HIV risk and pre‐exposure prophylaxis interest among women seeking post‐abortion care in Kenya: a cross‐sectional study

Abstract Introduction Post‐abortion clinics located in regions with high HIV burden may ideal locations to integrate counselling and delivery of HIV pre‐exposure prophylaxis (PrEP), aligning with normative goals for integrated delivery of HIV and reproductive health care. The objective of this study was to gauge the degree to which Kenyan women seeking care for a pregnancy loss, including induced abortion, are at risk for HIV and whether women would welcome an introduction to PrEP prior to discharge from post‐abortion care. Methods We conducted a mixed‐methods study from August 2019 to February 2020 with women ages 15 to 30 recruited sequentially as they were accessing post‐abortion care at public and private facilities in Thika and Kisumu, Kenya. Data collection was through a cross‐sectional survey and laboratory testing for common sexually transmitted infections (N = 200), and in‐depth interviews (N = 30). Descriptive statistics summarize PrEP knowledge and referrals and a multivariable log‐link binomial model estimated correlates of receiving a referral for PrEP. Qualitative data were analysed using inductive and deductive approaches. Results Among 200 HIV‐negative women (median age 21.0, interquartile range 19.0 to 22.0), the prevalence of Chlamydia trachomatis was 18.2% and Neisseria gonorrhoeae was 2.0%. Half of the women scored ≥5 on a validated tool that would correspond to an expected HIV incidence of 9.5% per year. Approximately half (55.8%) of women were familiar with PrEP prior to the study and 33.3% received a referral from study staff to a clinic offering PrEP. In qualitative interviews, women expressed interest in accessing PrEP from the gynaecology ward that provided post‐abortion care but they preferred alternative locations for PrEP refills. Conclusions Kenyan women accessing post‐abortion care have substantial HIV risk and were favourable about the idea of receiving support to initiate PrEP as part of care offered during post‐abortion care. These settings can be integrated into national PrEP programmes as locations providing PrEP referrals and initiation.


| INTRODUCTION
Young women living in settings with a high HIV burden face intersecting risks for unintended pregnancy, HIV and other sexually transmitted infections. To combat this, a priority strategy is to integrate programmes for reproductive health and HIV prevention, yielding provision of comprehensive care that can be delivered efficiently and in a cost-saving manner [1]. Integration of HIV prevention, including delivery of preexposure prophylaxis (PrEP), into existing reproductive health programmes has been demonstrated in clinics for family planning [2][3][4], antenatal and postnatal care [5], and safer conception [6,7] and is facilitated by national guidelines in many settings that promote PrEP use during pregnancy [8]. When women in these settings have been offered PrEP as an additional routine service, PrEP uptake has been moderate overall (22% in the largest study) with the greatest uptake in women whose partners are known to be living with HIV (94%) [2]. Continuation rates at 6-months appear low, mirroring observations of PrEP continuation among young women accessing PrEP as a stand-alone service [9,10].
Clinics providing abortion and post-abortion care are one type of reproductive health service yet they are often overlooked for programmatic innovation due to the vulnerability of patients and the complicated legal situation of abortion. Kenya has the highest abortion rate in the East Africa region, estimated at 48/1000 women [11]. In a 2012 facility-based survey, the peak age for induced abortion was estimated at 20 to 24 years (76 abortions/1000 women) with adolescents aged 15 to 19 years having a frequency of 38 abortions/1000 women [11]. Women living in settings with high HIV burden who access post-abortion care may have a high risk for HIV and other sexually transmitted infections because condom use is low and likely to continue despite the recent unintended pregnancy [12,13] A recent study in Kenya highlighted the frequency of repeated induced abortions among young women and unmet contraceptive need [11,14], highlighting the potential for poor sexual and reproductive health outcomes among these women [14].
Post-abortion clinics serve women with spontaneous and induced abortion who often need emergency care to be medically stabilized followed by treatment for complications and psychosocial and/or behavioural counselling [15]. Induced abortion is illegal in Kenya except to protect a woman's life and health [16]. Thus, these clinics do not initiate abortions but they do provide life-saving care and reduce the risk of maternal mortality. Currently in Kenya, 35% of maternal deaths are attributed to unsafe abortion [14,17]. A routine component of post-abortion care is counselling on contraceptive use and referral to family planning clinics. We postulate that this component of prevention counseling could be expanded to empower women to adopt effective HIV prevention strategies that are within their control, such as oral PrEP [18,19]. In order to explore the potential to integrate PrEP into post-abortion care, the objective of our study was to conduct formative work among young women seeking postabortion care in Kenya and estimate HIV risk and the potential for PrEP to be introduced.

