Unmet Need for Family Planning and Experience of Unintended Pregnancy Among Female Sex Workers in Urban Cameroon: Results From a National Cross-Sectional Study

Female sex workers (FSWs) in Cameroon have unmet need for effective contraception, and experience of unintended pregnancy and pregnancy termination is common. Reducing barriers to accessing high-quality, voluntary family planning services in FSW-focused community services is a key strategy to promote client-centered care, promote informed choice, reduce unintended pregnancies, and improve quality of life for FSWs.

Africa and was more than 60% across southern Africa. 2 Specific guidelines for HIV interventions among FSWs have shaped programs targeting HIV prevention, testing, and treatment.However, less attention has been given to FSWs' other comprehensive sexual and reproductive health services, including voluntary family planning.
FSWs may consider pregnancy prevention a more influential motivator for condom use than HIV prevention. 3Previous work has demonstrated that unintended pregnancy is a high-priority issue for FSWs, with unintended pregnancy incidence of 27 per 100 person-years among FSWs in lowand middle-income settings without presence of a sexual and reproductive health intervention. 4ne potential outcome of unintended pregnancy is termination.Among FSWs, experience of termination of pregnancy (TOP) is widespread in global estimates and often higher than national estimates among all women. 5In many countries where FSWs are most affected by unintended pregnancy, abortion policy is generally restrictive, 6 which can lead to women seeking unsafe abortions.7][8] In addition, unintended pregnancy leading to live births can have financial implications that may perpetuate vulnerability and HIV risks related to sex work. 9,10ale and female condoms remain important means of preventing unintended pregnancy and HIV and STIs among FSWs in 2019. 113][14][15] These factors include punitive laws and policies, local policing practices, economic insecurity, intersecting stigmas, physical and sexual violence, and drug and alcohol use during sex work. 16Even in settings with high levels of consistent condom use with paying partners, condom use between FSWs and their nonpaying or emotional partners is low due to interpersonal barriers to condom use. 17,18Previous studies have demonstrated associations between having regular, emotional partners and unintended pregnancy or abortion. 19,20n light of the reduced effectiveness of condoms with typical use as opposed to perfect use, 21,22 especially in the context of violence, 16 and differences in condom use with clients and nonpaying partners, dual-method contraceptive use-condoms combined with an effective nonbarrier method-for both HIV/STI and pregnancy prevention is recommended for key cisgender women populations. 23However, widespread studies have shown that FSWs across several lowand middle-income countries have significant unmet need for effective contraception and voluntary family planning services. 5,24The current use of modern nonbarrier methods among FSWs is estimated to be below 40% in numerous settings across sub-Saharan Africa, [25][26][27] and FSWs report most often using injectables or oral contraceptive pills. 14,26,28,29][32] Long-acting reversible contraceptives (LARCs), which include intrauterine devices (IUDs) and implants, are globally recommended for reducing unintended pregnancy due to high effectiveness, continuation rates, and higher tolerability compared to other nonbarrier methods, 31 but availability and utilization of LARCs remain low among FSWs in much of sub-Saharan Africa. 4,33omen in Cameroon most commonly use short-acting and user-dependent contraceptive methods.In 2015, among a household survey of women in urban Yaoundé using any contraceptive method, only 8% were using LARCs and an additional 5% reported using an oral contraceptive pill. 34Perceived efficacy and easy accessibility were the 2 main considerations for contraceptive choice, and only half of women were aware of LARCs.In Cameroon, LARCs are intended to be available through hospitals and selected clinics that provide family planning and reproductive health services. 34,35However, a recent evaluation indicates that stock-outs prevail, with 70%-80% of facilities reporting stock-outs of each offered LARC method on the day of assessment. 36The unmet need for contraception results in high numbers of unintended pregnancies. 36ameroon's abortion policy has been described as restrictive: legally permitted only in cases of rape, incest, or to preserve a woman's physical or mental health, and even then, it can be difficult to obtain legal standing for abortion. 379][40] Due to legal and regulatory restrictions, abortions are often unsafe.Notably, unsafe abortions have been demonstrated to account for one-quarter of registered maternal deaths in Cameroon. 41,42Although generalizable data are limited, abortions have been reported in both home and clinical settings and are most commonly provided by nurses, general practitioners, a friend, or through self-induction.The most commonly reported procedures are dilatation and curettage, manual vacuum aspiration, or misoprostal. 43There is no standard practice for postabortion care in Cameroon, but it may include further manual vacuum aspiration, sharp curettage, or misoprostal depending on resource availability. 37,44It is not clear whether postabortion care services routinely include family planning counseling and methods. 37s of 2016, more than 1 in 4 sex workers in Cameroon are estimated to be living with HIV 12 compared to around 1 in 20 among all women aged 15-64 years. 45Sex work is illegal in Cameroon, and arrests as well as police violence, raids, and extortion have been commonly reported. 46,47Earlier work has demonstrated that physical and sexual violence as well as depression are prevalent among FSWs in Cameroon and associated with condom nonuse, being offered more money for condomless sex, and condom failure. 46,48These circumstances may similarly affect risk of unintended pregnancy.However, there are limited studies evaluating unintended pregnancy, pregnancy outcomes, or related consequences with the majority of studies focused on HIV risks.In response, these analyses aim to fill a gap in knowledge of reproductive health among FSWs in Cameroon by characterizing the prevalence of unintended pregnancy, TOP, and contraceptive use.

