Mapping evidence on predictors of adverse sexual and reproductive health outcomes among young women in South Africa: A scoping review

Background Globally, most young women have an unmet need for sexual and reproductive health (SRH), which remains a public health concern. Identifying the predictors can help reduce this challenge. Aim This scoping review maps evidence on predictors of adverse sexual and reproductive health outcomes among young women in South Africa. Method Askey and O’Malley’s framework guided this review. We searched five databases from January 2000 to December 2020 using relevant keywords, Boolean terms and medical subject heading terms. All relevant extracted data were organised into the study themes, and summary of all the findings were reported in a narrative format. Results Nine studies met the inclusion criteria out of 1219 studies identified. Four out of the nine studies were national-based studies, while the remaining five studies were conducted in Western Cape (two), Eastern Cape (two) and KwaZulu-Natal (one). Out of the nine studies included, three reported predictors of unintended pregnancy, while six reported predictors of sexually transmitted infections and HIV among young women in South Africa. The most prevailing predictors of adverse sexual and reproductive health outcomes were gender-based violence and alcohol use, while other predictors were lower socio-economic status, place of residence, multiple sexual partnerships, low education and being between the ages of 20–24 years. Conclusion We conclude that gender-based violence and alcohol abuse are the most prevailing predictors of adverse sexual and reproductive health outcomes among young women in South Africa.

A study conducted in SSA estimated a 29% of unintended pregnancies among women of reproductive age between 15 years and 49 years, while young women between 15 and 24 years contributed 32%. 20 In 2020, an estimated 417 million young people had herpes simplex virus type 2 (HSV-2) infections. 21 World Health Organization (WHO) estimated an annual 357 million new episodes of four curable STIs, including chlamydia, gonorrhoea, syphilis and trichomoniasis globally. 22,23 Globally, 15% of women living with HIV are between the age of 15-24 years, of whom 80% reside in SSA, and for more than 30 years, this age group has remained much at higher risk of HIV infections than their male counterparts, thus reporting more than 380 000 new HIV infections yearly. 24 South Africa has been identified as one of the countries in SSA contributing significantly to the continuous increase in adverse sexual and reproductive health outcomes in terms of unintended pregnancy and STIs/HIV among young women. 25,26,27 This might be because of the high prevalence of early sexual debut among this cohort, 28 with about 41% of unintended pregnancies reported in a recent study. 27 Sexually transmitted infections and HIV prevalence have also been higher among South African young women, with approximately 9.8% STIs prevalence 29 and 21.0% HIV prevalence. 29 Furthermore, some studies argued that young women's socio-demographic, economic and structural characteristics in South Africa are some of the risk factors influencing adverse sexual and reproductive health outcomes. 30,31 However, no study in South Africa has synthesised evidence from the primary studies available on adverse sexual and reproductive health outcomes with a focus on unintended pregnancy and STIs/HIV using scoping review, which is important in actualising sustainable development goal 3 (SDG 3) by the year 2030, 32 global agenda goal of 90-90-90 towards the eradication of HIV 33 and other national strategic plans. 34,35 Thus, this scoping review maps evidence on predictors of adverse sexual and reproductive health outcomes among young women in South Africa. The results from this review will give an account of the prevailing factors predicting adverse sexual and reproductive health outcomes in terms of unintended pregnancy and STIs/HIV among young women in South Africa. Additionally, the result could also shape the direction of the national policymakers in developing targeted interventions that will address the adverse sexual and reproductive health outcomes among young women in South Africa, as no study in South Africa has been able to underscore this.

Design
We adopted the Arksey and O'Malley guidelines methodological frameworks 36 in designing this study, and we followed the checklist of Preferred Reporting Items for Systematic Reviews and Meta-analyses extension for scoping reviews (PRISMA-ScR) to report this study's findings. 30,31

Research question
This study research question was: 'What the predictors of adverse sexual and reproductive health outcomes, including unintended pregnancy, STIs/HIV among young women aged 15-24 in South Africa are?' The eligibility criteria for potential articles to address the research question were determined using the population concept context (PCC) framework, as depicted in Table 1 30 .

