Study design, setting and population

This paper is drawn from a larger ethnographic study conducted in Kilifi County between January and July 2021. The study aimed at understanding girls and young women abortion decision-making process, their care seeking pathways, as well as the different actors involved. The 2014 Kenya Demographic and Health Survey indicates that almost two out of ten girls in the Kilifi between the ages of 15 and 19, are reported to be pregnant for the first time or have had a child. This places Kilifi among the counties with a high prevalence rate of teenage pregnancy (21.8%). One of the reasons for this high teenage pregnancy rate is the low rate of 33% in usage of modern contraception in the county, against a national rate of 53% [18]. Consequently, unsafe abortions are reported to be common in the county, especially among 15-19-year-old girls, with some ending in complications treated in public health facilities. The study was conducted in two of the nine sub-counties: one urban (Kilifi North), and one rural (Kaloleni). In these two sub-counties, we collected data in two primary and two referral level facilities, as well as the neighboring communities. While the main study targeted various profiles of participants (including girls and young women who had abortion, community members, services providers and policy makers), this paper focuses on girls and young women who received post-abortion care following abortion complications as well as health care providers in charge of PAC services provision.

Data collection approach and participants’ characteristics

The data collection involved participant observation in the targeted public health facilities and surrounding communities, as well as in depth interviews with girls and young women who have experienced abortion and health care providers offering post-abortion care. Four junior researchers with backgrounds in anthropology and sociology were taken through face-to-face training for five days. The training included the study objectives and design, human research ethics, ethnographic approach, and interviewing and notes taking techniques. To check the validity and reliability of the study tools and documents, researchers did pretests and role-plays, and the tools refined accordingly.

After the training, the researchers immersed themselves in the targeted facilities and communities for six months. The researchers, first, liaised with the facility in-charges who showed and introduced them to specific wards and individuals in charge of providing PAC services. During each typical day of observation in the facilities, the researchers would assist the staff with entering records, registering patients and helping them to navigate the facility. Whenever a PAC patient was admitted, they would seek informed consent from the patient and follow the treatment procedure. For MVA procedures, they assisted by comforting and encouraging the patient, or holding a flashlight to enable the healthcare provider to clearly see the cervix and perform the procedure. This helped them to observe and document patients’ interactions with providers while engaging them in an informal conversation. Not only did this help them build rapport with service providers but also with the patients.

After observing the care process, researchers then asked consent from the patients (who were between 15 and 24 years) to be followed up and interviewed at a later point. In total, 16 girls and young women agreed and were followed up in their communities for six months. This involved visiting them at their home, accompanying them in their daily activities such as managing their shops, fetching water, or mentoring some girls and their family members. In addition to helping us understand their living conditions and factors that might have contributed to their vulnerability towards unintended pregnancy and unsafe abortion, it helped in building rapport with the participants who then could freely speak about their experiences. In the process of following up on these girls and young women, we also identified, through community health volunteers and youth advocates for reproductive health, five other girls and young women who had had abortions (including PAC), who were also recruited for the study and regularly followed up. This brought the total number of participants to 21. In addition to following them by observing their daily lives and informal conversations, we also conducted repeat in-depth interviews to get deeper understanding of their abortion and post abortion care experiences. The interviews and discussions were conducted in Swahili, and sometimes in English. The researchers also took field notes during these informal conversations with either health care providers or the PAC patients.

The immersion in the primary and public health facilities helped us to identify 12 health care providers offering MVA, and conduct in-depth interviews with them. The providers included seven clinical officers, four nurses and one medical doctor. The interviews and conversation with providers during the observations focused on understanding their experiences providing PAC, their perceptions of these services, and the challenges faced.

Data management and analysis

Formal interviews were audio-recorded after obtaining consent from the participants. Audio files and field notes were anonymized, and stored in a password-protected Google Drive project folder. All researchers installed two-step verification to access the data. The audio files were then transcribed verbatim and translated into English when needed. The data was analyzed using a thematic analysis approach [19] driven by theory and data. The study team and research assistants first went through both the study protocol and a sample of interview transcripts and observation notes to draw the emerging themes and sub-themes, and create the codebook. The majors themes in the codebook for this paper included experiences of pain, supplies and equipment, skills and staff training on PAC, records of PAC cases, informal services provision, values-beliefs-attitudes and services provision, patient-provider interactions (induced versus spontaneous, young versus adult, married versus single). The coding was done by three members of the team using Dedoose software. Inter-coder reliability test was first performed to ensure that every coder has the same understanding of the codes and to capture emerging codes. After that was achieved, the coders proceeded to code independently. Online weekly meetings were scheduled to discuss memos, emerging themes or duplication as well as get clarifications on unclear codes. Findings from interviews with participants and observation notes were presented through verbatim quotes adhering to the protection of the participants’ identities.

Ethical considerations

The study protocol was reviewed by the African Population and Health Research Center (APHRC) internal review committee to ensure scientific soundness (DOR/2020/036). We also sought and obtained ethical approval from the Amref Health Africa’s Ethics and Scientific Review Committee (ESRC) (AMREF-ESRC P909/2020). A research clearance was also obtained from the National Commission for Science Technology and Innovation (NACOSTI) (License No: NACOSTI/P/21/8421). As part of community entry, key stakeholders such as the local administrative and health authorities, County Commissioner, County Directors of Education and Health, health facility administrators, chiefs, assistant chiefs, and village elders were engaged first to seek their support. They were then introduced to the research assistants.