| Study design and participants
We conducted a cross-sectional mixed-methods study in Thika and Kisumu, Kenya among women who were seeking care following a pregnancy loss. From August 2019 to February 2020, women seeking post-abortion services were sequentially recruited by research staff who were contacted by nurses and clinical officers providing post-abortion care at Thika sub-County Hospital, KMET (a Kisumu-based non-governmental organization), and Jaramogi Oginga Odinga Teaching and Referral Hospital, when a potentially eligible women were seen at the facility. Eligibility criteria for these women included accessing care following a pregnancy loss and being 15 to 30 years. Women were excluded if living with HIV or with a social or physical condition that may make participation unsafe as determined by the study investigators. Enrolment was stopped once 200 women were accrued based on estimates of the sample size needed for moderate precision in estimates of STI prevalence.
All enrolled women completed a survey to characterize demographics, reproductive and medical history, sexual behaviour, relationship characteristics, use of alcohol and/or drugs, HIV risk perception and unintended pregnancy, experiences with the recent pregnancy and pregnancy loss, and knowledge of and experience with PrEP. The Awareness, Interest, Desire, Action marketing model was applied to experiences with PrEP to form a PrEP "knowledge-to-use" cascade. Heavy alcohol use was measured with the Rapid Alcohol Problem Screening tool [20] and all other items were measured with questions validated in recent PrEP and contraceptive studies [21,22].
Women provided a blood sample for Treponema pallidum testing (with Rapid Plasma Reagin tests) and urine sample for Neisseria gonorrhoeae and Chlamydia trachomatis testing using Aptima GenProbe (for women enrolled in Thika) or GeneXpert (for women enrolled in Kisumu). Results from STI testing were returned to women via phone or in-person based on participant preference; women with reactive results were given treatment according to national guidelines. If results from routinely conducted rapid HIV testing were available from the post-abortion clinic, they were abstracted from clinic records and if they were not available, rapid HIV antibody testing was conducted by study staff and results were returned to women in real time. After completing the survey, women were counselled about PrEP and all women were offered a referral for PrEP services at a nearby facility of their choice. The woman's preference for a referral/no referral was noted on the survey form. Research staff entered data from quantitative surveys and laboratory testing into a web-based platform (Comm-Care).
A purposive sample of women who completed the survey and had a high risk for HIV was selected for in-depth interviews. Drawing on standards for qualitative research, we aimed to enrol up to 30 women for this sub-sample [23]. To determine HIV risk, an empirically derived score that includes items incorporating women's age, marital status, alcohol use, partner provision of financial support, and whether the partner has additional sex partners was applied to each woman who completed the survey (i.e. the VOICE risk score with a range from zero to eight and we used a cutoff of >5 to denote high risk) [24]. A semi-structured discussion guide was developed to elicit conversation about HIV and pregnancy risk, perceptions about contraception and PrEP, and experiences with care following a pregnancy loss. Qualitative interviews conducted in the participant's preferred language (English, Kiswahili, or Dholuo) were recorded, transcribed and translated into English. If participants did not give permission for recording, detailed notes were taken in their place.

