beliefs preventing women from seeking out abortion services at healthcare facilities, many communities also use stigma, isolation and shame as tools to ensure that women do not break from tradition. Respondents explained that if discovered, women who seek abortion services are often publicly denounced as immoral and hypersexual. As a result, they may be shunned, disowned, or even killed. Overall, participants reported a large amount of variation in their personal experiences with social stigma and cultural beliefs related to abortion. However, the underlying belief that abortion is morally wrong and a sin remained unanimous, as well as the belief that women who have abortions are promiscuous and shameful. In response to strong social stigma about abortion, some respondents internalised these beliefs and expressed that women who seek abortions "deserve" to be stigmatised or mistreated.

Dealing with judgment and gossip from others
Similarly, many women (n = 10) felt that if other people were to find out about their abortion, people would likely judge, gossip about, or be disappointed in them.
"R: 'I am afraid if maybe my mother, sisters or friends get to hear about my abortion, they'll be disappointed so much and they'll say I am a stupid person who does not respect herself. My boyfriend maybe … will disrespect me and take me as a joker because … he might see me as a person who's not responsible with her life." [Respondent aged 23 years, interviewed at site A] However, there was also an overall acceptance from study participants that people will judge or gossip about them regardless. One participant explained, "it's life, gossip is everywhere so I just have to accept it" [Respondent aged 20 years, interviewed at site A].
Other women felt that there was no point in worrying about judgment or gossip because the decision had been made and the abortion had already taken place. This sense of self-reliance and self-determination exhibited by participants helps to illustrate a larger theme of abortion being a personal choice for them. These women did not worry about other people's views; rather, they focused largely on their own views. One participant explained that even if people did gossip, it would not affect her because " … if they do, it was a personal choice. I paid it, they had nothing to do with it" [Respondent aged 24  Many of these young women are responding to felt and internalised abortion stigma by prioritising their own healthcare needs. In doing so, they are asserting their agency and placing their own present and future well-being above the perceptions and beliefs of others. Responding to abortion stigma through selfreliance and agency Despite the heavy presence of abortion stigma, most participants expressed agency over their decision. A number of women, even those who felt strongly that abortion is wrong, still expressed that they made the right decision, suggesting a strong sense of self-reliance.
"Personally I believe that abortion is wrong … But when I decide[d] to abort [I did so] not because it is good, but [because] it may help. Because with the pregnancy I had to meet challenges I was not ready to face." [Respondent aged 21 years, interviewed at site B] Many respondents mentioned not worrying about judgment or feeling bad about their abortion because they felt that it was a private decision, being made between them, their provider and/ or their partner:

Discussion
Consideration of the larger public narrative about abortion in Nairobi has revealed that abortion remains not only widely stigmatised in Kenya but also criminalised due to its poorly understood legal status. 25 In this study, we gain a better understanding of young Kenyan women's experiences with abortion and PAC services at MSK facilities, including the quality of care they received, the barriers to care they faced and ways in which they responded to widespread abortion stigma.
This study suggests that young Kenyan women who access abortion and PAC services face significant psychosocial barriers. They are confronted by considerable fear of stigmatisation, based in negative religious and cultural beliefs held by their families, communities, and even within themselves. Yet, despite felt and internalised abortion stigma, participants were able to seek out the services they need. This suggests a level of agency and empowerment, which transcends age, marital status and other demographic characteristics. Participants made it overwhelmingly clear that when they were faced with a pregnancy they did not want, or they were not ready for, they chose to prioritise their own self-fulfilment (i.e. physical, financial, mental and emotional well-being) over the beliefs and desires of others. 26 This held true for both participants who felt that abortion was morally wrong and those who rejected negative beliefs about abortion. Both groups rationalised their abortion decision-making as legitimate behaviours, despite the taboo, because they saw the alternative as unfair to themselves and unfair to their potential future children. In this way, participants were willing to face potential gossip, judgment, and stigmatisation from others, in order to reject the idea of unprepared and unwilling motherhood. 26 These findings confirm prior literature that women who have abortions in the presence of perceived stigma develop individual stigma management strategies in order to promote their own positive coping. 7,26 Common stigma management strategies in this study included secrecy (i.e. limiting disclosure about their decision to have an abortion), and learning how to effectively deal with gossip and judgment from others. However, most prior studies which discuss the empowerment and positive coping of women after an abortion have been cited in the United States and Western Europe, not in an environment such as Kenya, where there is still considerable legal ambiguity, and where abortion is largely stigmatised. [27][28][29][30] Collectively, these studies recognise that among the risk factors for negative coping after an abortion are important interpersonal and contextual vulnerabilities, such as lack of social support, societal stigma, and negative cultural attitudes towards abortion. [27][28][29] As such, one would anticipate that the young Kenyan women interviewed in this study would be more prone towards expressing negative or conflicting emotions after having their abortions, as opposed to the decisional autonomy and self-confidence they displayed. As a result, in this study, we find evidence that even in restrictive (legally and culturally/religiously) contexts, women are able to adopt positive coping behaviours and feel empowered about their abortion decision-making.
Additionally, it is possible that since many of the women in this study heard about MSK from a friend who had sought services there, this created a form of social support that helped women feel more empowered about their decision. Social networks are key for information, but also just the knowledge that other women have had to seek these services could be boosting women's ability to feel positive about their abortions. A larger study that included the voices of women who were not able to obtain the abortions they wanted, or sought care elsewhere and had more negative experiences, would help us understand if social support played a role in women overcoming their own internal and societal stigma to seek the care they wanted.