The study participants were consented and confidentiality of collected information was also provided. Given the sensitivity of the issue of abortion, we requested and obtained waiver for written consent for participants who did not want to have their consent documented. Therefore, the consents were either written or oral. The participants were duly notified that the decision to participate or not to participate was voluntary, and they had the freedom to withdraw from the research if they wished to do so without coercion. For participants below 18 years, we did not seek parental consent due to the stigma and covert practice around abortion. In most cases, parents and guardians were not aware of the abortion, hence seeking consent from them would put girls at risk of stigma or abuse. Based on this reasoning, we requested and obtained a waiver of parental consent from the ethics committees.

When the cure hurts

Nearly all the girls and women who were treated using the MVA procedure narrated their experiences as ‘very painful’ as highlighted in the following quote:

I was seeing that I’m going to just die. I was feeling a lot of pain because there is no numbing injection. You feel the thing being pulled. That is, you feel it completely. It is not like giving birth at all. Giving birth is easier.

(18 years old, single, unemployed)

As described in the quotes and the field notes above, we witnessed patients going through such pain and equating it to the feeling of someone “inserting their hands through the vagina forcefully”, or “pulling out something”, hence making it unbearable. Patients who already experienced childbirth found the MVA pain worse than that of delivery. To express their pain, some women cried, shouted, screamed while others tried as much as they could to restrain themselves from shouting or letting out a sound that could alarm people outside the room. They would then moan and fidget silently, close their eyes tightly, let the tears flow down to their hair, bite their teeth together, sweat, sway their hips while squeezing their hands. When the pain became intense, they requested to hold our hands. Through the discussion and interviews, the participants explained that stigma around abortion affected the way they reacted to pain during MVA. Given the stigma surrounding abortion in their communities, some women and girls would avoid seeking post-abortion care in the facilities to avoid providers’ reactions or meeting members of their community in the facilities. Those who were brave enough to seek care explained that they needed to do it discreetly, which included bearing the pain in silence not to draw attention to their case. To echo this, a 32-year-old woman reported that she restrained from shouting for fear of being heard and let her secret out as highlighted in the following quote:

It was painful, but I could not remove the sound…you saw there was a window and the window was open, and there were people out there. If I could be noisy, people would start to ask and wonder what is going on there, and then they would follow up my case. So, I just had to endure the pain.

(32 years old, separated, casual worker)

Aside from stigma, some women and girls felt they had to endure the pain in exchange for PAC services. One of the women mentioned that she knew the provider should provide pain medication but chose not to ask for fear of how the provider might react. Moreover, some patients had to suppress the pain as much as they could because they believed the provider might be mad at them to an extent of not completing the MVA procedure as it happened in some cases where patients could not endure the pain, as highlighted in the following notes:

Mary (pseudonym) is crying in pain whenever the suction tube is inserted “Don’t do that, I don’t want, do it slowly”, she cries. …Mary keeps crying, putting her hand close to the vaginal opening. “Don’t do that, you will get a very bad infection” the doctor says. The doctor tries to insert another tube for the suction, Mary cries “I don’t want any more of this, it is painful. I want to go home, I want to go home!”, as she is trying to close her legs. At this point, the doctor has stopped the vacuum suction. She lies there, Dr. G says, “you can discharge her against doctor’s advice, we can’t force her if she doesn’t want.

(Observation notes, referral level facility, urban setting)

That day, the providers decided not to continue the procedure and termed the patient as “non-compliant”. Facility management and the girl’s mother had to intervene for the procedure to be completed late in the evening.

Unlike these providers, others would stop the procedure, and engage the patients in friendly chat with the aim of drifting their thoughts away from the pain. When the patients exhibited pain, they would tell them “sorry mama, sorry, try to cough, it will end soon”, “be strong, we are almost done”. With such constant engagement, they were able to complete the procedure. In situations where this moral support happened to be insufficient, they would resort to pain medication (when available), or request help from colleagues to hold the patient in order to complete the procedure.

We noticed that whenever they fidgeted, moved their hips, held the doctor’s hands or shouted, they interfered with the procedure, which ended up taking longer, hence prolonging the pain. Such MVA procedures would therefore last up to one hour instead of 15–20 minutes for procedures where the woman bears the pain and stays calm.

Factors driving the pain: Between lack of preparedness and providers attitudes

Nurse B began preparing to perform the MVA and he gathered the MVA kit that was in the sterilized solution. He started assembling it and he asked me to help him as he told me about the different cannulas and he noted that the white and yellow ones were missing. I put on gloves, he asked me to give him a condom and that was the first time I saw him using a condom during an MVA. He placed it inside the sanctioning part and he told me it helps make the work easier because it reduces the friction when opening it and closing it when pumping. He asked me to pass the lubricant, he applied it to the speculum and inserted it in the patient. He then told the patient “I’m sorry this is going to be painful” and then he inserted the cannula…

(Observation notes, referral level facility, urban setting).

These observation notes show that the provider did not use pain medication before starting the process as recommended. Hence, he knew the pain the patient was likely to go through, and went ahead to prepare her to expect it. In addition, the MVA kit was worn out, forcing him to use a condom to ensure the MVA draws properly. Overall, the study pointed out various factors driving the painful MVA procedures that we group into two categories: health system related factors (i.e. malfunctioning MVA kits, lack of training, lack of pain medication drugs) and provider related factors. Altogether, this led to pain medication not being used, or pain medication being used but ending up to be ineffective as illustrated in the field notes.

Lack of preparedness of health facilities to offer PAC

Limitations and strengths of the study

This study is one of the rare in sub-Saharan Africa that explores the lived experiences of the MVA pain during post-abortion care. The ethnographic approach used (through the participant observation for instance), allowed us to capture the emotions, feelings, as they were happening, as well get a deeper understanding of the factors driving the pain. However, as a qualitative study, the findings may not be generalizable to other settings. In addition, given the fact that the focus was only on public health facilities, the study findings do not inform on experiences and practices around pain in private health facilities.