Literature search strategy
We searched for publications on predictors of adverse sexual and reproductive health outcomes in terms of unintended pregnancy and STIs/HIV among young women in South Africa between January 2000 and December 2020. A preliminary search was performed on PubMed using the keywords 'predictors' AND 'sexual and reproductive health' AND 'young women' AND 'South Africa' to ensure a comprehensive search. The returned index terms on the search were then used to develop comprehensive terms using the peer review of electronic search strategies 2015 guideline checklist to assess, evaluate and revise our comprehensive vocabulary before performing our main search. We then searched using all identified Medical Subject Headings (MeSH) terms to ensure that all relevant studies were included. The following databases were searched African journals online, web of science, Scopus, PubMed and CINAHL.
To identify grey literature, WHO, United Nations Population Fund (UNFPA) and Guttmacher web pages were searched, while Google Scholar was searched for published theses. Finally, we searched the reference list of all review studies found in the search to determine more eligible studies. Eligible studies were exported to the Endnote version X9's library for abstract and full article screening after the keyword search and title screening. Articles published between January 2000 and December 2020 were included.

Eligibility criteria
The following inclusion criteria were applied: (1) studies on unintended pregnancy predictors among young women in

Study selection
After the retrieved articles were exported to EndNote X9's library, the study selection was in three phases. An independent reviewer conducted the title screening and duplicates removal. Another reviewer conducted the abstract and full article screening. Two independent reviewers were involved in this process by following the study inclusion and exclusion criteria. The research question derived from the eligibility criteria was used to develop a screening form used throughout the screening process. In cases of unresolved disagreement, another member of the research team (T.B.) provided the third opinion through a thorough discussion on why an article should be included or excluded.

Charting and extracting the data
A data charting form was developed to help determine the suitable variables to be extracted to answer the study research question. Data from included studies were extracted using the following domain author and publication year, study aim, study province or country, study design, research method, sample size (n), study population, age group, predictors of unintended pregnancy, predictors of STIs and HIV (key and significant findings)

Collating and summarising
We employed descriptive statistics to summarise the studies' key findings, locations and focus using Microsoft Excel 2020, while NVivo version 12 was used to perform content analysis.

Quality of evidence
A mixed-method assessment tool (MMAT), version 2018, was adopted and piloted by the two independent reviewers to assess the consistency of the selected studies. 33 The MMAT was used to determine the study's aim, adequacy, methodology, study design, data collection, study selection, data analysis, results presentation, author discussions and conclusions. An overall quality percentage score for each of the included studies was calculated, and scores were interpreted as high quality (76% -100%), average quality (51% -75%) and low quality (≤ 50%). Irrespective of the score, all assessed studies were included.

Ethical considerations
This study was approved by the Humanities and Social Science Research Ethics Committee of the University of KwaZulu-Natal on the 21st of February 2021, with protocol reference number HSSREC/00002192/2020.

Results
The PRISMA flow diagram depicted (Figure 1 36 ) shows the process of article screening, inclusion and exclusion. The initial search yielded 1224 results, from which we excluded 1096 because of duplicates and title unconformity. A total of 25 studies were assessed for full-text eligibility, which led to further exclusion of 16 that did not meet our inclusion criteria, leaving us with nine studies to be included in our analysis.