| Analytic methods
Descriptive statistics were used to summarize participant characteristics, experiences with post-abortion care, knowledge about PrEP, willingness to initiate PrEP in conjunction with post-abortion care, and the frequency with which PrEP referrals were given. To identify characteristics more common among women who received a PrEP referral, we calculated relative risks using a log-link binomial model; factors associated with PrEP referral at the level of p < 0.20 were included in a multivariable model. Qualitative interviews were analysed using inductive and deductive approaches [25]. An initial codebook was developed on questions in the guides and five randomly selected transcripts were read and coded in order to identify new codes. A team of 5 analysts (EC, VO, BK, DT, KN) grouped the codes into themes and broader topics and assigned definitions to each. The codebook with topics, themes, and codes was uploaded into Dedoose. During the coding process, new codes were added as they were identified and then applied to all transcripts. The analysts coded the transcripts independently: all coded the first transcript and came together to concur on their codes, then two people each coded the next three transcripts and came together to identify and adjust codes where discrepancies occurred. Finally, once coders were aligned in their understanding of codes, all remaining transcripts were distributed among coders and each was coded by one person. The team met biweekly over Zoom to discuss progress and emerging themes. Dedoose v8.0 (Los Angeles, CA, USA) was used to facilitate the coding and analysis process and SAS 9.4 (Cary, NC, USA) was used for all quantitative analysis.

| Ethics
The protocol was approved by the Scientific and Ethical Review Unit at the Kenya Medical Research Institute and the Human Subjects Division at the University of Washington. All participants completed written informed consent, including women ages 15 to 17 who were considered emancipated minors according to Kenya national research guidelines since they had recently been pregnant. Participants were reimbursed for their time and any expenses incurred (e.g. travel for interviews).

