Study site

This study was part of a large baseline survey to collect information on HIV, Sexual Reproductive Health (SRH) and Gender Based Violence (GBV) at district level. The study was designed to inform the design of interventions targeting AGYW aged 10–24 years in 20 target districts from all ten provinces of Zambia. Data was collected from: Central Province (Kabwe, Chisamba, and Mkushi), Copperbelt Province (Chililabombwe, Ndola, Kalulushi), Eastern Province (Petauke, Nyimba), Luapula Province (Mwense, Chembe), Lusaka Province (Chongwe), Muchinga Province (Chinsali), Northern Province (Kasama, Mungwi), North-Western Province (Mwinilunga), Southern Province (Sinazongwe, Kalomo, Mazabuka), Western Province (Limulunga, Mongu), Fig 1 below shows the provinces and districts selected in the study.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 1. Provinces and districts of the study area in Zambia.

https://doi.org/10.1371/journal.pone.0294545.g001

The Zambia AGYW program is part of the Global Fund Strategy (2017–2022) to reduce new HIV infections among AGYW by 58% by 2022 in13 sub-Saharan African countries including Botswana, Cameroon, Kenya, Lesotho, Malawi, Mozambique, Namibia, South Africa, Eswatini, Tanzania, Uganda, Zambia and Zimbabwe [18, 19].

Data collection procedures and methods

Data was collected using three data collection methods:

1. extensive desk/documents/literature reviews;
2. secondary data analyses and
3. primary data collection as described below.

a) Extensive desk/documents/literature review: Prior to data collection, we conducted a review of existing published (and unpublished) data on adolescent girls and young women’s access to and utilization of HIV and sexual and reproductive health services. The key documents reviewed included:

• Reports from key implementing partners targeting adolescent girls and young women (e.g. DREAMS, Churches Health Association of Zambia, Discover health, etc.)
• Zambia Demographic & Health Survey Report (2007, 2013/14, 2018)
• Zambia Population-based HIV Impact Assessment Report (2016)
• Joint UN Program on HIV/AIDS reports
• Annual reports from UNFPA and UNICEF
• Annual performance reports from the Ministry of Health (with a specific focus on sexual and reproductive health)
• National Strategy on Ending Child Marriage in Zambia 2016–2021
• Zambia National Health Strategic Plan 2017–2021 [5]
• Reports from the Ministry of Education and Ministry of Community Development and Social Services among others.
• Adolescent Health Strategy 2017–2021

We also reviewed published papers on adolescent girls and young women regarding their access to and utilization of HIV, GBV and sexual and reproductive health services [10].

The desk review that we undertook helped to:

• Clarify the context/scope of the study and the gaps in the existing information in relation to the proposed objectives.
• Provide a guideline on what primary information is required and needs to be collected during the subsequent phases.
• Guide in developing instruments to collect adequate data for this exercise.
• Provide a situational analysis of what was happening currently as documented on girls and women in- and out-of-school, second chance interventions/programs to get back girls and women out of school back to school, facilitators and barriers to access and uptake of HIV/SRH services as they related to population growth and market access approaches.

The desk review we proposed in this study was implemented as follows:

Purposive sampling was used to identify the relevant documents that were to be reviewed for the study. At the national level, the study team reviewed different documents including; the Adolescent Health Strategic Plan 2011 to 2015, the National Standards and Guidelines for Youth Friendly Health Services and the Reproductive, Maternal, Newborn, Child and Adolescent Health and Nutrition Communication and Advocacy Strategy 2018–2021. Additional Information was collected from the Ministry of Health and NGOs providing SRH services to adolescents and young people in Zambia.