Characteristics of the included studies
Four (44%) out of nine studies included were national-based studies, 37,38,39,40 while the remaining five (66%) studies were conducted in different provinces; Western Cape 41,42 and Eastern Cape 27,43 had two studies each and one study was conducted in KwaZulu-Natal. 26 Six (67%) were crosssectional study design, 26,27,37,38,39,40 , two (22%) studies followed longitudinal design 41,43 and one (11%) was a randomised control trial. 42 All the included studies were quantitative. The sample size of the included studies ranged from 88 41 to 11 904 40 respondents. The respondents were aged 15-24. • ArƟcles focused on women older than 24 years (n = 10) • ArƟcles that lacks clarity in the outcomes (n = 6) Records aŌer the removal of duplicates and Ɵtle screening (n = 128) Assessment of full-text arƟcles for eligibility (n = 25) ArƟcles included for quality assessment (n = 9) ArƟcles included in the study (n = 9) Screened records for abstract (n = 128) http://www.phcfm.org Open Access Table 2 indicated that three out of nine included studies reported predictors of unintended pregnancies among young women between the age of 15 years and 24 years in South Africa. 27,37,43 One of these was conducted among young women in Eastern Cape, South Africa, and reported that young women between 17 years and 24 years, young women who have ever had a life experience of sexual abuse and young women who ever drank alcohol were more likely to report unintended pregnancies. 27 The second was a longitudinal cohort study conducted among adolescent girls aged 15-18 years. The study showed that adolescent girls who had lifetime physical abuse, those within the lower socio-economic status and the assertion that teenage girls and her boyfriend were main mutual partners were predictors of unintended pregnancies. 43 The other was a nationwide representative cross-sectional study conducted among young women between the age of 15 years and 24 years in South Africa. The study concluded that young women with higher education, those not living together with a spouse, young women residing in the KwaZulu-Natal province and those having regular partners were more likely to report unintended pregnancies. 37

Predictors of STIs/HIV among young women in South Africa
Six studies reported predictors of HIV and AIDS among young women in South Africa, 26,38,39,40,41,42 as shown in Table 2.
A longitudinal study conducted in the Western Cape among HIV-positive and negative young women between the age of 17-21 years reported that persistence of high risk-human papilloma virus (HR-HPV) occurred higher among young women who were HIV-positive both at 6 and 12 months follow-up compared to HIV-negative young women. 41 One was a cross-sectional study conducted among young women aged 15-24 years in KwaZulu-Natal, South Africa, with the aim of investigating the acceptability and feasibility of home-based sampling of STIs and Bacterial Vaginosis (BV) and to measure prevalence and factors associated with STIs/BV. The study concluded that young women in school or those working were less likely to experience STIs/BV. However, young women who had ever drunk alcohol, those involved in genital touching and young women with HSV-2 were more likely to have STIs/ BV. 26 The randomised control trial of 388 young women between the ages of 16 years and 24 years conducted in the Western Cape showed that young women involved in alcohol use and those who reported having a sexual partner with STIs were more likely to report having concurrent HPV/STIs. 42 The national cross-sectional survey of 3092 adolescent girls and young women (AGYW) in South Africa reported that young women between 20 years and 24 years and AGYW who reported condom use at last sexual intercourse were more likely to report HIV infection. Adolescent girls and young women with sexual partners within five years of age, with tertiary level education, having low risk of alcohol use and having one sexual partner were less likely to report HIV infection. 38 Another was a wide range of national studies conducted by Pettifor and Rees. 40 It concluded that young women with higher odds of reporting HIV infection had a history of unusual vaginal discharge in the past 12 months, were of the black race, resided in the urban area and were aged 20-24 years. Besides, they were young women with educational levels below high school, young women who have been sexually active for more than 12 months, with an additional sexual life partner, and young women involved in condomless sex at most recent sex. Similarly, another nationally representative study of HIV infection among 1708 young women aged 15-24 years reporting one lifetime partner concluded that young women who had not completed high school and young women between the age of 20 years and 24 years were more likely to be infected with HIV. 39

