Reach, experience, and acceptability of an abortion self-care intervention in Bolivia: a mixed-methods evaluation

Abstract

In 2018, Ipas Bolivia launched an abortion self-care (ASC) community intervention with the goal of increasing access to supportive, well-informed abortion support provided by community agents (CAs). Between September 2019 and July 2020, Ipas conducted a mixed-methods evaluation to assess the reach, outcomes, and acceptability of the intervention. We used logbook data maintained by CAs to capture demographic characteristics and ASC outcomes of people supported. We also conducted in-depth interviews with 25 women who had received support and 22 CAs who had provided support. 530 people accessed ASC support through the intervention, most of whom were young, single, educated women accessing abortion in the first trimester. Among the 302 people who self-managed their abortions, 99% reported having a successful abortion. No women reported adverse events. All women interviewed expressed satisfaction with the support provided by the CA and, in particular, with the information, lack of judgement, and respect they felt from CAs. CAs spoke highly about their experience and viewed their participation as a way to increase people’s ability to exercise their reproductive rights. Obstacles included experiences of stigma, fears of legal repercussions, and difficulties dispelling misconceptions around abortion. Legal restrictions and abortion stigma continue to complicate access to safe abortion, and findings from this evaluation highlight important avenues for the effectiveness and expansion of ASC interventions, including legal support to people who have abortions and those who provide abortion support, building capacity of people as informed buyers, and ensuring that interventions reach rural and other often under-served people.

Introduction

Throughout the world and throughout history, people have always found ways to prevent and end unwanted pregnancies. What has changed in recent decades is the availability of safe and effective pills that allow for many modalities of care, including abortion self-care (ASC) or self-managed abortion. As with other self-care interventions,^1,2 self-managed abortion shifts the power and control of health care to people themselves. The self-care practice of self-managed abortion allows people to exercise their right to sexual and reproductive health and bodily autonomy by allowing individual control over their abortion process.³

People may need to or opt to self-manage their abortions for several reasons, including lack of access or preference for care outside of a medical setting. Lack of access to care within medical settings may be due to legal restrictions, distance, cost, and required time away from home, among other reasons.^4–6 These barriers to abortion access have been exacerbated by the COVID-19 pandemic as in-clinic abortion services were limited and deprioritised.⁷ Additionally, evidence suggests that people may prefer self-managing their own abortions because of increased privacy, autonomy, flexibility, and confidentiality, or because of perceived or experienced stigma and mistreatment within medical settings.^8–12 The benefit of ASC remains especially pronounced in legally restrictive settings where it is often one of the few pathways to safely have an abortion. Self-managed abortion using misoprostol (sometimes in combination with mifepristone) has been proven to be a safe, highly effective abortion method.^3,13,14 The discovery of misoprostol as an abortifacient and the advent of medications for abortion effectively gave people a safe and reliable way to have abortions outside of the healthcare system and provided many living in legally restrictive settings one of few options for safe abortion care. To ensure that people self-managing their abortions had access to information and support, feminist collectives created and championed a model of care known as abortion accompaniment.¹⁵ Abortion accompaniment models exist around the world – in Argentina, Mexico, Ecuador, Chile, Peru, Nigeria, and Venezuela, to name a few.^15–19 Accompaniment is generally provided by trained volunteer activists, without a clinical background, who provide a combination of information, emotional, and physical support, and sometimes also provide the medication and/or referrals to clinical care if desired or needed. In Bolivia, abortion is permissible in the case of rape, incest, or to protect the woman’s health.²⁰ However, studies have shown that abortion is difficult to access even when these conditions are met,^21,22 resulting in a high number of out-of-clinic abortions in the country, and a high rate of maternal mortality when methods used are unsafe.^23,24 This, along with the availability of medications for abortion, made Bolivia an ideal setting for the implementation of an intervention to support best outcomes when self-managed medical abortion (MA) is chosen by people who have abortions. A study conducted by Ipas Bolivia in 2018 found that many Bolivian women and adolescents who search for abortion-related information or services lack precise information.²⁵ Another 2010 study revealed high levels of confusion about MA, low success rates, and difficulty in accessing help for an abortion.²¹ In addition, evidence has shown that people who self-manage abortion may still desire access to health professionals and some have persistent concerns regarding the legal risks involved.^26–28 Taken together, there was a clear need for an abortion self-care (ASC) intervention in Bolivia inspired by the model of abortion accompaniment that combines accurate, accessible information with non-judgmental support and, if needed, timely clinical and legal assistance.