| Participant characteristics
A total of two hundred women were enrolled for this study, including 30 who were included for in-depth interviews ( Table 1). Two women were excluded from analyses because they were discovered to be living with HIV after study enrolment. Women were a median age of 21 (interquartile range [IQR]: 19 to 22), with 13 years of schooling (IQR 11 to 14), and a median of 0 to 1 child living. While 39.4% reported being married (55% of women in Thika and 17.9% of women in Kisumu), 74.2% reported receiving financial or material support from a partner and 69.2% did not earn an income of their own. Applying the VOICE risk scoring tool for HIV risk, the median score was 5 (IQR: 3 to 6) overall which would be commensurate with an expected HIV incidence of 9.5% (95% CI 7.6% to 12.6%) per year [24]. Women reported a median of 2 (IQR 0 to 8) vaginal sex acts during the past month (Thika: 4, IQR 1 to 20) and Kisumu: 2, IQR 0 to 8) and condom use was relatively low with 11.6% reporting that a condom was used during the most recent vaginal sex and 95.7% of women reporting never, rarely, or sometimes using a condom during the past month. Most women had one partner in the past three months and this was a new partner for 9.1% of women. Most women reported that their partner was HIV negative (67.2%) or that they did not know their partner's HIV status (32.3%) and 12.0% of women (6.4% in Thika and 22.9% in Kisumu) reported feeling obligated to have sex in exchange for money or goods. Twelve percent (12.2%) of women reported using alcohol and half of those using alcohol indicated heavy use. The prevalence of Neisseria gonorrhoeae was 2.0%, Chlamydia trachomatis was 18.2%, and Treponema pallidum was 0%; only 1% of women reported being diagnosed with an STI in the past six months. Women included in qualitative interviews were not substantially different from the whole cohort.
Approximately half (51.1%) of women reported using at least one measure to self-manage their abortion, including 28.8% who used misoprostol or another drug and 6.6% who reported their manual vacuum aspiration (MVA) as the measure taken. This question also elicited responses from 5.1% of women who reported a medical situation as being responsible for their pregnancy loss (e.g. an ectopic pregnancy) ( Table 2). When women provided an explanation for wanting to terminate their pregnancy, reasons included wanting to avoid missing opportunities, financial concerns, partners' lack of desire for the pregnancy, and stigma. Three-quarters of women had MVA at the post-abortion clinic (92.1% in Thika and 54.8% in Kisumu) in addition to treatments for infection (50.5%), pain (95.2%) and blood loss (4.5%). Most women desired more children in the future (median children desired=3, IQR 2 to 3) and 46.5% had never used a contraceptive method in their lifetime.
Knowledge about PrEP was moderate with 58.3% of women having heard of PrEP before the day survey data were collected (56.7% in Thika and 60.6% in Kisumu) and 48.0% having knowledge that PrEP is used to prevent HIV (43.4% in Thika and 72.6% in Kisumu, Figure 1). In qualitative interviews, young women described PrEP as a tool for people who have multiple partners, partners whose HIV status is unknown or known to be positive, and even more broadly for all relationships since it is common for men to have multiple partners. In addition to preventing HIV, women described the mental health benefits, such as reducing stress.
"She told me you take PrEP continuously if you have multiple sexual partners and you don't know about their HIV status to prevent you from HIV or you have a partner who is [HIV]-positive so you use PrEP so that the partner doesn't transfer the virus to you. " (22 year old, Thika) "You cannot trust these boyfriends of ours, you know like while I am here I don't know where he is and what he is doing. . .maybe he is with a lady now. Yes, I can even start using it even today. " (20 year old, Thika) "It prevents HIV and reduces stress because if your boyfriend says he won't use condoms, you have already taken it so you won't put yourself at risk." (23 year old, Thika) Nearly half of women (46.4%) agreed that they would want to start using PrEP. All women were offered a referral and 33.3% actually accepted a PrEP referral from the study team. There were site-specific differences with Kisumu generally having a greater frequency of knowledge and interest than Thika. In particular, 4.8% of women in Kisumu and 0.9% of women in Thika had used PrEP prior to their study participation. Women in Kisumu also had higher frequencies of wanting to start and receiving a referral (60.2% and 52.4%) than in Thika (36.3% and 19.3%). Women generally agreed or strongly agreed with statements relaying interest in PrEP (Figure 2). When probing women about how they learned about PrEP, there were some statements reflecting greater penetration of PrEP knowledge and experience into the community, such as statements about parents supporting PrEP use and learning about PrEP from a neighbour. Conversely, however, women also talked about the stigma associated with PrEP use and its relationship to HIV, indicating that these concerns are still very prevalent.
"Most of the people are saying that PrEP is more effective than condoms because during intercourse a condom might burst and that is already a risk and PrEP, is  (11,14) 13 (11,14) 12 (11,14) 13 (11,14) Married "You know even if someone is using, they don't want people to know, they can't be open to tell you that they use PrEP. Maybe they think that if they tell you, you will think that they are HIV infected, the person who has no knowledge of PrEP will say that 'so and so is infected with HIV' because they are always similar with those drugs of HIV. " (21 year old, Kisumu) Among women who did not know about PrEP during qualitative interviews, some recounted feeling too tired or overwhelmed by the experience of seeking postabortion care to remember what they had been told about PrEP by the research team. While women felt that PrEP could be initiated at a post-abortion clinic, they would prefer to seek PrEP refills from a different location, such as a family planning clinic or pharmacy, to avoid seeing people they know or recalling the negative experience of their pregnancy loss.
"I heard about it [PrEP] but I didn't understand about it well because at that time I was at the hospital I was tired and I didn't get anything about it but the study nurse did her work. I would like to know more about it. . .she told me that when I join the program, I will get to know about it. " (21 year old, Thika) "I wouldn't like to go there [MCH or gynecology ward] and then I remember how it [pregnancy loss] happened. . .for me I wouldn't want to go there. " (21 year old, Thika) "They can access PrEP from there (gynae ward) but the problem would be coming back. Some fear since you may bump into someone. " (21 year old, Thika) Numerous characteristics were different between women who were given a PrEP referral and those who were not (Table 3). PrEP referrals were more commonly given to women who were married or cohabiting with a partner, receiving financial or material support from a partner, had a higher score on the VOICE risk score, had fewer vaginal sex acts, and used a condom during the most recent sex act. However, none of these associations were statistically significant in our multivariable model.

