"When it comes to time of removal, nothing is straightforward": A qualitative study of experiences with barriers to removal of long-acting reversible contraception in Western Kenya.

OBJECTIVE: Barriers to removal of long-acting reversible contraception (LARC) threaten reproductive self-determination, but their influence on contraceptive behaviors is not well understood. We describe perspectives of women in Western Kenya concerning LARC removal barriers. STUDY
DESIGN: We used a qualitative descriptive approach with conventional content analysis to analyze transcripts for content and themes from eight focus group discussions (n = 55 participants) and one client journey mapping workshop (n = 9 participants) with women ages 18-49 in Western Kenya who were currently using or had formerly used contraceptives.
FINDINGS: Our primary themes concerned women's experience of LARC removal barriers and the impact on their behaviors and attitudes towards contraception. Women described providers being unwilling to remove LARC, regardless of rationale (including expiration, seeking pregnancy, or experiencing intolerable side effects) or demanding unaffordable fees. Women were reluctant to try LARC for fear of having to use the method for its entire lifespan even if they did not like it. Women saw LARC removal barriers as increasing their risk of unintended pregnancy through non-replacement of expired devices and fostering distrust in the health system.
CONCLUSION: Barriers to LARC removal may discourage utilization of LARC and contraceptive services generally, which can undermine women's efforts to achieve reproductive self-determination. IMPLICATIONS: Our findings affirm the importance of timely LARC removal to ensure that family planning programs uphold women's reproductive autonomy.

Introduction

At the 2012 London Summit on Family Planning, the Kenyan government made ambitious commitments to increase access to modern contraception in support of reproductive health and rights. Since the London Summit, the prevalence of modern contraceptives has increased from 39% to 60% among married women and 45% to 54% among unmarried women [1]. Implant use increased dramatically from 5% to 30% of the total method mix among married women and 3% to 15% among sexually active unmarried women [1], [2], [3]. Implants and intrauterine devices (IUDs) are forms of long-acting reversible contraception (LARC), the most effective, non-permanent modern contraceptive methods. Both methods require access to a health provider trained in insertion and removal.

The accessibility of LARC removal is a key element of a high-quality, woman-centered, and rights-based approach to care [4,5]. The Kenya Health Information System (KHIS) reports that most LARC insertions occur at public facilities, and more than 99% of facilities that offer insertions also have staff trained to perform removals [3,6]. Yet, Performance Monitoring for Action (PMA2020) reports that 9% of removal attempts in the last 12 months were not successfully completed, with higher rates in rural areas [3]. Technical capacity may not be the only barrier to LARC removal. Despite political will to support the infrastructure to ensure LARC removals, women have mentioned provider refusal throughout sub-Saharan Africa in recent qualitative studies about LARC [7], [8], [9]. Our objective here is to characterize, in greater depth, women's perceptions of LARC removal in Western Kenya, and how their experiences influence their contraceptive use.

Material and methods

This is a secondary analysis of a subset of data collected in a larger mixed-methods parent study designed to identify and contextualize facility-level barriers to family planning uptake and continuation in Western Kenya. We use qualitative data collected between October 2018 and February 2019. The University of North Carolina at Chapel Hill and the Kenya Medical Research Institute institutional review boards approved all procedures.

Focus group discussions

We conducted eight focus group discussions (FGDs), each containing six to eight women. Women were eligible if they had ever used contraception and were between the ages of 18 and 49. We conducted FGDs in four counties in Western Kenya (Kisumu, Kisii, Busia, and Bungoma) in order to be inclusive of the four main tribes that reside in Western Kenya and enhance transferability of our findings [10]. We stratified FGDs by county, contraceptive status (currently uses versus formerly used), and urbanicity (urban versus rural residence).

To recruit, local community health volunteers approached eligible women in the community to obtain permission for a call by study personnel. On the call, study personnel described the study, explained why the woman had been contacted, explained how the results would be used, and determined interest. Women who chose to participate in FGDs or the client journey map workshop provided oral and written informed consent. We provided a travel allowance of approximately six US dollars.