To respond to this need, Ipas Bolivia began the development of an ASC intervention in 2018. This intervention uses as a blueprint the extensive work of abortion accompaniment groups and aims to provide support and information to people having abortions through volunteer community agents (CAs) trained by Ipas. Support as operationalised in this intervention includes close person-centred assistance through compassionate, non-judgmental emotional and physical support to people having abortions throughout the abortion process. This intervention, having been developed by an international NGO (iNGO) and not by grassroots feminist activists, sets it apart from abortion accompaniment models. Rather, this novel intervention can be situated among abortion accompaniment and other non-clinical models of care (telehealth, harm reduction) as another prototype for provision of safe abortion care in legally restrictive settings.

Between June 2019 and July 2020, we conducted a mixed-methods evaluation to assess the reach of this novel intervention and the experiences and outcomes of women* accessing abortion information and support. We also evaluated the acceptability of the intervention among CAs. While studies exist that examine the clinical outcomes and client satisfaction of non-clinical models such as accompaniment, fewer studies assess whether these models are acceptable and sustainable by those providing the accompaniment or support, in addition to the people served by these models.^15,16,18,29 The perspectives of people providing abortion support are particularly important in legally restrictive settings, where participation could result in experienced or perceived stigma or legal repercussions. In addition, there is a stated need for further evidence on the social and emotional experiences of people self-managing their abortions.⁸ Our study aims to fill these gaps.

Methods

In January 2019, Ipas Bolivia launched an ASC intervention using trained community agents to provide accurate information on MA and support to people opting for self-managed MA in two departments in Bolivia. After a successful initial pilot, the ASC intervention was expanded to two additional departments in September 2019. The ASC intervention was implemented in 12 municipalities within these four departments.

Community agents (CAs) form the key pillar of the intervention. CAs are volunteers trained by Ipas to provide information and support to people needing MA, MA-related information, or information on contraception. CAs generally are not clinicians or health providers, but leaders and members of community organisations active in their communities and who, through Ipas training and support, are seen as community resource persons in reproductive health issues, including information on contraception and the abortion law in Bolivia, and in the provision of non-judgmental, empathetic support. They are not the decision makers or providers nor gatekeepers to access, but rather, there to support people in accessing resources they want or need, using evidence-based information.

Ipas Bolivia selected CAs from among a pool of leaders and members of community organisations that had already been trained by Ipas Bolivia in a range of sexual and reproductive health topics between 2016 and 2019. Ipas Bolivia conducted trainings for selected CAs between February 2019 and April 2020 and trained a total of 79 CAs. Characteristics of the 79 CAs are presented in Table 1 and showcase a diversity of age, gender, education, and occupation.

Table 1.

Characteristics of CAs, September 2019–July 2020 (n = 79)

	Total n = 79a
	n	%
Gender
Female	64	81%
Male	12	15%
Non binary	3	4%
Age (n = 76)
19 and under	0	0%
20–29	31	41%
30–39	18	24%
40–49	16	21%
50+	10	13%
Highest Education Level (n = 78)
Primary school	1	1%
Secondary school	11	14%
Technical school	17	22%
University or higher	49	63%
Occupation (n = 77)
Teacher/professor	4	5%
Healthcare worker (pharmacists, psychologists, social workers)	5	6%
Student	24	31%
Other professional (consultants, administrative professionals, public officials, managers)	40	52%
Homemaker	4	5%
Department/Municipality
Department A	20	25%
Department B	22	28%
Department C	26	33%
Department D	11	14%

^aWhere data are missing, total n is noted.

Selected CAs had shown commitment to sexual and reproductive rights and abortion access through continued work with Ipas, and expressed interest and willingness to strengthen their knowledge on abortion, specifically ASC, through participation in the ASC intervention. These selected CAs then participated in a three-part training series to build their capacity to provide information and support to people interested in self-managing their abortions using MA. The trainings covered the following ASC-related topics: how people who opt for self-care should use MA, including dosing information, how to counsel people opting for self-managed MA, how to assess eligibility to use MA, how to provide information to survivors of sexual violence, how to support someone who has taken MA, and how to refer people to providers in the case of pregnancies greater than 12 weeks gestation, any clinical complications, and for contraception if desired.

In addition to training, Ipas staff provided support and follow-up to CAs as needed, including answering any questions they might have, reinforcing training content, and providing additional information and resources. Additionally, Ipas Bolivia conducted skills assessments following each training, which included role-playing counselling and ensuring the correct information was provided based on the individual needs and situations of people having abortions.

Prior to training on ASC, CAs routinely disseminated information about prevention of sexual violence, the abortion law in Bolivia and contraceptive methods through social media, one-on-one conversations with community members and in-person presentations or discussions during local community events. After training on ASC, CAs integrated more abortion-related content and messaging in their routine outreach and dissemination activities. These outreach and dissemination activities served to increase general awareness within the CA’s local community of their extensive knowledge about SRHR issues including abortion, and to build trust and confidence in approaching CAs with any questions about abortion, including how to self-manage an abortion. Ipas also supported and fostered new partnerships between CAs and other local community groups that worked with or supported marginalised, under-represented populations, such as organisations that support sexual violence survivors and sex workers. By fostering these partnerships, these organisations were made aware of Ipas’s ASC intervention and were able to refer anyone who wanted or needed support to CAs in their community.