b) Secondary data analysis: Secondary analysis of Zambia Population-based HIV Impact Assessment (ZAMPHIA 2016) and Zambia Demographic and Health Survey (2007, 2013/14, 2018) [20] was conducted to generate trends in HIV prevalence and sexual risk behaviours (e.g. transactional sex, teenage pregnancy and motherhood, having multiple sexual partners, and non-use of condoms in high-risk sexual relationships); health-seeking behaviours, access to HIV services (e.g. HIV counselling and testing, pre or post -exposure prophylaxis, condom promotion and provision, and sexually transmitted infections (STI) screening and treatment) and access to SRH services including post-abortion care, contraceptive services among AGYW.

c) Primary data collection: This comprised of administration of survey questionnaires (quantitative interviews) to randomly selected adolescent girls and young women living within the selected districts. All AGYWs enrolled for the quantitative component of this study were subjected to HIV and syphilis testing to determine the sero-prevalence of the two diseases. Besides the quantitative interviews, focus group discussions, in-depth interviews and key informant interviews were conducted with purposely selected respondents to explore facilitators, opportunities and barriers to utilization of HIV, GBV and SRH services among AGYW, among other aspects. The data collection methods varied by each objective. However, in this paper, we only concentrate on the quantitative findings.

Study design and population

This was a cross-sectional, mixed-methods study. A convergent research design was employed to combine the qualitative and quantitative data collected among the in-and-out-of-school AGYW aged 10–24 years across the 20 districts. This implies that data collection and analysis of both quantitative and qualitative data was conducted simultaneously and results were integrated in the interpretation phase (Moseholm and Fetters, 2017) [21]. In-school AGYW were those that were currently in school at the time of the survey while out-of-school AGYW were those dropped out of school prior to school completion or adolescents and young women who were out of formal education or enrolled in a vocational training institution at the time of the interview. Hence, this excluded enrolling out-of-school AGYW who were not in school at the time of the survey because they completed school.

The study population included in-school and out-of-school adolescent girls and young women aged 10–24 years, resident in the selected districts. Besides AGYW, we also interviewed selected key informants involved in HIV, GBV and SRH programming including implementing partners. A list of key informants is listed in Table 1 below.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 1. List of major sources of key informants.

https://doi.org/10.1371/journal.pone.0294545.t001

Sample size determination

Using the prevalence formula, a sample size calculation showed that a total of 12560 was needed to detect an effect size if present in the target population. The sample size was calculated per age group using the total projected population for the year 2020 and the HIV prevalence as estimated by the Zambia Population-based HIV Impact Assessment (ZAMPHIA) [6]. The population projections were obtained from the Central Statistical Agency website. To obtain the sample size per age group, the following formula for sample surveys was used Lwanga and Lemeshow. [22].

Where;

P is the anticipated prevalence of the study outcome (here we used the estimated HIV prevalence among in target population (Ministry of Health) [5]), N is the target population size in each age group, d is the precision set at 0.8% (taking into account a Bonferroni correction) and Z is z-score set to 1.96 at α = 5% significance level.

To correct for non-response, the estimated sample size in each district and in each age-group was adjusted using the formula below

Adjusted Sample size =

Where;

• n is the calculated sample size in each age-group and district and,
• is the anticipated non-response rate (10%).

After adjusting for non-response, total sample size was 13960.

The calculated sample size of 13,960 (Table 2) was then allocated among the in-school and the out-of-school using estimates national level estimates from the 2015 Zambia Living Conditions Monitoring Survey (LCMS) on the proportion of females in school and out-of-school. Using the 2015 LCMS, we assumed 84.8% of 10–14-year-olds, 73.4% of 15–19-year-olds and 22.5% of 20–24-year-olds were in school.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 2. Sample size determination by selected district.

https://doi.org/10.1371/journal.pone.0294545.t002

However, we managed to collect a total of 12, 813 (corresponding to 92% response rate) constituting 5979 (46.7%) in-school and 6834 (53.3%) of out-of-school participants. This was slightly above the estimated minimum sample size for the study.

Sampling procedures

The sampling procedures was done with respect to addressing the following study objectives:

1. (a) Objective 1: To determine socio-demographic, sexual, health and behavioural status among adolescent girls and young women in targeted districts in Zambia

Approach.