Discussion
Our scoping review was conducted to map existing literature on the predictors of adverse sexual and reproductive health outcomes among young women in South Africa.
Three (33%) of the nine included studies reported predictors of unintended pregnancy among young women in South Africa. 27,37,43 Sexual abuse, use of alcohol, 27 physical abuse, 37,43 an assertion on the main mutual partnership, lower socioeconomic status, 43 higher educational attainment, living arrangement, provincial residence and having a regular partner 37 were the identified predictors.
Physical abuse as a predictor of unintended pregnancy was reported twice in the study by Christofides and Jewkes 43 and Ibisomi and Odimegwu, 37 signifying a key predictor of unintended pregnancy among young women in South Africa. This is similar to a study conducted among Bangladeshi women who concluded that unintended pregnancy is higher among women who had ever experienced physical violence, 44 and the same conclusion was reported in a similar study conducted in Peru among pregnant women. 45 Sexual abuse and alcohol use were identified as predictors of unintended pregnancy among young women in South Africa. 27 This is in concordance with a study conducted in Nepal by Acharya and Paudel 46 who concluded that young women who had ever experienced sexual violence had a higher likelihood (2.3%) of reporting unintended pregnancy. Similarly, a study conducted by Ahinkorah and Seidu 47 in 22 SSA countries also linked sexual violence to unintended   pregnancy. In the same vein, a self-reported periconceptional substance use study conducted among women reported that women who use alcohol were more likely to report unintended pregnancy. 48 We found that higher educational attainment, not living together with a spouse, residing in the KwaZulu-Natal province and having a regular partner were predictors of unintended pregnancy among young women in South Africa. 37 These predictors are similar to the conclusion of a study conducted on individual and contextual factors associated with unintended pregnancy among 6791 AGYW (aged 15-24 years) in 10 SSA countries.
We found higher education attainment and variation in type place of residence as predictors of unintended pregnancy among young women in South Africa. 49 Women's level of education has always been a critical factor in decision-making, and the impact of educational attainment on sexual and reproductive health, particularly unintended pregnancy, has been discussed in the past. 50 This study result aligns with a study conducted in Ghana by Ameyaw 51 and a study conducted in Ethiopia by Habte and Teklu. 52 As young women with a higher level of education are often employed, it could be that the young women often miss their family planning service appointment or find it difficult to plan their reproductive health needs because of work pressure. 53 Lower socio-economic status was also a predictor of unintended pregnancy identified in this study. 43 . This finding is in line with a study conducted by Iseyemi, Zhao 54 that women with lower socio-economic status tend to report unintended pregnancy compared to their counterparts with higher socio-economic status. This finding might be because young women with lower socio-economic status could not make independent financial decisions or could not afford suitable contraception methods. 49,55 The remaining six studies included in this study reported predictors of STIs/HIV among young women in South Africa. 26,38,39,40,41,42 The study findings show that being HIVpositive was reported as a predictor to HR-HPV among young women in South Africa. 41 Alcohol use, touching of the genital area, being HSV-2 positive were predictors of having STIs/BV 26 and having a sexual partner with STIs was reported as a predictor of having concurrent HPV/STIs. 42 Similarly, a study conducted by Seth and Wingood 56 among young African American women also identified alcohol use as a strong predictor of having STIs. In the same vein, being exposed to any form of STIs, including HIV and HS-2, has been reported to be a predictor of having any other form of STIs in the later life of young women. 57 Furthermore, our study reported that the use of condoms during sexual intercourse in the last 5-year was reported as a predictor of HIV, 38 young South African women whose population group were black, those who have been reporting unusual vaginal discharge in the past 12 months, those residing in the urban area, those with lower educational attainment, being sexually active, having multiple sexual partners, those involved in condomless sexual intercourse 40 and young women between the age of 20-24 years were more likely to report having HIV. 39 This study's results are in corroboration with a study conducted in Uganda by Nankinga and Misinde, 58 which concluded that having multiple sexual partners remains one of the strong predictors of having STIs, including HIV. Similarly, another study conducted in Ethiopia supported these findings. 59 We also identified involving in condomless sexual intercourse as a predictor of reporting being HIVpositive among young women in South Africa; however, a study conducted in Malawi opposed this finding as the study reported that there is no significant relationship between condom use during sex and being HIV-positive, 60 while another study conducted in Uganda supported our findings that condom use during sexual intercourse reduces the likelihood of being HIV-positive. 61 However, our study also reported that condom use during sexual intercourse in the last 5-year could be a risk factor for HIV. 38 This may be that the practice of condom use started after the young women had contracted the virus as such condomising during sexual activities could halt the transmission of the virus to their sexual partners. 62 Finally, our findings indicated that young women in South Africa between the age of 20-24 years, those with less education and those residing in urban areas were more likely to report being infected with HIV. These findings are in line with mapping studies conducted in Uganda and Malawi by Chimoyi and Musenge 61 and Nutor and Duah, 63 respectively, concluded that young women who reside in the urban area, those less educated and those between the age of 20-24 years have higher odds of reporting being HIV-positive compared to their counterparts who reside in a rural area, more educated and those between the of age 15-19 years.