CAs were available to help women confirm their pregnancy using urine tests from pharmacies and assess their gestational age using a variety of gestational dating resources. CAs were trained on how to use a traditional calendar, as well as physical and online or app-based gestational or pregnancy wheels. Oftentimes, CAs were approached by women who already confirmed pregnancy in a clinic or hospital with an ultrasound. In these instances, the ultrasound information was shared with the CA and used to determine gestational age. Where needed, CAs provided MA information, including correct dosage, side effects, warning signs of potential complications for which the women should seek immediate care, and information on post-abortion contraception. In line with clinical recommendations,³⁰ CAs advised women to take 800 mcg of misoprostol every three hours until expulsion (misoprostol only regimen) or 200 mcg of mifepristone followed by 800 mcg of misoprostol 24 hours later and every subsequent three hours until expulsion (combined regimen). The sublingual route of administration was recommended. Misoprostol is available by prescription from pharmacists and doctors in Bolivia, and sometimes through informal sellers who sell misoprostol of varying quality and price. Mifepristone was added to the essential drug list in Bolivia in 2019 and is available from doctors in limited health facilities and can be more difficult to access. CAs provided people self-managing their abortions with information on trusted misoprostol or mifepristone suppliers and provided in-person, phone, or hybrid in-person and phone support throughout the abortion process. In the case of incomplete abortions, clinical complications, or if the person desired contraception not available through pharmacies, CAs referred women to a facility or Ipas-trained provider. For women with gestational ages >12 weeks, CAs referred clients to appropriate providers for the abortion care. CAs relied on the support of a network of trusted doctors and pharmacists trained in accessing MA safely as recommended by the WHO guidelines,³¹ including how to counsel someone on using MA, how to assess eligibility to use MA for an abortion, dosing information and, in the case of doctors, how to address abortion complications from MA. Ipas also engaged attorneys and SRHR advocates, especially ones familiar with abortion and human rights laws, to support, address any questions, and provide legal assistance if needed or desired.

Between September 2019 and July 2020, Ipas conducted a mixed-methods evaluation to respond to three key questions: (1) who was reached by the ASC intervention in Bolivia? (2) what were the experiences and outcomes of women self-managing their abortions through the ASC intervention? and (3) was the intervention acceptable to the CAs providing support?

To assess the reach of the intervention, we conducted a retrospective review of ASC logbooks maintained by CAs. With women’s consent, CAs maintained confidential logbooks to collect data on ASC support provided and demographic characteristics of people accessing ASC support, including gender, age, marital status, education level, and gestational age. After providing ASC information and support to those requesting it, CAs completed paper logbooks capturing this demographic information. Identifiable information, including name and contact information, was collected for women who consented to follow-up; this information was kept separately and linked to the logbook via an assigned ID number. Ipas consultants collected and consolidated completed logbooks across CAs each month, and the information was entered digitally and analysed at the Ipas Bolivia office.

CA logbooks also captured quantitative experience and outcome data for women accessing information and support through the intervention. During follow-up contacts with women who consented to follow-up, CAs asked women what medications they used (misoprostol, or mifepristone and misoprostol), how many doses of misoprostol they used (integer), whether they experienced any adverse events (yes/no), and whether they received additional care (yes/no). Three weeks after the abortion, women were advised to take urine pregnancy tests and share results with CAs to confirm a successful abortion. Data on successful abortion (yes/no) was then captured in logbooks. During follow-up contacts, CAs confirmed the medication used by each woman by confirming the brand(s) of the medication obtained. Descriptive analyses on demographic and experience and outcome data were conducted using Stata SE 14.

Experience and outcomes of women accessing information and support through the intervention were also assessed through in-depth interviews (IDIs). Ipas Bolivia conducted IDIs with 25 women who utilised ASC services. Gender identity was asked of all ASC participants by the CA and recorded in the CA logbooks. Any cis-gendered woman who self-managed their abortion through the ASC intervention 2–4 weeks prior to recruitment was eligible for participation in IDIs. There were no age limits for eligibility in IDIs. Recruitment was prioritised for women who had self-managed their abortion in the past two weeks, progressing to three weeks and then four weeks, until the number of required participants was recruited. The timeframe allowed for distance from the event to allow the woman to reflect on her experience while also recognising the need to minimise potential bias from loss of recall. CAs contacted women who had previously consented to follow-up and did an initial eligibility phone screening. The CA confirmed that the woman had an abortion in the past 2–4 weeks, informed her of the evaluation, and asked if she was willing to be contacted for an interview. If she agreed, the CA provided her contact information to the Ipas trained interviewer. The interviewer completed the recruitment and informed consent process and scheduled the interview. The interview was done over the phone either immediately after the recruitment and informed consent process or scheduled for another more convenient time. Interviewers confirmed with each participant if they had auditory privacy to respond freely to the interview questions. Interview guides focused on receipt of information and experiences self-managing their abortion, including how women were treated by CAs and their satisfaction with the model. Interviewees were offered a reimbursement of 100 Bolivianos (equivalent to $14 USD) in recognition of their time upon completion of the interview.