Our in-country research partner, Innovations for Poverty Action-Kenya (IPA-K), hired and trained local female data collectors with experience in FGD facilitation. Each facilitator conducted the FGDs in private locations considered neutral by community partners. Facilitators were multilingual, used women's preferred languages (two FGDs were primarily in Luo, one in Ekegusii, and five in Kiswahili), and interpreted to help the women understand each other, if needed. To achieve the aims of the parent study, the facilitators used a semi-structured guide of 18 predetermined open-ended questions to explore facility-level barriers women faced when trying to access contraception, such as “What have your experiences been like when accessing family planning?” A local female trained note-taker took detailed field notes during the FGDs and managed the audio recording devices. A Kenyan firm transcribed and translated the audio recordings into English.

Our methodology was qualitative description, a pragmatic data-near approach that is suitable for generating policy-relevant insight into health services [11,12]. A female American doctoral-level nurse researcher (LB) and a female Kenyan public health professional (LO) performed coding. Coders read transcripts holistically to obtain a sense of their entirety, writing memos about emergent codes and themes. Using conventional content analysis [13], coders then read transcripts word-for-word to derive codes inductively, which they organized into categories and used to identify themes. Coders produced a codebook with definitions, exemplars, and exclusions to support consistent coding. Coders conducted daily Skype sessions to discuss the cultural context and talk through differences in coding until consensus was reached. Coders managed coding with NVivo 11.0 (QSR International).

We enhanced the dependability of our analysis by maintaining an audit trail, credibility through the) prolonged engagement in the community (11 years) by the study primary investigator (KT), and confirmability through reflexive practice [10].

Client journey maps

We synthesized findings from the FGDs to create a client journey map to identify opportunities to improve client-facing systems [14], [15], [16]. The client journey map visually represents the sequence of women's interactions with the healthcare system when seeking contraception, including the barriers they encounter.

Using the same eligibility criteria and recruitment procedures established for the FGDs, we recruited additional women for a client journey map workshop. An experienced Kenyan researcher fluent in English and Kiswahili facilitated the workshop in Kiswahili in a private meeting room in Kisumu. A note taker and the study primary investigator (KT) also attended. During the workshop, women sat in a semi-circle around the map printed on 3-foot by 6-foot cloth and hung on the wall. The facilitator asked women to discuss which barriers were the most frequent, impactful, and important to address. We used the client journey map workshop to conduct member checks to enhance credibility [10] and to stimulate more discussion. We audio-recorded the workshop, which a Kenyan firm transcribed and translated. We analyzed the client journey map data in the same manner used for the FGDs.

Because the parent study was focused on facility-level barriers to obtaining contraceptives, we did not ask specific questions about LARC removals in the interview guide. However, women spontaneously provided rich descriptions of their experiences seeking LARC removal, including the effects on their lives. We present those findings here.

Results

Fifty-five women participated in the eight FGDs (Table 1). We approached 240 women, obtained consent from 88 women, and 55 women participated. Thirty-three of the women who consented to participate canceled the day before or did not arrive on the morning of the activity. Each FGD had six to eight women, ranging in age from 18 to 46. An additional nine women, ages 27–41, participated in the client journey map workshop. The average FGD length was 103 minutes (81–128 minutes), and the client journey mapping workshop was four hours including breaks. In response to questions about experiences accessing contraceptive services, women described LARC removal barriers in six of the eight FGDs as well as the client journey mapping workshop. Here, we present women's experiences of LARC removal barriers and their perceptions of the impact of those barriers.

Table 1

Characteristics of participants in 8 focus group discussions (FGDs) and 1 client journey mapping workshop (CJMW).