METHODS Study Overview and Population
This is a secondary analysis of data collected through a cross-sectional respondent-driven sampling (RDS) study in 5 cities of Cameroon: Bamenda, Bertoua, Douala, Kribi, and Yaoundé.These represent the 2 largest cities and cities with high absolute and relative population of sex workers. 49This study was conducted from December 2015 to October 2016 as a baseline assessment to inform service provision models for the Continuum of prevention, care, and treatment of HIV/AIDS with Most at-risk Populations (CHAMP) program, implemented by an alliance of community-based organizations (CBOs) and led by the nongovernmental organization CARE.
Women were eligible to participate if they were assigned the female sex at birth, reported sex work as their principal source of income in the previous year, were 18 years or older, spoke French or English, and had resided in the city of recruitment for at least 3 months before recruitment.

Study Recruitment and Procedures
Methods have been reported in more detail previously. 12,50Briefly, 6 FSWs were purposively selected as "seeds" based on their peer networks.Seeds and subsequent participants were provided with referral coupons to recruit up to 3 other FSWs from their social network for participation.Sampling continued until recruitment across cities reached the desired sample size (Bamenda: 340, Bertoua: 301, Douala: 460, Kribi: 578, Yaoundé: 571).Sample size calculations were based on the estimated HIV prevalence by region and included a design effect of 2, alpha of 0.01, and were intended to measure HIV prevalence with a precision of 63%.
Trained interviewers assessed participant eligibility, obtained informed verbal consent, and administered a 45-60-minute questionnaire on individual, community, network, and structurallevel HIV risks; sexual and reproductive health; and service engagement.Questionnaires and biological data were linked by a confidential unique identifying code, and data were nonidentifiable.Women diagnosed with HIV were supported for linkage to treatment; others were linked to community-based HIV prevention and support services.All participants were reimbursed 2,000 FCFA (US$4) and received an additional 1,000 FCFA (US$2) for each successful recruit. Ethical

Outcomes
The 2 primary outcomes of interest were lifetime experience of TOP and current use of a nonbarrier contraceptive.Given the complexity of measuring unintended pregnancy and bias associated with retrospective reporting of pregnancy intentions, 51 TOPs was used as an indirect measure of unintended pregnancy as well as an indicator of potential risk to women's health due to unsafe TOP. 8 Respondents were asked the number of lifetime pregnancies experienced followed by whether they had ever experienced the following outcomes: live birth, stillbirth, miscarriage, or voluntary termination.Participants were also asked to select all methods currently being used to prevent pregnancy.Nonbarrier contraceptive use considered current use of an oral contraceptive pill, IUD, injectable method, implant, or female These analyses aim to fill a gap in knowledge of reproductive health among FSWs in Cameroon.
sterilization.Secondary outcomes considered were unintended pregnancy, defined as ever having a self-reported an unplanned or unwanted pregnancy and LARC use.Unintended pregnancy was defined as a pregnancy that was unplanned or unwanted.LARC use included current use of a contraceptive implant or IUD.