Acceptability of the intervention among CAs was assessed through qualitative analysis of IDIs with CAs. We conducted a total of 22 interviews with CAs. Interview guides with CAs focused on obstacles and facilitators to implementing the ASC model, as well as motivations to participate. To capture a diverse sample of CAs across the four intervention departments, Ipas Bolivia intentionally sampled female and male CAs from each of the 12 municipalities that include urban and rural locations. Ipas consultants called all selected CAs to inform them of the study, inquire about their interest in participating, and ask permission to provide their contact information to an Ipas-trained interviewer. All CAs contacted agreed to participate and gave the consultant permission to share their name and phone number with an interviewer. One of three trained interviewers called CAs to provide more information about the interview, perform informed consent and arrange a time to conduct the full interview. Interviewers also confirmed with each participating CA if they had auditory privacy to respond freely to the questions of the interview. Interviews were conducted until at least two CAs were interviewed per municipality, including at least one male CA. CAs interviewed were also offered a 100 Bolivianos (equivalent to $14 USD) reimbursement in recognition of their time upon completion of the interview.

All IDIs were conducted in Spanish either in-person or by phone by a trained interviewer selected and monitored by an external research agency. Ipas staff and consultants did not conduct interviews in an attempt to mitigate the bias inherent in Ipas both implementing and evaluating the intervention. Interviewers were trained in research ethics and trained to assure participants that responses would remain anonymous and confidential, including to Ipas, and that there would be no negative repercussions for declining participation or for sharing any challenges or areas of improvement from their experience in the intervention. Interviews lasted between one and two hours. Interviews were recorded and transcribed verbatim; two Ipas staff members experienced in qualitative data analysis analysed transcripts in Spanish using Dedoose version #8.3.35. The two-person analysis team read all the interviews and developed a codebook based on the structured interview guide. The analysis team double-coded an initial subset of interviews with CAs and women with reconciliation meetings to resolve any discrepancies in coding. Once there was coder agreement and consistency in the use of codes, the remaining transcripts were single-coded by one member of the analysis team. To mitigate bias in the analysis of the CA and IDI transcripts, an independent research consultant experienced in qualitative data analysis subsequently reviewed all transcripts and coding to assess analysis quality and to support analysis and interpretation for this paper. Multilingual members of the research team translated all quotes from Spanish into English, applying meaningful translation rather than literal translation to convey message or meaning more appropriately.

This evaluation received institutional review board approval from the Allendale Investigational Review Board in the United States (Protocol ID: Bolivia_MASU_Mar2020). The research team was unable to identify an appropriate local research ethics committee in Bolivia to review this study.

Results

Reach of the intervention

Between September 2019 and July 2020, 530 people accessed ASC information and support through Ipas’s ASC intervention (Table 2). The large majority (96%) were women, but some men also utilised the model for ASC information. These may include male partners, friends, and family members of women seeking ASC information. A majority of people accessing ASC information and support were less than 29 years old (84%) and were single (70%). Most (62%) had technical school or university-level schooling or higher, and nearly all accessed ASC information and support for a pregnancy in the first trimester.

Table 2.

Characteristics of individuals who accessed ASC services from Community Agents, September 2019–July 2020 (n = 530)

	Total n = 530a
	n	(%)
Gender
Women	510	96%
Men	20	4%
Age
19 and under	124	23%
20–29	324	61%
30–39	72	14%
40–49	10	2%
Relationship status
Married	56	11%
Cohabitating, not married	103	19%
Single	371	70%
Highest educational attended (n = 529)
Primary school	17	3%
Secondary school	182	34%
Technical school or university	330	62%
Gestational age (n = 412)
<13 weeks	410	100%
13 weeks or over	2	0%

^aWhere data are missing, total n is noted.

