Sexual and reproductive health among adolescent girls and young women in Mombasa, Kenya.

This secondary data analysis of a cross-sectional survey conducted in Mombasa, Kenya characterises sexual and reproductive health (SRH) indicators among adolescent girls and young women (AGYW) engaged in casual and transactional sexual relationships as well as sex work. It describes the association between awareness of local HIV programmes and SRH services uptake for AGYW engaged in sex work. Thirty-eight percent of the participants reported a history of pregnancy. Among participants not trying to get pregnant, 27% stated that they were not currently using any form of contraception. Of the participants who had an abortion, 59% were completed under unsafe conditions. For AGYW engaged in sex work, awareness of local HIV prevention programmes was associated with increased STI testing within the last year (29%) as well as at least one HIV test (99%) compared to those who were not aware of local programming (18% and 92%, respectively); however, only 26% of participants engaged in sex work had heard of local HIV prevention programmes. There were no associations between awareness of local HIV programming and rates of dual contraception use, safe abortion, most recent birth attended by a skilled health professional or testing for HIV during pregnancy. Our study found high need for SRH services, particularly, access to contraception and safe abortion. Continued efforts are required to improve access to the full spectrum of SRH interventions, including family planning services and access to safe abortion in addition to HIV prevention to promote health equity.

Introduction

Adolescent girls and young women (AGYW), aged 15–24 years, are disproportionately impacted by HIV globally. In 2015, young women constituted 58% of newly acquired HIV infections and 60% of all young people living with HIV worldwide.[1] At the end of 2013, an estimated 2.1 million adolescents (aged 10–19) in sub-Saharan Africa were living with HIV; HIV-related deaths have not decreased in this population and HIV remains the number one killer of adolescents there.[2,3] In Kenya, AGYW are twice as likely to be living with HIV as males aged 15–24 years[4] and AGYW reporting the exchange of sex for money, gifts, or favours, are at a five times higher odds of contracting HIV than those who do not engage in transactional sex.[5]

In Kenya, there are unmet contraception needs; among sexually active never married young women, the unmet contraceptive need is as high as 74% among those 15–19 years of age and 39% among those 20–24 years of age.[6,7] For AGYW engaging in sex work (SW), dual contraception or the simultaneous use of both condoms and a female-controlled modern non-barrier method, is encouraged to reduce the risk of unwanted pregnancy as well as the transmission of HIV/STIs. In a study of female sex workers (SWs) in Kenya, reported dual contraception was 38%; however, consistency of condom use was found to vary depending on context.[8] While female SWs frequently recognised the importance of dual contraception for HIV/STIs and pregnancy prevention, barriers to consistent use included fear of losing clients, concern that HIV testing was a prerequisite to access, wait times and side effects such as vaginal bleeding or lack of regular menstrual cycle.[9]

Access to effective forms of modern contraception and safe abortion services is the central component to the promotion of SRH rights; yet globally, an estimated 25 million abortions are completed under unsafe conditions.[10] The WHO estimates that unsafe abortion in sub-Saharan Africa accounts for 10% of all maternal deaths and strongly recommends that safe abortion services should be accessible to all women choosing to terminate a pregnancy.[1] Data from Kenya suggest an estimated 86% of women engaged in SW report at least one abortion in their lifetime with 50% reporting more than one.[11] In Kenya, restrictive policies and the criminalisation of the provision of sexual and reproductive services to unmarried adolescents or those below the age of majority, reinforce stigma and prevent young women from seeking these fundamental services.[12] Abortion is legal only when the woman’s life is deemed at risk due to the pregnancy and abortion services are often not affordable.[13,14]

Globally, 24 million women receive inadequate or no antenatal care.[9] In Kenya, antenatal care and the presence of a skilled health professional at birth have increased over the past twenty years; however, maternal and neonatal mortality remain high (510 per 100,000 live births and 22.2 per 1,000 live births, respectively) and significant heterogeneity exists in the quality of health system delivery.[15] AGYW living in the poorest conditions receive fewer essential services during antenatal care and are least likely to have a skilled health professional at their delivery.[6] Data related to maternal health services and HIV testing in pregnancy are scarce for AGYW in Kenya, particularly for those engaging in SW.