To create the Adolescent Girls Multilevel Vulnerability Index (AGI), we followed a four-step process as outlined in the Handbook on Constructing Composite Indicators (OECD 2008) [24]. The first step was the development of a theoretical framework to inform selection of indicators into a composite score. We matched the framework with the available data sources and indicators to guide us in constructing a composite index. Lastly, we tested the validity and robustness of the vulnerability index through a sensitivity analysis. We illustrate each step in detail below.

Step 1: Developing the theoretical framework. For the purpose of the index, we followed the United Nations definition of adolescence, which refers to individuals between the ages of 10 and 19 years (UNICEF 2003) [25]. The AGI is focused on adolescent girls rather than boys because of the unique risks faced by girls in this age group. Consequently, we concentrate on the girl child in this study. The indicators included in the index were based on research and expertise on adolescent girls. Risks such as child marriage, limited economic opportunities, and parenthood have different effects on the outcome of the adolescent girls. Vulnerability was defined as a relationship between the context in which a girl lives and a set of factors that put her “at-risk” for negative outcomes. These factors are better explained by socio-ecological model and are presented at multiple levels and (Fischhoff, Nightingale, and Iannotta 2001) [26], including the level of the:

1. individual girl (e.g., a chronic health condition like asthma), genetic (e.g., predispositions to disease), and psychological (e.g., thoughts, emotions, and behaviours);
2. her proximal social settings, (such as familial structures (e.g., orphan), peer networks (e.g., older, more experienced peers), classrooms (e.g., teachers who do not show up for work), and community characteristics (e.g., violence, gangs)
3. more distal levels (girl can also be influenced negatively by societal practices and beliefs e.g., female genital cutting and a policy environment that does not recognize her rights)

In addition, there was a need to promote positive interactions between the development of a young person and her context to increase the likelihood that she will thrive (Lerner et al. 2011) [27]. This calls for programs and policies to focus not only on prevention and problem reduction, but also on creating contexts to promote thriving. It has been noted that vulnerability affects girls at different levels listed above. The AGI index conceptualizes vulnerability using the multi-dimensional measures lens as well as at multiple levels of risk. Therefore, the AGI alludes to the notion that comprehensive actions are more effective than population-based actions that focus on a single issue.

Step 2: Selecting the data and potential indicators. For individual and household level indicators, we collected primary data using the quantitative questionnaire. The questionnaire collected data on the following indicators:

1. Individual level: We selected the indicators relating to schooling and family structure, marital status, maternal health, high-risk sex and family formation.
2. Household level: Data was collected on housing conditions and household and family structure.

For the community risk factors, we conducted secondary data analysis of the 2018 Zambia Demographic Health Survey.

1. Community risk factors: Data included child marriage, illiteracy, and low comprehensive knowledge of HIV/AIDS.

Step 3: Constructing the index. Vulnerability index. Vulnerability was assessed as risk factors within three dimensions (individual, household and community, see Table 3 below). Data was summed to create the composite indicator with multiple cut-off points at different levels. The indicators were dichotomised and coded as 0 for “no” and 1 for “yes.” Variables were standardized so that each contributed equally to the overall summary statistic. The denominator included all girls in the sample. At the individual level for girls aged 10 to 14, a girl was classified as vulnerable if they were at least two years behind grade for age or were not in school and/or not living with their parents. The cut off for individual girls ages 15 to 19 years, one was considered to have individual level vulnerability if she has ever been married, or given birth or currently married, or did not attend secondary school, or engaged in high-risk sex, that is, sex under the age of 15 or multiple non regular partners. At household level, a girl (10–19 years) was classified as vulnerable if she experienced any two of the following five conditions: no access to improved source of water, no access to improved sanitation, household head has no education, food insecurity (no access to food in a day), and non-family support, that is, ever consulted other for social support other than a family member. At the community level, girls were considered vulnerable if they lived in a community characterised by any one of the following: high rate of early marriage before the age of 18, high rate of illiteracy, increased prevalence of HIV, and low comprehensive knowledge of HIV. At each level, a score of 1 was given if a girl experienced these measures and 0 otherwise. We then used principal component analysis on the scored data (0/1) to derive the vulnerability index. The vulnerability index was divided into tertiles (low, medium and high) with the highest tertiles representing the most vulnerable group.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 3. Categorization of vulnerability index.