Covariates of Interest
The association of multiple covariates with TOP and nonbarrier contraception were considered.Covariates related to service access included previous HIV testing, receipt of information on HIV in the previous 6 months, and member of a CBO working with FSWs.
HIV status was categorized as HIV-negative, HIV-positive previously diagnosed, HIV-positive newly diagnosed, and indeterminate based on serological testing and self-reported HIV status.
Covariates related to other contraceptive use and sexual behavior included ever using emergency contraception, number of clients in the past month, condom use with clients, condom use with regular nonpaying partners, number of regular nonpaying partners in the past year, number of times experienced condom failure in the past year, and whether engaged in sex work during last pregnancy.Consistency of condom use was assessed among participants reporting at least 1 of given partner type on an average week and was assessed separately for regular and casual paying and nonpaying partners.Condom use was defined as consistent if they always reported using condoms during vaginal and anal sex with the given partner type and inconsistent if they reported using condoms less than always.In regression analysis, categorization of condom use with clients (regular and casual nonpaying partners combined) also included not having vaginal or anal sex with clients.In addition, future pregnancy intentions were considered, defined as intending to have children in the future (yes, no, does not know).
Covariates related to structural determinants of health included ever experiencing physical violence or assault, ever being forced to have sex, and experience of health-related stigma.Healthrelated stigma included reporting any of the following in relation to involvement in sex work: ever being afraid of seeking health services, avoidance of health services, mistreatment in a health center, heard health providers gossip, denied health services, or forced to have an HIV test.

Analytical Approach
Data were combined from 5 independent study sites.Primary outcomes are presented as both crude and RDS-adjusted proportions, the latter were calculated by study site and averaged to estimate overall RDS-adjusted proportion.
On account of primary outcomes being common, modified Poisson regression with robust variance was used to approximate prevalence ratios (PR) for each primary outcome. 52,53Models were corrected for RDS-weighting and clustering by seed.Firstly, bivariate associations were assessed for variables with theoretical association to the given outcomes.A multivariable model was developed for each primary outcome with variables with a P value <.1 in bivariate association.Any regular nonpaying sex partner(s) in the past year and consistent condom use with clients were included a priori in both models, and age group was included a priori when modelling associations with history of TOP.To prevent collinearity between HIV testing history and previous HIV diagnosis, only HIV status was considered in multivariable models.
Site-specific RDS sampling weights were computed using the Gile's SS estimator, which generates weight by estimating the probability of being included in the study using self-reported network size, using RDS Analyst (version 0.42, Los Angeles, CA).All other analyses were conducted using Stata version 14 (StataCorp, College Station, TX).Unless indicated, missing data were omitted from analyses.
Among women with any pregnancy experience, 93.1% (1,915/2,058) had any live births, and 32.3% (657/2,035) of women reported any births since beginning to sell sex.At last pregnancy 34.2% (650/1,903) of women concurrently engaged in sex work.
Overall, 64.4% (1,452/2,255) women intended to become pregnant in the future.The most common contraceptive methods reported being currently used were male condoms (76.5%), female condoms (30.9%), and the rhythm method (22.7%); more than 1 response was permissible.