Integration of broader SRH interventions for AGYW within existing HIV prevention programmes has been promoted since the 1980s; however, the integration of vertical HIV programming with existing state health systems has been challenging.[10,16-18] The WHO has developed a framework of six recommended heath intervention categories to advance the SRH rights of women living with HIV.[1] Ideally, HIV programmes should work with local health systems to provide sexual health counselling, violence against women services, family planning, STI and cervical cancer services, safe abortion services and antenatal care and maternal health services.[19,20]

The Transitions Study[21] was conducted among AGYW recruited from settings associated with SW in Mombasa, Kenya. The objective of this paper is to describe the SRH indicators among AGYW engaging in casual sex, transactional sex and formal SW and to determine whether awareness of existing HIV programming is associated with SRH service uptake among AGYW engaging in SW.

Methods

We performed a secondary analysis of a cross-sectional survey from The Transitions Study among AGYW aged 14–24 years in Mombasa, Kenya.[21] Questions related to socio-demographic characteristics, risk behaviours, structural vulnerabilities, programme use and STI/HIV testing practices were based on Integrated Behavioural and Biological Assessment (IBBA) Guidelines.[22] Sex partner concurrency was measured using the UNAIDS (Joint United Nations Programme on HIV/AIDS) guidelines, which recommend measuring concurrency as “point-prevalence” six months before the interview, as well as measuring self-reporting of concurrency. Questions related to sexual partnerships and timing of events were based on a number of different validated tools and methodological approaches for capturing key events.[21] Prior to study implementation, the questionnaire was pilot tested in all three study sites.

Data collection occurred between April and November 2015. Participants were recruited from “hotspots”, which can be defined as locations where female SWs meet clients such as bars, nightclubs, hotels and public spaces. Recruitment was conducted through the support of community mobilisers, who were formerly or currently engaged in SW, and who were working with the International Centre for Reproductive Health (ICRH – Kenya). AGYW were eligible to participate if they were 14–24 years of age, sexually active (ever had vaginal or anal intercourse) and could provide written informed consent. Trained interviewers administered the questionnaire in English or Kiswahili, the local language.

Definitions

Participants were classified as engaging in SW if they self-identified as a sex worker or ever reported sex in exchange for money wherein the price was negotiated prior to the sex event. Participants were classified as engaging in transactional sex (TS) if they could not be classified as engaging in SW but reported at least one sex partner in their lifetime where there was an expectation and receipt of money/goods in return for sex, but the price was not pre-negotiated. The remainder of the participants were classified as engaging in casual sex (CS).[21,23] During the interview, participants were asked two criteria questions: (1) Have you ever had sex with a man with the expectation that you would receive money, gifts, goods or other resources in return? and (2) Have you ever had sex with a man where the price of sex was negotiated before the sex event? Participants self-selected into the SW group if they answered “yes” to both questions, and conversely into the CS group if they answered “no” to both questions. If participants answered “yes” to the first question but “no” to the second, they were directed into the TS group. For participants who answered “yes” to both questions, but later indicated they had never had clients or they had never considered themselves sex workers, they were redirected to the TS group.

Modern female-controlled non-barrier methods of contraception included female sterilisation, birth control pills, IUD, injectables and implants. Dual contraception use was defined as at least one modern female-controlled non-barrier method and condoms. Unsafe abortion was defined as abortions completed under unsafe conditions: participant’s home, someone else’s home or unlicensed clinic. Safe abortions were defined as abortions completed in a public/government facility or private, non-governmental, community- or faith-based organisation (NGO/CBO/FBO) facility. Births completed in public/government facilities or private/NGO/CBO/FBO were considered to have been attended by a skilled health professional.

All participants provided written informed consent. We did not seek a waiver of parental consent for participants under the age of 18. The justification was that in this context, participants under the age of 18 fall into a mature minor categorisation. Furthermore, seeking parental consent could potentially result in harm for participants if their sexual behaviours were disclosed to their parents. Ethical approval was obtained from the Human Research Ethics Board at the University of Manitoba in Canada (HS16557 [H2013:295]), the Kenyatta National Hospital/University of Nairobi Ethics and Research Committee (P497/10/2013).