https://doi.org/10.1371/journal.pone.0294545.t003

Measurement of other variables

Similar to Matovu et al. [11], the dependent variables were: 1) sexual-risk behaviour and 2) prevalence of HIV and syphilis infections, assessed separately among in-and-out-of-school AGYW. Adolescent girls and young women were deemed to have engaged in sexual-risk behaviour if they: a) reported a history of sexually transmitted infections; or b) reported that they had their first-time sexual experience before the age of 15; or c) had sexual intercourse with multiple (2+) sexual partners in the past 12 months; or d) did not use a condom or used condoms inconsistently with their most recent sexual partner. No attempt was made to create one composite variable of sexual-risk behaviours because each behaviour was considered to constitute a level of HIV/STI risk on its own. HIV and syphilis prevalence was determined as a percentage of those tested who tested positive for HIV or syphilis.

The independent variables included age-group (categorised as 10–14, 15–19, and 20–24 years), highest level of education attained at the time of the survey (in-school AGYW were asked about their current class of attendance), marital status (categorised as ‘never married’, ‘in a relationship but not married’, ‘married or in union’ and ‘divorced/widowed/separated’), history of HIV testing (ever tested for HIV; tested for HIV in the past 12 months), alcohol use before sex, wealth tertile (categorised as ‘low’, ‘middle’, and ‘high’), comprehensive knowledge of HIV (categorised as ‘low’, ‘medium’, and ‘high’) and vulnerability index (categorised as ‘low’, ‘middle’, and ‘high’). A detailed description of how wealth tertile, comprehensive HIV knowledge and vulnerability index were derived are given in the sections above.

Data analysis (objectives 1, 2, and 3)

Descriptive statistics such as frequencies, proportions, means, medians, standard deviation, and interquartile range were used as appropriate to summarise the characteristics of the study participants, by district, age-group, and schooling status. HIV and syphilis status were determined and presented in form of percentages, that is, frequencies and percentages of those tested for HIV and syphilis. HIV and syphilis status were cross tabulated by district, age-groups and stratified by schooling status to determine differences in the prevalence of these diseases by those characteristics. The results were presented in tables and figures as appropriate. To model possible associations between socio-demographic factors and an adolescent being HIV or Syphilis positive a logistic regression model was used.

Ethical considerations

Ethics approval was sought from UNZABREC (REF. 2460–202) The protocol was then submitted to the Zambia National Health Research Authority (Ref No: NHRA000010/10/02/2022) for registration and Approval. All participating implementing partners (IPs) were asked for permission to have their staff participate in the study. A letter of introduction from the ministry of education and ministry of health was also obtained to ease coordination with the selected organizations and partners.

Informed consent was obtained from all the participants prior to participation, and we ensured confidentiality of data obtained. The study used written informed consent from each participant prior to data collection solicited by a trained study team member. Potential respondents were given an opportunity to ask any questions about the research before they are asked if they were willing to participate. Potential respondents who agreed to participate were interviewed. The study participants were provided with written informed consent and information participant sheets detailing the study, the risks and benefits, and emphasis on the protection of confidentiality.

For adolescents less than the age of 18 years, consent was sought from their parents for the structured interview, FGD, IDI participation and the HIV and Syphilis testing. Assent was also sought from the adolescent’s 10–17 years old. Permission to conduct the structured interview, FDG or IDI and blood tests was only considered granted if both the parents/guardians and the adolescent signed the consent and assent forms respectively. However, it was also explained to the adolescents that they can opt out of the study at any time during the interview or blood test. Results from the qualitative analysis will be published in a separate paper (submitted).