DISCUSSION
FSWs across 5 cities in Cameroon experienced a high burden of unintended pregnancy and TOP.
Given the history of inconsistent or no condom use, nonbarrier contraceptive use is highly indicated to prevent unintended pregnancies and associated health consequences, but coverage among FSWs was low.These results remind us that the health needs of FSWs in Cameroon extend beyond HIV and underscore the importance of removing structural barriers as integral to optimizing coverage of health care services.The evidence suggests that community-based and FSW-focused services in Cameroon do not currently meet these broader needs.Comprehensive services, including access to client-centered, voluntary family planning counseling and services, are needed to effectively address FSWs' fertility and reproductive health needs. 54ompared to other studies of FSWs, experience of TOP was similar or higher than observed across Southern Africa. 29,55TOP estimates among FSWs were twice as high as those reported among young women in urban areas or women seen in antenatal care (ANC) in other Cameroon-based studies. 38,40TOP performed in unsafe settings, even if supported by a health care worker, may lack appropriate counseling and information and miss opportunities to prevent future unintended pregnancy. 6,8,56Although this information was not collected here, among women with history of TOP in Cameroon, less than one-third reported currently using a non-barrier contraceptive and nearly one-quarter reported inconsistent condom use with clients.Subsequently, these women are likely at risk of future unintended pregnancies.
Findings from this study suggest that high burden of targeted violence in the context of criminalization experienced by FSWs may increase Continued vulnerability to TOP. 16 Links between violence and unintended pregnancy have also been reported in other settings, 57 and associations between violence and HIV acquisition among FSWs are well-described. 14,16,58The relationships between violence and unintended pregnancy are likely similarly multifaceted. 58,59Violence and condom failure are commonly related, 15,46 and past year condom failure was exceptionally common among this group of FSWs.In addition, it is possible that unintended pregnancy may itself trigger intimate partner violence. 60Given that physical violence is not factored into the conditions for legal abortion, FSWs experiencing physical violence leading to or arising from unintended pregnancy may perceive a lack of support to safely consider their options.Including access to emergency contraception and HIV postexposure prophylaxis as part of services for violence prevention and response could prevent unintended pregnancy and HIV acquisition resulting from sexual or physical violence.Programs that encourage social cohesion may help strengthen existing networks for violence coping and prevention mechanisms among FSWs. 16,61niformed officers have been common perpetrators of violence in this and other settings, 16 suggesting the utility of training these officers in the public health outcomes of punitive enforcement as well as legal advocacy to overcome policy-level Abbreviations: aPR, adjusted prevalence ratio; CI, confidence interval; FSW, female sex worker; PR, prevalence ratio; RDS, respondent-driven sampling; TOP, termination of pregnancy.a Excluded from adjusted analyses due to collinearity with HIV status.