Statistical analysis

We describe the demographic characteristics of participants using frequencies, means and standard deviations. Two-tailed Student’s t-tests and chi-squared tests were used to compare means and proportions of normally distributed data and the Wilcoxon rank-sum test was used to compare outcomes that were not normally distributed. Demographic variables were examined across the three study groups: AGYW engaging in SW, TS or CS. Because of the small sample size of the TS group, behavioural overlap across the CS and TS groups, and that those engaging in SW would be at the highest risk of HIV,[4] the CS and TS groups were combined and compared to the SW group for SRH outcome indicators: age at first sex, ever pregnant, age at first pregnancy, number of live births, most recent birth attended by a skilled health personnel, ever had an abortion, total number of abortions, unsafe abortion and current contraception use. Women who responded “I am trying to get pregnant” (n = 3) were excluded from the analysis on contraception use.

Because existing HIV-targeted intervention prevention programmes in Mombasa were at the time designed to target only women engaged in SW, only the SW group was included for analysis examining the association of awareness of local HIV prevention programming with SRH services uptake. Participants engaging in SW were coded as being aware of HIV programming if they responded “yes” to Are you aware of any NGOs, CBOs, or FBOs working on the prevention of HIV/AIDS for young women AND/OR female sex workers? Because the sub-groups within the SW group were similar demographically, we used univariate logistic regression to examine the association between awareness of HIV programmes and the outcome variables: tested for STI in the last year, ever tested for HIV, dual contraception use (female sterilisation or birth control pills or IUD or injectables or implants and condoms), most recent abortion completed safely (completed in public/government facility or private/NGO/CBO/FBO facility), most recent birth attended by a skilled health professional (completed in public/government facility or private/NGO/CBO/FBO facility) and tested for HIV during most recent pregnancy to calculate odds ratios (ORs) and 95% confidence intervals (CIs). All data were analysed using SAS Version 9.4 software (SAS Institute, Cary, NC).

Results

Of the 1419 women screened, 1304 (91.9%) were eligible and 1299 (99.6%) consented to participate. As shown in Table 1, the average age of participants overall was 19.3 years (SD = 2.6 years), most participants could read and write (97%) and most were unmarried (97%). There was a difference in age across the three study groups, with participants engaged in SW being slightly older (20.0 years, p < .01) than AGYW engaged in CS (18.9 years, p < .01) or TS (19.2 years, p < .01).

Table 1.

Demographic characteristics across AGYW engaged in casual sex, transactional sex and SW in Mombasa, Kenya

Variable	Total (N = 1299)	Casual sex n = 714 (55%)	Transactional Sex n = 177 (14%)	Formal sex Work n = 408 (31%)	P
Age in years M (SD)	19.30 (2.56)	18.94 (2.61)	19.21 (2.38)	19.96 (2.43)	< .01
Literate	1262 (97.07%)	693 (97.06%)	171 (96.61%)	397 (97.03%)	.17
Married	45 (3.46%)	31 (4.34%)	6 (3.39%)	8 (1.96%)	.11
Religion
Roman Catholic	378 (29.01%)
Protestant/Other	601 (46.27%)	178 (24.93%)	63 (35.59%)	137 (33.58%)
Christian	306 (23.56%)	341 (47.76%)	83 (46.89%)	177 (43.38%)	< .01
Muslim	14 (1.08%)	192 (26.89%)	30 (16.95%)	84 (20.59%)
Other religion/		3 (0.42%)	1 (0.56%)	10 (2.45%)
Unanswered

Sexual and reproductive health indicators

As shown in Table 2, participants reported an average age of 16.0 years (SD = 2.4 years) at first sex. Participants engaged in SW were younger at the time of first sex compared to AGYW in the TS and CS groups combined (15.5 years vs. 16.3 years, p < .01). Overall, 38% of participants reported a history of at least one pregnancy. Participants engaged in SW were more likely to report a pregnancy compared to AGYW in the other two groups combined (57% vs. 29%, p < 0.01)

Table 2.