| DISCUSSION
To our knowledge, these are the first data to explore knowledge and interest in HIV PrEP among Kenyan women accessing facility-based care following an abortion. We found that these women are good candidates for PrEP based on their desire and potential for being exposed to HIV. Very few women had actual experience with PrEP; many learned about it for the first time from our research staff following their survey and one-third received a referral to a PrEP clinic. In recent family planning and maternal and child health PrEP  programmes in Kenya, approximately 20% of women have received a PrEP referral, and so women seeking post-abortion care may accept a PrEP referral with similar or slightly greater frequency, an estimate that could be used for policy and programme planning [15]. The overall results from our study suggest that post-abortion care and safe abortion clinics, important components of reproductive health programming, can be incorporated into policy guidance and programmes that aim to link and provide direct services for PrEP and other HIV and STI strategies. Women's mental health varies dramatically after an induced abortion often with changes in depression, anxiety, and substance use [26,27]. Women in our study were willing to talk with clinic providers and our study staff about their potential for HIV exposure and to realize different methods for HIV prevention. An important key finding from the qualitative data is that women expressed a strong desire to avoid the postabortion clinic in the future. Thus, women who initiate PrEP through post-abortion clinics would need strong linkage to alternative PrEP programmes, such as integrated family planning clinics, in order to receive refills and accompanying HIV testing [2,5]. In addition, programmes that feature caring and friendly providers whose attitudes are not judgmental are better suited to younger women [28] and community-based PrEP education services could support young women with unanswered questions regarding PrEP and HIV/STI prevention.
We found higher frequencies of PrEP knowledge and interest in women in Kisumu than Thika. HIV incidence in Kisumu is the highest in Kenya and this region has been scaling up PrEP and other prevention interventions very quickly in the past five years [29][30][31]. Thus, it is likely that community awareness of PrEP and normalization for using HIV prevention has penetrated further into the community. This penetration has been cited by women in South Africa for providing social support and normalizing PrEP use and it is likely to foster opportunities for PrEP provision at a wide range of health clinics and facilities, including post-abortion clinics [32].
This formative study was designed to inform a PrEP delivery programme that will be integrated into existing postabortion clinics in Kenya and to begin developing rapport for research to be conducted in these settings. Thus, our study is limited in size, we were not actually providing PrEP to study participants, and we do not have data on whether PrEP referrals were realized by participants. Additionally, our study is based on sequentially recruiting women accessing postabortion care at health facilities, which is somewhat of a convenience sample and misses a large segment of women who induce abortion at home and do not need to seek medical care. Since abortion and seeking post-abortion care is highly stigmatized, there are no data to determine if our sample is representative of all women who seek post-abortion care. Our efforts to talk to pharmacists that provide medical abortion pills did not yield methods to include them in this study but these are an important source to learn from and consider as PrEP providers in future work [33]. A final limitation is that we targeted our in-depth interview sampling to women with high risk as determined through a tool developed and validated with women from South Africa. Since women in our study were Kenyan, the tool may not be fully applicable, although one-fifth of the women in our study were infected with Chlamydia trachomatis or Neisseria gonorrhoeae, and thus likely at substantial risk for HIV.

| CONCLUSIONS
Post-abortion care is an existing service that can be better leveraged to integrate HIV and STI services and other facets of reproductive health (e.g. HPV screening and vaccination, postpartum family planning). An added benefit of integration is the opportunity to reduce the stigma associated with attending a post-abortion clinic since multiple services are available. Women in this study expressed openness to receiving these types of services and engaging with comprehensive services once their immediate needs related to pregnancy loss were met. To advance the delivery of comprehensive sexual and reproductive health services, post-abortion clinics can become a component of national programmes for PrEP and other HIV and STI services. conducted data analysis. RH and NM prepared the first draft of the paper. All authors read and approved the final manuscript.

A C K N O W L E D G E M E N T S
We acknowledge the women who participated in this study for their willingness to share their experiences and opinions with us. This study was funded by the Children's Investment Fund Foundation (R-1808-02992). All authors declare no conflicts of interest.

F U N D I N G
This work was supported by Children's Investment Fund Foundation (R-1808-02992).