Group	County name	Setting	Contraception	Number of participants	Mean age (range)	Percent married	Number of children per participant	Discussed LARC removal
FGD1	Kisumu	Rural	Formerly used	6	28 (19–35)	100%	2–4	Yes
FGD2	Kisumu	Urban	Currently uses	7	29 (20–33)	57%	1–4	Yes
FGD3	Kisii	Rural	Formerly used	8	31 (18–45)	88%	1–4	Yes
FGD4	Kisii	Urban	Currently uses	7	32 (25–40)	71%	1–3	No
FGD5	Bungoma	Rural	Currently uses	7	26 (20–33)	100%	1–4	Yes
FGD6	Bungoma	Urban	Formerly used	7	36 (24–46)	43%	1–7	Yes
FGD7	Busia	Rural	Currently uses	6	29 (22–42)	100%	1–7	No
FGD8	Busia	Urban	Formerly used	7	29 (24–36)	100%	0–5	Yes
CJMW	Kisumu	Urban	Currently uses	9	32 (27–41)	89%	2–4	Yes

Women's experiences of encountering barriers to LARC removal

Women described being frustrated by providers who refused to remove LARC: “If any type of family planning cannot work for you, you should just go there so that it can be removed. There is no discussion.” (FGD6, Participant 4, urban, formerly used contraceptives.)

Women described providers pressuring them to use methods of LARC for the full duration, regardless of side effects or desire for pregnancy. Providers refused women both with and without agreement of male partners.

Maybe you agreed with your husband that you remove and get pregnant or maybe it's affecting you negatively. If you go to explain that to them, as in to them to remove it, you will become enemies… Like they will respond to you rudely. They want it to last a while at least… They want the time that they placed it for to elapse. (FGD5, Participant 4, currently uses contraceptives, rural).

Weight loss, backache, diminished sexual desire, or bleeding were side effects that some women deemed sufficient justification for removal but providers did not. Some providers offered counseling or medication when refusing to remove the device: “After, they gave me this brufen [sic]. I took it, but it did not help I just continued feeling bad." (FGD2, Participant 4, currently uses contraceptives, urban)

Women from multiple FGDs described providers assessing whether the amount of bleeding was a clinical indication for removal:

If you start getting the effects, like you bleed too much, at times you feel dizzy. When you go to tell him, he will just write for you some drugs and you may continue bleeding and worsening… I used to bleed so much until when I went there, they used to tell me to open my mouth and they check my tongue and say that I am lying and that they haven't seen signs of too much blood coming out. (FGD5, Participant 4, currently uses contraceptives, rural).

Cost barriers to LARC removal affected women in every FGD where LARC removal was mentioned, as well as in the client journey map workshop. Fees ranged from 200 KSH to 3000 KSH, with 500 KSH being most commonly reported. While some providers only solicited fees for unexpired devices, others always required fees: “If you don't have the cash, then that method will not be removed, even if the date has reached for its removal.” (FGD1, Participant 4, formerly used contraceptives, rural)

One woman suspected the fees were not sanctioned:

The providers insisted that I must pay some money for a referral. I go to that hospital, bring such amount of money, and yet I did not have that money and yet I'm not feeling okay. So, I experienced discomfort for quite some time, then at long last I found some money at a point where my situation was worsening, and I removed it. Removing that thing in a public hospital is not easy. You must bribe the providers. Unless there is a forum with another organization then at that time it can be removed for free. (Client Journey Map Workshop, Participant 2, currently uses contraceptives, urban)

Some women explained that public awareness campaigns about free contraception led them to expect removals to be free, so they were unprepared to make payments at the time of their appointment. Some women did not think it was appropriate that insertions were free but removals costly:

They do mistreat. Someone might come, for example, in a private hospital, something that they offered her free of charge. When it comes that time that she wants to remove it – maybe she wants to get a baby or maybe it has been bothering her like the way we have been explaining our problems here – so she needs to get rid of it so that she can be free… She is then told that a certain amount of money is needed. She is told a high amount of money. (FGD3, Participant 4, formerly used contraceptives, rural)

Women varied in their price sensitivity. Although fees deterred some women, others were able to obtain sufficient cash to pay for LARC removals. Other women emphasized that the unanticipated requests for fees often necessitated additional trips. A small portion of the women did not remark on provider refusal or provider fees as a barrier to LARC removal; instead, they emphasized the inconvenience:

In the hospital, I think they can change timetable for family planning. You see, they should have scheduled family planning. There are different methods. The hospital should put the timetable on the wall…. They can set aside a date for removal only because all of the days to remove and insert takes time until they go for lunch and come back, and they still have a lot of clients waiting. (FGD5, Participant 3, currently uses contraceptives, rural).