SRH indicators among AGYW engaged in casual and transactional sex or SW in Mombasa, Kenya

	Total (N = 1299)	Casual/transactional sex n = 891	Formal sex work n = 408	P
Age at first sex M (SD) (n = 1216)	16.04 (2.42)	16.27 (2.32)	15.54 (2.54)	< .01
Ever pregnant (n = 1299)	493 (37.95%)	259 (29.07%)	234 (57.35%)	< .01
Age at first pregnancy M (SD) (n = 486)	17.79 (2.25)	17.98 (2.25)	17.57 (2.23)	0.03
Number of live births (n = 493)
None	115 (23.33%)	69 (26.64%)	46 (19.66%)
1	348 (70.59%)	180 (69.50%)	168 (71.79%)	0.03
> 1	30 (6.09%)	10 (3.86%)	20 (8.55%)
Most recent birth attended by a skilled health personnela (n = 378)	319 (84.39%)	164 (86.32%)	155 (82.45%)	0.3
Ever had abortion (n = 490)	105 (21.43%)	49 (19.07%)	56 (24.03%)	0.19
Age at first abortion M (SD) (n = 106)	18.30 (2.36)	18.02 (2.07)	18.55 (2.59)	0.23
Total number of abortions (n = 490)
None	385 (78.57%)	208 (80.93%)	177 (75.97%)
1	86 (17.52%)	46 (17.90%)	40 (17.17%)	< .01
>1	19 (3.87%)	3 (1.17%)	16 (6.87%)
Unsafe abortionb (n = 105)
Participant’s home	27 (25.71%)	11 (22.45%)	16 (28.57%)
In someone else’s home	1 (0.95%)	0	1 (1.79%)	0.5
Unlicensed clinic	34 (32.38%)	15 (30.61%)	19 (33.93%)
Current contraception usec (n = 1299)
Not using any methods to avoid pregnancy	354 (27.25%)	282 (31.65%)	72 (17.65%)	< .01
Modern method	814 (62.66%)	487 (54.66%)	327 (80.15%)	< .01
Birth control pills/IUD/injectables/implants	356 (27.41%)	179 (20.09%)	177 (43.38%)	< .01
Condoms	479 (36.87%)	305 (34.23%)	174 (42.65%)	< .01
Emergency contraception	113 (8.70%)	74 (8.31%)	39 (9.56%)	0.46
Rhythm/withdrawal	173 (13.32%)	147 (13.60%)	26 (6.37%)	< .01

aPublic/government facility or private/NGO/CBO/FBO facility.

bParticipant’s home, someone else’s home or unlicensed clinic.

cAmong participants not intending to get pregnant; multiple answers possible.

Among all participants who had been pregnant (n = 493), the average age at first pregnancy was 17.8 years (SD = 2.3 years). Participants engaged in SW were younger at the age of first pregnancy compared to AGYW in the other two groups combined (17.6 years, SD = 2.2 vs. 18.0, SD = 2.3, p < 0.01). Most participants reported one live birth (71%), with 84% of the most recent births attended by a skilled health professional. Participants engaged in SW were more likely to report more than one live birth compared to AGYW in the other groups (9% vs. 4%; p < 0.01). There was no difference across the groups for the presence of a skilled health personnel at the most recent birth (p = 0.30).

Of the participants who had ever been pregnant (n = 493), most (79%) reported that they had never had an abortion; but among those who had (n = 105), the average age of first abortion was 18.3 years (SD = 2.4 years) and 59% of the most recent abortions were completed under unsafe conditions (participant’s home, someone else’s home or unlicensed clinic). The average age at the time of first abortion was not different between AGYW engaged in SW (18.6 years, SD = 2.6) and those engaged in CS or TS (18.0 years, SD = 2.1; p = 0.23). There was also no difference in whether the most recent abortion was unsafe across groups (64% vs. 53%; p = 0.24). However, SWs were significantly more likely to report multiple abortions as compared to AGYW engaging in other forms of sex (7% vs. 1%; p < 0.01).