Experiences of the intervention

Outcomes

Of the 530 people who accessed ASC information through the intervention, most (302; 57%) opted to move ahead with ASC (Table 3). Among them, a majority (63%) reported using misoprostol only, while 37% used misoprostol with mifepristone. Among women who took misoprostol only, the median number of doses (800 mcg) taken was two. Among women who took a combined mifepristone and misoprostol regimen, the median number of doses (200 mcg mifepristone followed by 800 mcg of misoprostol) was one. The vast majority (99%) self-reported having a successful abortion. Two women (1%) self-reported an unsuccessful termination when asked. When probed further, one woman reported she took the pills but did not have an expulsion and was referred to a facility to complete the abortion. Another woman experienced more bleeding than she had expected but when probed by the CA, the amount of bleeding described was consistent with the normal course of termination with medical abortion. She was referred to a private provider for care. No women reported any adverse events, and only six women (2%) reported receiving additional care. Five of these six women sought additional care to confirm their abortion was complete. One woman sought additional care from a private provider because she could not obtain a third dose of misoprostol needed to complete her abortion.

Table 3.

Outcomes of ASC among women who self-managed their abortions (n = 302)

	Total n = 302a
	n	(%)
Medical abortion type used (n = 300)
Misoprostol alone	189	63%
Mifepristone & Misoprostol	111	37%
Self report of successful termination of pregnancy (n = 301)
Yes	299	99%
No	2	1%
Adverse event
No	302	100%
Yes	0	0%
Received additional care (n = 294)
Yes	6	2%
No	288	98%
Median number of doses taken, among misoprostol only group (n = 189)	2
Median number of doses taken, among mife/miso group (n = 111)	1

^aWhere data are missing, total n is noted.

Experiences with the intervention

All women interviewed (n = 25) expressed satisfaction with the ASC intervention and, in particular, with the CAs. This includes women receiving support from female, male, and non-binary CAs. No one expressed discomfort or dissatisfaction having received support from CAs (male or non-binary) who did not self-identify as women. Interviewees highlighted the value of the information, support, advice, and sense of security they felt from CAs, as well as the lack of judgment and respect for their decisions and autonomy as expressed in the following quotes:

“It allowed me to breathe again, knowing that this moment would pass and that it was going to take a big weight off of me  …  he [the CA] treated me well because I could ask him any question or worry I had and he, without any … I don’t know, without treating me poorly he advised me, he told me what I had to do, he responded to me, he always cleared up all my questions. I really have no complaints, truly.” [Woman opting for ASC, Department C]

“[The CA] made me feel calm, and also, that I could trust her  …  Honestly, she took the time to explain everything, any question, she even told me she would be available at any time, and for me, that was great, because sometimes I had questions at dawn, in the morning, very early, and she was available – any question I had, she was very accessible.” [Woman opting for ASC, Department B]

“Above all I felt happy and sure that, if I was going to need help, that I wasn’t alone, that there were people to give you a hand, to support you  …  she [the CA] was a great help, a great community agent  …  even at times when I need psychological support she gives it to me.” [Woman opting for ASC, Department A]

Despite being highly satisfied with the support they received from CAs, it is important to note that expectations for abortion care among women interviewed were quite low, such that simply receiving nonjudgmental care set the model apart:

“[The CA] treated me well because  …  well let’s just say I thought she was going to treat me poorly because of what I was doing.” [Woman opting for ASC, Department C]

“I felt good. Like I said the most important thing is that he didn’t judge me  …  I think someone else may have punished me or looked at me bad, for getting pregnant and wanting to  …  lose it, no?” [Woman opting for ASC, Department A]

Women who self-managed their abortions were also asked about any adverse events or additional care received during or after the abortion process. Most commonly, interviewees mentioned experiencing expected effects of bleeding and cramping. The only side effect reported was nausea and, in line with quantitative results, no interviewees reported any serious adverse events; only two people reported receiving in-clinic care after their abortion, one for a check-up and one for an IUD. In sharing anecdotes about the side effects they experienced, interviewees again highlighted the support and advice provided by CAs in navigating their care, as shown in the following quote:

“No (I didn’t have any problems or complications), I called [the CA] and he  …  he told me that it [my bleeding] was normal, he told me to wait and later if I was bleeding more that I should call him, or go directly to the hospital.” [Woman opting for ASC, Department A]

Finally, interviewees were asked what recommendations they would have for the CA. By and large, these recommendations reflect the satisfaction interviewees reported with the model; most respondents simply recommended that the CAs continue doing the work they are doing, as expressed in the following quotes:

“That they keep providing assistance. Because sometimes when there’s no good information  …  tragedies happen, things that shouldn’t happen, because of the misinformation that people give you, you know? Sometimes when you don’t have the information or support from someone, that’s  …  I think that’s bad.” [Woman opting for ASC, Department A]

“That they keep it up  …  that they continue to be patient and, that they continue inspiring calm.” [Woman opting for ASC, Department D]