How barriers to LARC removal impact women

Anticipation of LARC removal barriers decreased some women's willingness to try LARC because they believed they would be forced to use it for the entire duration, even if they had negative side effects or wanted to get pregnant.

I later refused because I thought, if a day comes that insert this thing and it happens that I don't have money for the removal – how will it be? (FGD8, Participant 4, formerly used contraceptives, urban)

Women were not certain why providers advocated for LARC use, though they suspected it was because the methods they requested were out of stock. Some women felt that providers pressured them into initiating LARC and then disparaged them when they sought removal of a method they did not really want to begin with.

Participant 6: Such a thing has happened to me, mostly this implant, when you have placed, let's say you had it for three years, but due to the challenges you are facing, you decide it to be removed, so for them to get an opportunity to come and remove it from you, it's not easy task. They can bad mouth you that you are the one who wanted it and agreed, even if there are challenges you need to persevere till the date it was scheduled to be removed. So, it's not easy for them to accept and remove the implant before its scheduled date has reached. So, at that point, you can be told any form of bad words.

Facilitator: Did that deter you from using the desired method of family planning?

Participant 6: Yeah, it discouraged me.

Facilitator: And did you go back to get it removed?

Participant 6: After then, they came and removed it after which that was the end of me and them. (FGD1, Participant 6, formerly used contraceptives, rural)

In one FDG, a woman characterized LARC removal barriers as producing an unwillingness to patronize any health facilities for other effective modern methods, not just the facility where they experienced barriers to LARC removal. Multiple women also asserted that barriers to the removal and replacement of expired LARC put women at risk of unintended pregnancy

Let me say the difficulty we sometimes get, sometimes your appointment date has reached for the device to be removed or to be given an injection, and sometimes you don't have the money. So that forces you to look for the money because you think, if I don't go, I might get pregnant. Now, that is a difficulty we face, sometimes it's your appointment date but you don't have the money. (FGD2, Participant 3, currently uses contraceptives, urban)

Discussion

Women in Western Kenya described barriers to accessing LARC removal, including provider refusals and prohibitive (and perhaps informal) fees. Some women indicated that these concerns and experiences contributed to their reluctance to use LARC, specifically, or contraception generally. Our findings suggest inattention to LARC removal may undermine the goals of the 2012 London Summit on Family Planning to make voluntary family planning services more accessible.

Women reported a range of costs associated with removals, which they characterized as possibly unsanctioned. While some women are able to afford removals, one third of Kenyans live under the international poverty standard of $1.90 USD (at data collection, 100 KSH was equivalent to approximately $1 USD) [17]. Women believed LARC insertion was covered by public free family planning programs but reported uncertainty about whether coverage extended to removal. This uncertainty seems to reflect ambiguity and coverage gaps in existent policy. Ensuring LARC removal is free of user fees both in policy and in practice would increase accessibility. Public service announcements clarifying if and when fees should be collected would improve the navigability of a rapidly evolving policy landscape [18].

Our findings corroborate other studies from sub-Saharan Africa that found women encounter barriers to LARC removal prior to expiration [7], [8], [9]. Women from Ethiopia and Ghana report challenges obtaining removal of an unexpired implant [8,9]. Like the women who participated in our study, Ethiopian women reported that providers urged them to use medications to manage symptoms [8]. Almost half of the women in a Ghanaian study reported that removal cost more if the device was unexpired [9]. Unlike in our study, most Ghanaian women obtained removals if they expressed desire to get pregnant or brought husband approval for removal [9]. We also note that women attributed beliefs about removal barriers both to personal experience and stories, adding to the evidence that rumor may play a significant role in sexual and reproductive health behaviors [19,20].