Implication for policy and public health
This study's findings are relevant to policy and public health. Predictors of adverse sexual and reproductive health outcomes among young women in South Africa identified in this study are important in developing required interventions and policies to halt its continuous increase. Among identified predictors of adverse sexual and reproductive health outcomes among young women in South Africa, genderbased violence was most prevailing to unintended pregnancy, while alcohol use was identified as the common predictor to both unintended pregnancy and STIs/HIV, and this calls for the attention of policymakers, government agencies, non-governmental organisations and key stakeholders in policy formation and implementation to develop and implement behaviour and social policy that will be targeted at eliminating gender-based violence exposure and alcohol use abuse among young women in South Africa to reduce or eliminate adverse sexual and reproductive health outcomes.
Regarding public health implications, this study's findings also identified residences (area of residence or province of residence) and other health behavioural factors as predictors of adverse sexual and reproductive health outcomes. This will provide public health providers with the blueprint that will guide their programmes and interventions. Furthermore, it will help ensure optimal utilisation of scarce resources to reduce adverse sexual and reproductive health outcomes, including unintended pregnancy and STIs/HIV among young women in South Africa.

Implication for research
These research findings identified gender-based violence exposure and alcohol use as the most prevailing predictors of adverse sexual and reproductive health outcomes among young women in South Africa. However, what is unknown is the degree to which these two predictors influence these adverse outcomes at household and community levels; as such future studies should consider conducting primary research that will help the policymakers and stakeholders in understanding the extent to which adverse sexual and reproductive health outcomes among young women in South Africa are being influenced by these predictors (genderbased violence exposure and alcohol use) at both household and community levels.

Strengths and limitations
This study has numerous strengths. Firstly, scoping review allows different study designs to be included in a study, and this allowed the authors to systematically search and select relevant literature to describe and map the available evidence on adverse sexual and reproductive health outcomes among young women in South Africa. This study design also allowed us to further identify literature gaps useful to inform future research and policy implications. To the best of our knowledge, this study is the first scoping review focusing on adverse sexual and reproductive health outcomes among young women in South Africa in terms of unintended pregnancy and STIs/HIV covering 20 years range. However, there is no study without limitation. The study possibly failed to capture some other relevant articles because we searched fewer databases in the study. Also, we were not involved in the study design and data collection of any of the studies included in this study. Despite these limitations, this review provided evidence useful to guide future research in eliminating or reducing adverse sexual and reproductive health outcomes among young women in South Africa.

Conclusion and recommendation
This study has revealed evidence of existing literature on adverse sexual and reproductive health outcomes among young women in South Africa while identifying gaps for future research. We conclude that predictors of adverse sexual and reproductive health outcomes (i.e. unintended pregnancy and STIs/HIV among young women in South Africa) were lower socio-economic status, place of residence, multiple sexual partners, lower educational level and being between the ages of 20-24 years. However, the most prevailing predictors among all the identified predictors were gender-based violence (sexual and physical abuse) and alcohol use. This study findings call for the attention of key stakeholders towards policy formation and development of behaviour and social intervention, such as encouraging the young women to speak up after the incidents and constant community engagement by government or non-governmental organisations on the adverse impact of gender-based violence and alcohol use on sexual and reproductive health outcomes of young women in South Africa. The community engagement can be in the form of playlets and one-on-one discussions, while the victims who speak up should be given rewards to encourage other young women to speak up. Particular interest should be channelled towards the marginalised communities in the urban areas and hard-reach areas in rural South Africa.