Acceptability of the model among CAs accompanying women in self-care

All community agents interviewed (n = 22) spoke very highly of their experience providing ASC information and support to people in their communities. They especially appreciated the support received in training and spoke about the materials (written and pictorial) they were given and the support and continued education they received from Ipas in implementation of the intervention, as expressed in the following quotes:

“The flipcharts helped me a lot, and the little flyers. We don’t have electricity, so the flipcharts were a great help.” [Community Agent, Department B]

“The positives are that they [Ipas] are always training us, no? They’re always finding a way to make sure that we’re protected, that we can do our work without fear  …  whenever we have any questions we can call them [Ipas] at any time and ask, ‘look this is happening, what do I do?’ and they always answer us. For us those two things are very valuable, that they’re always training us, and that we can always count on them to help resolve any problem.” [Community Agent, Department A]

“Any problem I have, any difficulty, I am always in communication with [Ipas]  …  when I have any issue [they] give me all the answers I need.” [Community Agent, Department D]

CAs also spoke extensively about how good they felt in being able to offer support and information to people who otherwise have few safe options to terminate a pregnancy, as demonstrated in the following quotes:

“Well above all it’s to reduce the number of women who are dying, no? …  there’s a high rate of women dying in our society and, that’s what most motivates me … ” [Community Agent, Department B]

“What motivates me is the ability to help women so that they know their rights, because it’s my right to decide whether I want to be a mother or not, no? So no one can decide for me or make me do something I don’t want to do … so for me that’s the motivation, knowing that I’m doing something so women aren’t stuck.” [Community Agent, Department A]

In addition, CAs spoke about the positive effects of the intervention, both on themselves and on the people receiving support. Through their participation in the intervention, CAs were able to form a network of support and assistance, and observed an increase in the knowledge and strength of the people they supported, as expressed in the following quotes:

“We had good coordination as a group. There were times I couldn’t help, and immediately the rest of the group was there [to help].” [Community Agent, Department A]

“The women themselves, some have overcome the fear, the shame of approaching a community agent and asking their question, and receiving orientation.” [Community Agent, Department A]

Obstacles

CAs also identified some obstacles to implementation of the intervention in Bolivia. Some CAs spoke about the stigma they experienced when talking with the public, family, and friends about abortion, which limited their ability to disseminate information about the intervention, as expressed in the following quote:

“The problem for us is fear, no? The stigma of being known as pro-abortion, that we’re promoting abortion. This has always been our main obstacle.” [Community Agent, Department A]

Other obstacles mentioned by the CAs included time limitations due to the fact that all were volunteers, with particular emphasis on the time required to reverse much of the misinformation and misconceptions around abortion that exist in Bolivia. In addition, many CAs mentioned the lack of official office space, job titles, or certifications, and the lack of economic resources to enable them to reach more beneficiaries. The quotes below detail some of these obstacles:

“The travel above all [is a problem]  …  as I’ve said the area there is rural, so getting to the municipios or colonias is very complicated.” [Community Agent, Department B]

“At times I didn’t have [phone] credit … that was an obstacle, because the girls generally use Whatsapp …” [Community Agent, Department C]

“The main obstacles, for me at least  …  here we don’t have an office, and apart from not having an office, we don’t have a certificate that guarantees that you are a person with certification, a person that has training in this field, in sexual and reproductive health.” [Community Agent, Department A]

“Obstacles include the issue of time, and lack of knowledge around the theme [of abortion]. When I say time I mean  …  the community agents need to find time for the different activities, to disseminate the information, to orient people on contraceptive methods, the person [getting an abortion] needs to be sure about their decision, [those conversations] can’t happen out in the open …” [Community Agent, Department B]

“The main problem with working on ASC is the misinformation that exists … it’s an obstacle because these women come to us with a preconceived notion of how the process should be or what they will take and we have to do a whole process of un-learning, you know? To try and teach them the correct way to save their lives in case they want to terminate their pregnancy.” [Community Agent, Department D]

Many CAs also spoke about the importance of establishing trust with the people accessing support through the intervention. This trust ran both ways, given fears of legal repercussions among people accessing information and services and the CAs providing information and support. CAs felt this necessitated the investment of additional time to establish trust with women opting for ASC. The lack of trust between CAs and people desiring ASC information was also believed to result in fewer people seeking support, especially in more rural settings, as reflected in the following quotes:

“I needed to get to know the person [accessing services] better, because we don’t know who they might work with, or what reaction they might have  …  at first the problem I had was that I didn’t know the woman well, I needed to get to know her, I needed to get to the highest level of confidence [with her], which at first I didn’t do but after, with more experience, I was able to orient them much better.” [Community Agent, Department B]