Limitations
This study had several limitations. First, all women interviewed in this study had successfully reached an MSK facility in time to receive an abortion. Therefore, we do not have insights from women who wished to seek PAC services and faced insurmountable barriers. Thus, the sense of agency and confidence described by our respondents about their decision might be reflective of women who already had more agency and were thus better able to obtain the services they desired. Research interviewing women not able to receive abortion services would further elucidate barriers to care. The women in this study were able to obtain PAC services while facing strong family, community and religious stigma, so these barriers can be overcome to some extent. Research comparing women who received and did not receive abortions could help identify factors enabling women to overcome these barriers.
This study has other limitations related to selection bias. The women who participated in our study may have been more inclined to be interviewed because they had positive experiences with MSK facility staff or because they had not yet experienced pain as a result of the abortion procedure. 31 Selection bias may also have occurred in the sense that participants who received the initial abortion procedure were more predisposed to participate than their counterparts who returned for a PAC follow-up visit. In addition, MSK healthcare clinics are not representative of the diverse range of healthcare centres in Nairobi. As MSK provides high quality, private abortion services, its clients are likely to be slightly older, more financially stable, unmarried, and with less severe symptoms than clients who attend public healthcare facilities. Thus, we cannot generalise these study findings to young Kenyan women more broadly, or even young Kenyan women in urban Nairobi. The experiences of women who receive abortion and PAC services at public healthcare facilities in Nairobi and elsewhere are likely to differ from MSK clients.

Conclusion
Despite widespread abortion stigma and the lack of clarity around legal status, young Kenyan women in this study were able to assert their own healthcare needs over the perceptions and beliefs of others. These women were positive and empowered in their decision to reject unwanted or early motherhood, justifying their decision by situating it relative to their other priorities in life. They saw the alternative of continuing with the pregnancy as unfair to themselves and their potential future children, and as a result were willing to deal with gossip, judgment, and community-level stigma in order to defend their right to choose. 26 Although wide-scale social norming is still needed to begin destigmatising abortion throughout Kenyan society as a whole, these findings provide hope that a subset of young women are able to obtain services and feel empowered about their decision despite existing stigma.
In addition, our finding that most women in this study sought services at MSK clinics because they heard positive things from a friend, suggests that finding ways to leverage social networks, or even create social networks centred around disclosure, may be good intervention approaches. These interventions could help other women obtain induced or PAC services by connecting women who successfully overcame stigma to seek services, with women hoping to obtain services. Since many women wishing to seek an abortion, especially those facing high internalised or felt stigma, might not feel comfortable asking their friends, a social media approach has potential. Most young women, especially in this urban setting, have access to mobile phones. This approach will only be appropriate for populations similar to our study population who have access to mobile phones and are highly literate; different approaches would be needed to provide information and support to women in other parts in Kenya, especially rural areas. More research is needed to understand the specific experiences, barriers, and avenues for support for these other populations.