Of the participants who were not trying to get pregnant (n = 1296), 27% stated that they were not currently using any contraceptive methods. AGYW engaged in CS or TS were more likely to report not using any methods to prevent pregnancy in the past one month (32%) compared to AGYW engaging in SW (18%; p < 0.01). Overall, among participants who reported using contraception in the past one month (n = 942), the most common form of contraception was condoms (37%) followed by birth control pills/IUD/injectables/implants (27%), rhythm/withdrawal methods (13%) and emergency contraception (9%). Overall, participants engaged in SW were more likely to report the use of modern methods of contraception (80%) compared to AGYW engaging in CS and TS (55%, p < 0.01). Participants engaged in SW were more likely to report the use of birth control pills/IUD/injectables/implants (43%) compared to AGYW engaging in CS and TS (20%, p < 0.01). AGYW engaged in SW were more likely to report the use of condoms (43%) compared to AGYW engaging in CS and TS (34%, p < 0.01). However, dual contraception was uncommon among participants engaged in SW with only 8.9% of participants reporting the use of female sterilisation or birth control pills or IUD or injectables or implants with condoms.

Awareness of local HIV programmes and SRH services uptake among AGYW engaged in SW

As shown in Table 3, 74% (n = 302) of AGYW engaged in SW were unaware of local HIV programming. Awareness of local HIV programmes was associated with a greater odds of STI testing in the last year, with 29% of AGYW engaged in SW who were aware of local HIV programmes being tested for an STI within the last year compared to 18% of women who were unaware of these services (OR 1.80, 95% CI [1.74, 3.00]). Similar patterns were seen with HIV testing, with 99% of AGYW engaged in SW who were aware of local HIV programmes reporting at least one HIV test in their life compared to 92% of women who were unaware of these services (OR 8.49, 95% CI [1.13, 63.61]).

Table 3.

SRH indicators among AGYW engaged in SW in Mombasa, Kenya

	Total	Aware of local programming (n = 105)	Not aware of local programming (n = 302)	Crude OR	95% CI	P
Tested for an STI in the last year (n = 407)	85 (20.88%)	30 (28.57%)	55 (18.21%)	1.80	1.74, 3.00	.03
Ever tested for HIV (n = 406)	382 (94.09%)	103 (99.04%)	279 (92.38%)	8.49	1.13, 63.61	.04
Dual contraception usea (n = 407)	36 (8.85%)	7 (6.67%)	29 (9.60%)	0.67	0.63, 3.50	.36
Safe abortionb (n = 56)	19 (33.93%)	7 (50.00%)	12 (28.57%)	2.50	0.72, 8.67	.15
Most recent birth attended by a skilled health professionalc (n = 188)	155 (82.45%)	43 (87.76%)	112 (80.58%)	1.73	0.22, 1.50	.26
Tested for HIV during most recent pregnancy (n = 234)	180 (76.92%)	48 (80.00%)	132 (75.86%)	1.27	0.62, 2.62	.51

aFemale sterilisation or birth control pills or IUD or Injectables or Implants and Condoms.

bPublic/government facility or private/NGO/CBO/FBO facility.

cPublic/government facility or private/NGO/CBO/FBO facility.

Ninety-four percent of participants engaged in SW reported ever being tested for HIV and 77% reported that they had been tested for HIV during their most recent pregnancy. Awareness of local HIV programming was not associated with improved rates of safe abortion (OR 2.50, 95%CI [0.72, 8.67]), dual contraception (OR 0.67, 95% CI [0.63, 3.50]), most recent birth attended by a skilled health professional (OR 1.73, 95%CI [0.22, 1.50]) or being tested for HIV during the most recent pregnancy (OR 1.27, 95%CI [0.62, 2.62]).

Discussion

This study found that AGYW engaged in SW were younger at the time of first sex, more likely to report a history of pregnancy and had more than one live birth compared to those engaged in CS and TS. In the general Kenyan population, the average age at first sex for females is 18.0 years and this indicator has remained stable over time.[6] In our study, participants reported a younger age at first sex of 16.5 years. Furthermore, AGYW engaged in SW were younger at the time of first sex (15.8 years) compared to AGYW in the TS and CS groups combined (16.7 years). Additionally, from general population surveys, an estimated 18% of AGYW aged 15–19 years have given birth to their first child and the average age at first birth overall is 20.3 years.[6] By contrast, in our study, a significantly higher proportion of AGYW aged 14–24 years reported a history of pregnancy (38%) with an average age at first birth of 17.8 years. AGYW engaged in SW were twice as likely to report a history of pregnancy (57%) compared to AGYW in the TS and CS groups combined (29%). Taken together, our study highlights the young age of first sex and pregnancy among a group of AGYW congregating at hotspots, thus identifying a group of AGYW with unmet SRH needs; the hotspots used in this research could be potentially utilised as locations in which to reach these AGYW.