“I think [one of the main obstacles] is fear  …  fear that someone can break confidentiality, respect, privacy  …  because here we all know each other in a sense, so someone could realize ‘oh, this is the cousin of so-and-so, the niece of so-and-so,’ so that’s a big obstacle among us  …  it’s a small town, and people know each other …” [Community Agent, Department A]

“Not all women know that  …  we can help them in these types of situations. We need all women to know that they can come to us, they can trust us, I think that’s the main obstacle  …  maybe we can’t make it such a public theme because we live in a small place and they run the risk of being stigmatized, of people saying, ‘oh no, they’re looking for the people who promote abortion!’ It could create confusion, people may not understand what we’re really trying to do.” [Community Agent, Department D]

Discussion

This paper presents evaluation findings from a novel ASC intervention launched in four departments in Bolivia in 2019. Similar models that employ non-clinical support for medical abortion exist throughout the world in a diverse range of legal settings, including Pakistan, Argentina, Mexico, Peru, Ireland, Tanzania, Nigeria, and Indonesia.^{13,15–19,28,32–35} The value of ASC has become even more pronounced where legal restrictions on abortion proliferate, and as the COVID-19 pandemic led under-staffed, under-resourced health facilities to deprioritise in-clinic abortion services. While full decriminalisation of abortion is essential to ensure access to abortion, as emphasised in the recent WHO abortion guidelines,¹⁴ decriminalisation may take some time to achieve. In the meantime, the ASC intervention and the accompaniment model of care are important extra-clinical models of care that can help address the requirements for safe self-managed abortion as outlined by clinical guidance³⁶ and countries should be encouraged to push for protections for those who choose to self-manage their abortions. The recent US Supreme Court decision to overturn Roe v. Wade lends increased urgency to this call for both domestic and international abortion advocates as anti-abortion rights activists become emboldened.

In its first year of implementation the ASC intervention in Bolivia reached 530 people. These numbers continue to grow, with an additional 1,098 people reached in the second year of implementation. As word spreads about the intervention, we anticipate more people being served each year. Evaluation data shows that women who self-managed their abortions after receiving information and support through the Ipas intervention had high rates of success and no adverse events. These findings contribute to the literature on the safety and effectiveness of non-clinical models of abortion care and are consistent with existing literature on the safety and effectiveness of self-managed abortion using MA with accompaniment and telehealth support.^3,4,13,17,33 In-depth interviews reveal more about the experiences of women self-managing their abortions; ASC users interviewed spoke highly of the model, with particular emphasis on the non-judgmental support of the CAs. The fact that all women interviewed recounted positive experiences with accompaniment, including those supported by male and non-binary CAs, demonstrate that people of all genders can be an important resource in providing abortion information and support and should not be overlooked in such interventions. Overall, findings show that users of Bolivia’s ASC intervention are both interested in and capable of self-managing their abortions safely and had positive experiences with the intervention.

Data from participants in the intervention highlight that the people reached in the first year of implementation were largely young, single, and highly educated. These findings reveal that less educated poorer women were less likely to access support for ASC, despite evidence that their unmet needs for family planning and safe abortion services may be greatest.^37–40 This may be due to most CAs being based in larger cities and their catchment areas skewing urban. In addition, CA difficulty in accessing certain hard-to-reach areas, privacy and legal concerns for people interested in ASC, social stigma and legal restrictions, and a lack of access to phones and internet access which facilitate learning about ASC and accessing support pose additional barriers to reaching rural communities. This is an important avenue of focus for future expansion of the intervention.

Finally, this evaluation is novel in that it assessed whether the ASC intervention was acceptable not just to women accessing support but also to CAs providing support. This information is particularly important in legally restrictive settings where community members who provide information and support may face stigma and legal repercussions for their participation. All CAs interviewed spoke very highly about the intervention. CAs shared a deep appreciation for the training they received from Ipas and viewed their participation as a way to decrease maternal mortality and increase the exercise of reproductive rights in the country. However, CAs vocalised appreciation of and continued need for Ipas support in providing abortion support, especially in the case of less common or serious adverse events. One of the goals of the ASC model is to create a sustainable network of CAs and reproductive rights defenders that can function with minimal external support. This includes ensuring that CAs have the confidence, resources, and motivation to continue providing support. Findings show that CA motivation to participate in the ASC intervention is particularly high. However, the above interest in continued training and Ipas support may signal that one year of implementation is not enough to ensure continuity of ASC intervention services without external support. When coupled with the difficulties CAs shared in travelling to remote areas, having sufficient funds to provide support via phone, and the time required of CAs, there is an expressed need for further investment of resources to ensure CAs are able to continue providing services. In particular, sustainability of interventions like this do require resources, including funds for travel and phone credits as needed, to follow up with people having abortions. Additionally, greater funding is needed to support accompaniment groups and other non-clinical models of care that fill a critical gap in the healthcare landscape, particularly in countries where reproductive rights are limited or under attack.