Continued
structural determinants in promoting greater protection and access to health and justice for FSWs. 47,61nintended pregnancy is largely preventable through effective nonbarrier contraceptives or consistent and correct condom use.FSWs in Cameroon were generally reliant on condoms for contraception, with limited use of nonbarrier methods and even lower use of more effective LARCs.This finding is consistent with the low use of LARCs in general across Cameroon, 34 which despite government committment 62 has not expanded at the same rate as observed in countries across eastern and southern Africa. 63We also identified differences in contraceptive access within Cameroon, with relatively higher access in Bamenda despite regional sociopolitical unrest.A similar finding has been previously reported in relation to ANC services in Cameroon and may be due to long-standing investment and resource mobilization in the region where Bamenda is located. 64espite the reliance on condoms, most women did not report consistent condom use, particularly with nonpaying partners.There is substantial evidence from Cameroon and further abroad that pervasive structural factors, such as violence, stigmatization, price premiums for condomless sex, legal status, and policing, as well as social considerations, undermine FSWs agency to negotiate condom use consistently. 46,48,65Although current fertility intentions were not known, half of women were susceptible to pregnancy due to inadequate contraceptive use.Access to contraceptive choice and counseling is an essential part of sexual and reproductive health services. 54Expanding access to nonbarrier contraceptives and emergency contraception can reduce the risk of unintended pregnancy and related health, social, and economic consequences and provide women greater control over the timing and number of their children. 66upplementary interventions targeting the social and structural influences on condom use are still needed to promote the comprehensive protection of dual-method use for HIV and STI prevention. 11,16Alternative user-controlled interventions for HIV prevention such as through pre-exposure prophylaxis may also benefit FSWs. 11,66Programs implementing single-purpose interventions such as contraception, pre-exposure prophylaxis, and antiretroviral therapy to FSWs can benefit from maintaining a comprehensive view of sexual and reproductive health.For example, proactively considering existing contraceptive mix and accessibility as pre-exposure prophylaxis is scaled up can maximize opportunities to prevent unintended pregnancies and counter potential changes in condom negotiation power or resource allocation. 66,67In addition to meeting the broader needs of FSWs, STI education, screening, prevention, and treatment programs can help prevent the potential STIrelated risks associated with IUD use in the context of voluntary family planning services. 68ith the exception of HIV testing, indicators of HIV service access-in particular FSW-oriented community services-were associated with substantially lower coverage of nonbarrier contraceptives.In comparison, membership in an FSWoriented CBO was associated with higher recent HIV testing (data not shown).In Cameroon, most community-based facilities serving FSWs currently offer services fairly focused on HIV prevention, testing, and treatment support, with family planning often limited to counseling. 69Although these services appear to be an effective way to deliver HIV testing to FSWs, 70 they may miss opportunities for contraceptive provision.Requiring women to access multiple service points to meet their sexual and reproductive health needs may further disincentivize contraceptive use and act as a structural barrier. 71This may be particularly relevant given there are already well-characterized barriers affecting FSWs' access to family planning through public clinics. 72,73Further qualitative research to assess personal considerations, barriers, and facilitators of voluntary contraceptive use among FSWs in Cameroon, including emergency contraception, could better inform local family planning delivery strategies and integration of services adapted to the needs and preferences of FSWs.Elsewhere, it has been demonstrated that FSWled mobilization for HIV services through community drop-in centers has been associated with reductions in HIV incidence among FSWs in Tanzania. 70Expanding this community empowerment model for integrated family planning, STI, and HIV service delivery may represent an effective strategy to reduce unintended pregnancies and be responsive to critical health care needs.
Although there is growing evidence that integrated HIV and family planning delivery in a single location facilitates contraceptive uptake, 71,74,75 there are several considerations for implementation within existing community-based HIV services.There is empirical evidence that providing multiple contraceptive options improves uptake. 76Elsewhere, making longer-acting methods readily available to FSWs at a drop-in center alongside counseling on methods suited to a woman's needs and current fertility intentions significantly increased demand and uptake of effective contraceptives. 75However, the predominance of lower cadres of health care workers and peer outreach workers who often staff community-led FSW services may limit the number of staff with specific training and skills for contraception administration, especially when considering voluntary LARCs. 77,78Physical space with adequate privacy and space for an examination table required for LARC administration may also be limited in drop-in centers. 78Within the setting of community-based services, means to improve contraception availability and coverage may include task-sharing 79 and collaborations with services already providing family planning or ANC to offer periodic clinics with outside staff within the community-based settings.Active, facilitated referrals may reduce the number of women falling through the service gap 80 but may still pose barriers to some women.Lastly, not all FSWs in Cameroon are regularly accessing HIV testing 12 and the majority were not associated with an FSW-focused CBO at the time of study.A single model of HIV and family planning integration is unlikely to reach all women, and considering alternative points of service access for multipronged, bidirectional approaches to integration and family planning service delivery may optimize coverage. 81,82s for many working cisgender women, FSWs are often mothers with ongoing fertility desires. 83 pregnancy.Given pregnancy experience, including a sizeable proportion of pregnancies occurring after women began sex work, FSWs may benefit from greater links between FSW-oriented services and ANC.Approximately one-third of FSWs engaged in sex work during their last pregnancy, with potential for ongoing acquisition risks and resulting vertical transmission, especially if HIV is newly acquired in this period and missed by traditional approaches for preventing vertical transmission. 846][87] In our case, we identified associations between health care-related stigma and current use of nonbarrier contraception.Strengthening bidirectional collaborations between FSW-oriented services and ANC may foster more equitable access to quality and respectful family planning and ANC services for FSWs, including women living with HIV.