Our data found that more than one quarter of AGYW were not using any form of contraception, despite not wanting to become pregnant. Qualitative research has offered important insights as to why contraception rates in AGYW engaged in SW in Kenya remain low. Corneli et al. (2016) found that female SWs in Kenya perceive unsupportive clinic infrastructure (long wait times, fees, inconvenient operating hours, perceived compulsory HIV testing), perceived differential treatment from male and female staff and negative partner influences (including paying and nonpaying partners) as barriers to accessing contraceptive services.[24] Male-dominated reproductive decision making, social network approval, fears of side effects, such as infertility, and stigmatisation have also been identified as barriers to contraception use in Kenya.[25-28] Evaluation of more accessible drop-in centres, peer-led outreach work and interventions to reduce criminalisation and stigmatisation directed at public healthcare providers warrant additional study. Education programming targeting male partners could also improve contraception use.[25]

Overall, participants engaged in SW were more likely to report the use of modern methods of contraception (80%) compared to AGYW engaging in CS and TS (55%, p < 0.01). However, dual contraception was uncommon among participants engaged in SW with only 8.9% of participants reporting the use of modern female-controlled contraception with condoms. The use of dual contraception in this population of AGYW engaged in SW was lower than expected on the basis of previous research.[8] Additional research elucidating why various methods are chosen in particular settings, such as regular clients vs. new clients, would enable programmes to facilitate a more tailored approach to encourage consistent protection from HIV/STIs as well as unwanted pregnancy.

While data are limited, previous research has suggested an estimated 86% of women engaged in SW in Kenya had a history of at least one abortion with 50% reporting a history of more than one.[11] In this study population of AGYW, we found 24% already reported at least one abortion by the average age of 19 years and 7% reported more than one. Of the participants who had an abortion (n = 105), 59% were completed under unsafe conditions. There are clear associations between unsafe abortion and maternal mortality; this study highlights high rates of abortion in AGYW engaging in SW and the need for continued effort to integrate services that meet the holistic SRH needs of this population.

Among AGYW engaged in SW, only 26% had heard of local HIV programmes. While safe abortion, dual contraception use, most recent birth attended by a skilled health professional and HIV testing during most recent pregnancy were not associated with awareness of local HIV programming, there was evidence that awareness of existing programmes was associated with STI testing in the last year and a history of at least one HIV test. Importantly, most of the women, who reported having tested, had been tested during pregnancy, demonstrating promising synergy between HIV programming and broader SRH programmes for antenatal care.

This study found high rates of unmet SRH needs, particularly contraception use and access to safe abortion, among AGYW engaged in CS, TS and SW, congregating at hotspots in Mombasa, Kenya. These sites could offer important geographic opportunities for HIV and SHR integration through programme outreach. Additional research could evaluate peer-led efforts, male partner-directed education programming or public media campaigns to address some of the known barriers to SRH service uptake in this setting. Among AGYW engaged in SW, we found a low awareness of local HIV programmes. Additional research is urgently needed to understand how to facilitate awareness of existing programming. Furthermore, integration of SRH services by leveraging existing HIV programmes within the community, particularly family planning and access to safe abortion, is needed to improve the access and uptake of SRH services and promote health equity.

Limitations

Topics, including contraception, pregnancy and abortion, can be sensitive and the quality of the recorded responses is subject to recall biases such as social desirability. As well, as this was a cross-sectional study, we cannot infer causality with respect to awareness of programmes and use/access of receipt of services. Due to small sample size, we did not control for demographic variables; it is likely that higher rates of multiple abortions in the SW group could be partly a function of greater age.