We note important limitations as the parent study was not designed to examine barriers to LARC removal. Our study lacks provider perspectives on why they may be counseling women against removal or charging fees for removals. Senderowicz argues that refusals to provide LARC removal constitute a structural form of contraception coercion, reflecting drivers ranging from individual providers’ gender ideology to a cascade of pressure on providers originating from national or donor-driven contraceptive uptake targets [7]. Clarifying the reason providers denied removals should guide future intervention strategies. There are systems-level approaches for ensuring provider and a healthcare system capacity to respond to the management of difficult implant removals [5,21]. However, providers may be capable but refuse for other reasons. Recent in-depth interviews with Kenyan providers found some refused to remove implants without extra compensation due to the difficulty of the procedure [22]. Research in other contexts indicates that providers may be reluctant to perform LARC removals because of concerns about unintended pregnancy, believing women will like the method if they can endure initial side effects [23,24]. Hospital administrations exert significant influence on care provision [25] and may be concerned about the cost of supplies involved in removal, such as gloves, and removal kits. Improving access to LARC removal in Western Kenya necessitates understanding the concerns of stakeholders in the healthcare system and supply chain that may exert pressure on providers.

We note additional limitations, including perspectives absent from our sample. We intentionally excluded women who never used contraception because our primary research question addressed experiences with facility-level barriers. Employing CHVs to recruit current and former contraception users likely contributed to the absence of women who are disengaged from the healthcare system. Our study lacked adolescents and women disclosing sex work, who may be particularly sensitive to barriers to LARC removal [26,27]. We did not collect data about women's use of specific methods; contraceptive behaviors after experiencing or hearing about other women experiencing LARC removal barriers; or use of methods that do not require a clinical encounter, such as fertility-awareness based methods or condoms. Implants constitute a much larger share of the national method mix than IUDs, but often in our data, women did not consistently specify which they were discussing; barriers, such as provider skill, may be specific to the method. Another ambiguity is whether women were talking about the public or private sectors; many women went back and forth between the two frequently, rather than being a consistent patron of one or the other. We concluded that we did not reach data saturation concerning LARC removal barriers, let alone LARC removal generally. Characterizing positive experiences obtaining LARC removals might provide insight into healthcare provision that achieves the goals of the London Summit. Despite the many limitations we describe, our study highlights women's perspectives and identifies a poorly understood barrier to reproductive autonomy.

In order to address the LARC removal barriers raised by women in our study, interventions addressing different levels of the health system may be needed. Revising insurance service packages to include LARC removal explicitly, including among women for whom maternity coverage has expired, would address policy neglect that contributes to structural contraceptive coercion. Program implementation guides and clinical practice guidelines could also be revised to make it clear that removals are considered an essential part of service provision, and thus are considered part of the reimbursable bundle of LARC services. Provider-level interventions that address competency, comfort, and attitudes regarding LARC removal may also be warranted, though further research is needed to better understand the drivers of provider refusal in order to ensure interventions are appropriately designed and tailored. Finally, we recommend leveraging the existing data collection infrastructure to track the frequency and location of provider refusals and fees for LARC removal to develop targeted interventions if certain sectors, facilities, or individual providers are found to account for large proportions of unsuccessful attempts at LARC removal. Given disparities across the country [7], disaggregating such data by geography and key sociodemographic characteristics could foreground equity and foster accountability.

Conclusion

It is concerning that women in Western Kenya report struggling to access LARC removal despite such access being recognized as a human right by WHO [4], particularly as barriers to removal are recognized as contributing to the failure of the previous generation of contraceptive implants [28]. To achieve the ambitious international family planning agenda and promote reproductive autonomy, it will be important to detect and address barriers to LARC removal.