Our evaluation has limitations. In-depth interviews were only conducted among a subset of CAs and women who sought support through the ASC intervention. Respondents were selected via convenience sampling. Thus, it is possible that CAs and women who did not participate in in-depth interviews are meaningfully different from those who did; CAs and women who have limited connectivity or time, or who had negative experiences with the intervention may have been less willing or able to participate in IDIs and this may explain the overwhelmingly positive feedback about the intervention. Weighting the sample to capture more CAs and women from regions with higher abortion caseloads may have served to obscure findings unique to smaller, more rural areas. However, we felt weighting was necessary not only to capture the most common experiences, but also to protect confidentiality that may be compromised when interviewing CAs and women in settings with lower caseloads. Finally, while non-clinical models of abortion care, including accompaniment models from which this intervention was derived, exist in many other settings, this intervention is distinct from those models and unique to the Bolivian abortion context. As such, findings may not be generalisable to other countries or to departments in Bolivia where this intervention was not implemented.

Additionally, the design and implementation of the intervention and evaluation were funded by Ipas, and authors of this paper are current and former Ipas staff. Ipas’s organisational mission aims to advance reproductive justice by expanding access to abortion; and the authors of this paper also hold the same beliefs and values of advancing reproductive justice. We acknowledge that our views, values, and beliefs have influenced how the evaluation was designed and implemented, and the results presented in this paper. Although we aimed to be as neutral as possible in the collection, interpretation, and presentation of the data – as presented in the methods section – we recognise that this aspiration cannot be fully achieved.

While self-managed abortion using MA is a safe and effective way for people to access abortion even in legally restrictive settings, any discussion of self-managed abortion in such settings necessitates acknowledgement of the myriad ways that legal restrictions continue to hinder abortion access. Ultimately, unhindered access to abortion cannot be realised where legal restrictions and stigma prevail. Stigma and fear of legal repercussions were evident in IDIs with both women and CAs; all service providers, and especially non-clinical models operating outside of the formal health sector, should strongly consider including lawyers as an additional support to people having abortions and those providing services and support. Further, restrictions on abortion drive information and resources underground, leading to the profusion of illicit market abortion medication sellers. While the Ipas intervention attempts to ensure the safety of people self-managing their abortions through trusted pharmacists, trained doctors, and CAs, and diffusion of information in the community, models of care in which people seeking abortions have to purchase the pills themselves at pharmacies or in illicit markets should focus on growing the capacity of these individuals as informed buyers. Resources such as the Ipas Buyers Guide⁴¹ can provide people self-managing abortions with information about how to identify abortion medications and fair costs and minimise price gauging.⁴²

Conclusion

Overall, our findings show that the ASC intervention by CAs in Bolivia was well received both by women accessing ASC and by those supporting them. WHO guidance has highlighted the importance of models like this one to ensure access to safe abortion, especially in settings like Bolivia where people who can get pregnant have limited access to abortion.

This intervention joins a cohort of non-clinical models in other settings that have also proven accessible and acceptable to the people it aims to serve. Our study adds to this evidence base by demonstrating that such interventions are acceptable not only to people accessing services but also to the people providing services, and offers further support to an already extensive roster of literature on the safety and desirability of non-clinical models of abortion care. Nonetheless, legal restrictions and abortion stigma continue to complicate access to safe abortion even where non-clinical models of care such as the ASC intervention and accompaniment models exist. Findings from this evaluation highlight important avenues for the effectiveness and expansion of such interventions, including legal support to CAs and people having abortions, building capacity for people having abortions as informed buyers, and ensuring that interventions reach rural and other often under-served people and settings.

Funding Statement

This study was funded by Ipas.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Ethical approval

Ethical review and approval for this evaluation was obtained from the Allendale Institutional Review Board in the United States (Protocol ID: Bolivia_MASU_Mar2020). The research team was unable to identify an appropriate local research ethics committee in Bolivia to review this study.

Authors’ contributions

MM, AY, and DI led the design and implementation of the accompaniment model. VNA and CA supported evaluation design and data collection. VNA, CA, and SAK led data analysis and interpretation of results. MM, AY, and DI were major contributors to interpreting the data analysis and conceptualising the manuscript content. VNA and SAK led writing and revisions of the manuscript. MM supervised the design and implementation of the ASC intervention, design, and implementation of the evaluation, and manuscript development. All authors read and approved the final manuscript.

Data availability statement

The quantitative and qualitative data that support the findings of this study are not publicly available due to the sensitive nature of the evaluation and risks to breach of confidentiality. The data are available from the corresponding author, VNA, on reasonable request.