A separate consent form was obtained for HIV and Syphilis testing. Adolescents who consented to participate in answering the structured questionnaire could further consent to having an HIV test and/or a Syphilis test only. Therefore, only Adolescents who consented to both the structured interview and the HIV and/or Syphilis test were tested. The results of the blood tests were not shared with anyone except the participants if requested.

To ensure privacy, HIV testing, and counselling was conducted within the dwelling or premises of the participants’ residence. Adolescents who consented to the HIV and/or Syphilis test, were tested by a laboratory scientist after counselling. Counselling sessions was conducted one on one with a trained counsellor. Adolescents who participated in HIV and/or Syphilis testing were offered both pre and post HIV counselling.

Covid prevention measures

To prevent the spread of Covid-19 among the research assistants and the participants, disposable masks were provided to both the participants and the research assistants. Before commencement of interviews, both the research assistants and the participants sanitized their hands with the alcohol-based hand sanitizer that were provided. Furthermore, a distance of at least 1 meter was maintained during the interview.

Distribution of samples over districts stratified by in-and-out of school participants

Table 4 shows the distribution of the 12,813 (91.8% response rate) study participants. Of those interviewed, 5,979 (46.7%) were in school while 6,834 (53.3%) were out of school. The age band 15–19 had the highest number interviewed from the in-school, 3150 respondents, while the age band 20–24 had the highest interviewees among the out of school, with 5,241 respondents. Mwinilunga, Chinsali, Chisamba and Chembe districts had the highest number of respondents, while Sinazongwe and Mungwi districts contributed the least.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 4. Distribution of the 12813 individuals interviewed for the AGYW study.

https://doi.org/10.1371/journal.pone.0294545.t004

Respondents’ characteristics

Table 5 shows the distribution of AGYW participants by selected background characteristics. The overall age distribution was such that 12.6% (n = 1617) of those interviewed were aged 10 to 14 years, 35.4% (n = 4536) were aged 15–19 years, and 52.0% (n = 6660) were aged 20–24 years. By school status, 46.7% (n = 5979) were in-school while 53.3% (n = 6834) were out-of-school. About half 52.1% (n = 3117) of in-school AGYW were in secondary school, 42.5% (n = 2540) were in primary school while only 5.4% (n = 322) were in vocational or tertiary institutions. For the out-of-school AGWY, 37.9% (n = 2591) attained primary education, 29.0% (n = 1982) attained secondary education while 25.5% (n = 1742) had vocational or other tertiary education.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 5. Percentage distribution of the participants, overall and by school status.

https://doi.org/10.1371/journal.pone.0294545.t005

Overall, 21.6% (n = 2765) of AGYW reported that they were married or cohabiting while 19.4% (n = 2481) were in a relationship but not married. As one would expect, a smaller proportion of AGYW of the in-school, {1.5% (n = 92)} compared to {39.1% (n = 2673)} of out-of-school AGYW reported that they were currently married or cohabiting. Slightly above three quarters 77.5% (n = 4632) of the in-school and about a third 32.8% (n = 2244) of out-of-school AGYW had never married while 1.9% (n = 114) and 8.4% (n = 577) respectively were divorced or widowed at the time of the interviews. Among those who were married or cohabiting, the age of their partners varied by schooling status: 31.1% (n = 134) of the in-school and 7.2% (n = 42) of the out-of-school were married to partners aged 15 to 19 years; 39.2% (n = 169) of the in-school and 35.6% (n = 209) out of school were married to partners aged 20 to 24 years; and, 20.0% (n = 86) of in-school and 46.7% (n = 274) of out-of-school were married to partners aged 25 years and above.