Limitations
These findings are subject to several limitations.Indicators related to sexual and reproductive health were asked as part of a broader survey on HIV risk and vulnerability.There was limited space to comprehensively explore domains related to unintended pregnancy and TOP experiences.Because we did not assess current pregnancy intentions, we could not determine the current unmet need for contraception.Pregnancy outcomes were assessed based on lifetime experience rather than specific to each pregnancy, and contraception was assessed in a different timeframe based on current use.TOP was used as an indirect measure of unintended pregnancy due to biases associated with direct, retrospective measurement of unintended pregnancy. 51Some discrepancy is expected given that not all unintended pregnancies will result in TOP and TOP may also occur after planned pregnancies.With the exception of HIV status, all indicators were based on self-report and may be subject to social desirability and recall biases.However, despite social desirability biases, reported history of termination was high and reported use of current contraception low.Importantly, data collected were cross-sectional and causality cannot be inferred.Lastly, findings are based on FSWs recruited from 5 urban cities; access to contraception and TOP likely differs among women residing in rural areas, and findings cannot be generalized.

CONCLUSION
In conclusion, experience with unintended pregnancy is common among FSWs in Cameroon, and women seeking TOP likely face risks associated with unsafe abortion given the restrictive legal environment.The ultimate direct and indirect impacts of TOP on health and mortality cannot be ascertained from these cross-sectional data, and longitudinal follow-up of women is needed to understand sexual and reproductive health outcomes that transcend HIV.The structural drivers of excess violence among FSWs remain a critical barrier to upholding the sexual and reproductive health and human rights of FSWs consistent with the United Nations Declaration of Human Rights.Reliance on condoms alone for preventing both HIV and pregnancy are failing the health needs of FSWs in Cameroon.Although the potential preventive impact of condoms has been substantial and continued promotion is critical, improved access to high-quality family planning counseling and a wider range of contraceptives, including nonbarrier and particularly long-acting contraceptives, is also necessary to improve clientcentered care, promote informed choice, and reduce unintended pregnancies.Ultimately, mitigating structural and facility-level barriers to the coverage of high-quality, voluntary family planning services and method choice in FSW-focused community services represents a key strategy to overcome current barriers to access and move towards optimizing sexual and reproductive health outcomes among FSWs.

TABLE 1 .
Sociodemographic Characteristics of Female Sex Workers in 5 Cities, Cameroon a Minor changes in denominator due to missing data.b Income through formal and informal work with a threshold of 50,000 XAF (US$90) per month based on national minimum wage (36,270 XAF/month).

TABLE 2 .
Continued Abbreviations: FSW, female sex worker.aReportedamong individuals reporting at least 1 of given partner type in an average week and refers to condom use during vaginal and anal sex.b Incudes regular and casual paying partners and regular and casual nonpaying partners.cRespondentscould select more than one contraceptive and percentages do not add up to 100.d Includes drinking whisky (n=22), traditional medicine (n=8), salted water (n=7), paracetamol/aspirin/quinine (n=5), menopause (n=5), Nescafé (n=3).currentlyusingany nonbarrier contraceptive method (Table3).Nonbarrier contraception was trending higher among women not living with HIV (27.6%, 469/1,698) compared to women living with HIV (22.9%, 126/550, P=.072).Approximately half (51.2%, 1,152/2,251) of women reported either nonbarrier contraceptive use or consistent condom use with all partners.

TABLE 3 .
Crude and RDS-Adjusted Prevalence of TOP and Current Nonbarrier Contraceptive Method Use by Study Site and Overall Abbreviations: CI, confidence interval; RDS, respondent-driven sampling; TOP, termination of pregnancy.

TABLE 4 .
Associations with Experience of TOP Among FSWs in Univariate and Multivariable Poisson Regression Corrected for RDS-Weighting and Clustering by Seed, Cameroon

TABLE 5 .
Associations with Nonbarrier Contraceptive Method Use Among FSWs in Univariate and Multivariable Poisson Regression Corrected for RDS-Weighting and Clustering by Seed, Cameroon

TABLE 5 .
Continued a Excluded from adjusted analyses due to collinearity with HIV status.
Consequently, they would benefit from family planning services that are adaptive to changes in fertility intentions and contraceptive needs, including services for safer conception and