The population’s ability to receive and synthesize information on matters that affect their own health was assessed by the indicator literacy levels. Literacy levels were determined by asking the AGYW to read a plain english or local language paragraph that was embedded into the questionnaire. Individuals were considered to be literate if they were able to read the paragraph with ease (with clear and correct pronunciation of words). A higher proportion 34.5% (n = 2061) of the in-school AGYW were able to read the text with ease compared to 21.2% (n = 1446) of their out-of-school counterparts. Similarly, a third 29.4% (n = 1755) of the in-school AGYW were able to read but with difficulty, compared to 18.3% (n = 1250) for the out-of-school. Furthermore, only 5.2% (n = 308) of the in-school AGYW could not read at all, and 3.7% (n = 254) among the out-of-school could not read at all. These findings have implications for the use of written messages particularly among out-of-school AGYW. It should be noted that more than half of the out-of-school and 31.0% (n = 1855) in-school AGYW had missing information or non-response for this indicator.

Slightly above half 52.2% (n = 3120) of the in-school AGYW reported living with both parents compared to only 26.1% (n = 1784) of their Out-of-school counterparts. Among in-school AGYW, 22.6% (n = 1352) lived with other relatives, 19.8% (1186) with mother only and 27.9% (n = 1673) lived with either a partner, father only, other relative, friend or alone. For out-of-school AGYW 36.2% (n = 2471) lived with a partner, 16.1% (n = 1103) with other relative, 15.6% (n = 1067) with mother only, and 6.0% (n = 408) with either a friend, father only or alone.

When asked if they had ever tested for HIV, 77.0% (n = 5260) of the out-of-school said yes compared to only 37.8% (n = 2262) among the in-school AGYW. Not surprisingly, only a few of the HIV tests were done within the 12 months preceding the survey, with 6.4% (n = 381) among the in-school and 16.3% (n = 1113) among the out-of-school. Interestingly, only a third 30.1% (n = 1797) of the in-school AGYW and half 50.6% (n = 3456) of the out-of-school AGYW had high comprehensive knowledge of HIV. Similarly, 40.9% (n = 2444) of the in-school AGYW and 24.2% (n = 1652) of the out-of-school counterparts had low comprehensive knowledge of HIV. The use of condoms was low with only a quarter 25.6% (n = 506) of in-school AGYW and 11.0% (n = 577) of out-of-school reported using condoms always. Only a third 33.4% (661) of in-school and a quarter 26.5% (1396) of the out-of-school AGYW reported using condoms sometimes with their partner.

The wealth tertile was used as a proxy of household level wealth. AGYW in the high tertile were 41.3% (n = 2467) among the in-school and 39.6% (n = 2707) among the out-of-school. Majority of the AGYW fell in the medium tertile with 58.3% (n = 34485) among the in-school and 60.2% (n = 4111) of the out-of-school AGYW. Vulnerability index was used to map out community risks, vulnerability, and susceptibility to GBV/Syphilis/HIV in the AGYW. The index is a tool that combines different indicators. High vulnerability index means a participant was very vulnerable in need of priority attention. Half 53.1% (n = 3177) of the in-school were classified as highly vulnerable compared to only 16.7% (n = 1141) among the out-of-school. Conversely, 83.0% (n = 5671) of the out-of-school AGYW were classified as having moderate vulnerability compared to 46.9% (n = 2802) among the in-school AGYW.

Sexual debut experiences among AGYW

The overall mean age at first sex among AGYW interviewed was 16.6 years (Table 6), 16.2 years for in-school and 16.8 years for out of school.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 6. Distribution of first-time sexual behaviour experiences among the participants, overall and by school status.

https://doi.org/10.1371/journal.pone.0294545.t006

Overall, most of the respondents had first time sex with either their boyfriend (80.4%) or husband (15.6%), with 2.4% of the in school reported to have had their sexual debut in marriage compared to 21.0% among out of school AGYW. Conversely, 91.6% of in-school reported to have had their sexual debut with boyfriends compared to 75.9% among the out of school AGYW. This means that overall, only 4% of respondents had their sexual debut with someone other than their boyfriend or husband. Table 6 also shows the age of the partner the AGYW respondents had their sexual debut with. Overall, partner age ranges 1–2 and 3–4 years older and AGYW interviewed, this age group constituted the majority of their partners at sexual debut (31.3% and 30.2%, respectively). The in school AGYW were more likely to have a partners of age range 1–2 years compared to their out of school counterparts (38.3% vs.28.5%, respectively) at sexual debut. Few respondents from both in-school and out of school (3.24% vs. 3.42%, respectively) had their partners who were 10 or more years at sexual debut.

Number of sexual partners

Self-reported number of sexual partners.

From Table 7, the AGYW who had ever had sex were asked about the number of sexual partners that they had sex with in the last 12 months and in their lifetime. For sex within the last 12 months, overall, the majority of the AGYW 72.9% (n = 4978) reported engaging in sex with one (1) sexual partner in the past 12 months, with minimal differences between the in-school 74.8% (n = 1415) and the out-of-school 72.2% (n = 3563). However, 17.3% (n = 1177) AGYW reported engaging in sex with 2+ partners in the past 12 months. By age-group, the proportion of AGYW reporting having one sexual partner in the past 12 months was decreasing slightly with increasing age. That is, 83.5% (n = 71) for the 10–14 years, 76.7% (n = 1426) for the 15 to 19 years, and 71.3% (n = 3481) for the 20 to 24 years. On the other hand, the proportion reporting 2+ partners in the past year increased with increasing age with 4.7% (n = 4) for the 10 to `4 years, 15.8% (n = 294) for the 15 to 19 years and 18% (n = 879) for the 20 to 24 years. Only 9.8% (n = 670) reported having no sexual partner in the past 12 months.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 7. Number of sexual partners by school status and age-group.

https://doi.org/10.1371/journal.pone.0294545.t007

Multiple sexual partnerships, condom use at last sex and consistent condom use by district of residence

Multiple sexual partnerships.

Table 8 shows the percentage of respondents that reported multiple sexual partners, condom use at last sex and consistent condom use by district of residence. The proportion of those reporting 2 or more sexual partners was highest in Kalomo (60.9%), Nyimba (39.6%) and Mwense (36.4%). Districts with the lowest proportion of AGYW reporting 2+ sexual partners in the past 12 months include Chililabombwe (6.5%), Mungwi (9.3%) and Mkushi (9.9%). There is a need to focus on high-risk districts (with the highest proportions reporting 2+ sexual partners) with targeted HIV prevention efforts to reduce AGYW’s vulnerability to HIV infection.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 8. Percentage of respondents that had sex in the past 12 months, and of these, percentage that reported multiple sexual partners, condom use at last sex and consistent condom use by district of residence.

https://doi.org/10.1371/journal.pone.0294545.t008

Prevalence and factors associated with hiv and syphilis

Overall prevalence of HIV and syphilis.

Out of 12, 457 participants the overall prevalence of HIV was 466 (3.7%) 95% CI: 3.4–4.1). The weighted prevalence of HIV was higher in the out-of-school compared to the compared to the in-school participants (5.5% vs 2.0%), Similarly, the prevalence of syphilis was higher in the out-of-school than the in-school participants (4.1% vs 1.5%). When the HIV prevalence was stratified by age group, highest prevalence was reported among those aged 20–24 years (5.8%) while the least was among those aged 10–14 years (0.7%). Likewise, highest prevalence of syphilis was reported among participants aged 20–24 years (4.7%) and lowest was among those aged 10–14 years (0.5%) as shown in Table 9.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 9. HIV and syphilis prevalence among AGYW, by school status and age-group.

https://doi.org/10.1371/journal.pone.0294545.t009

Prevalence of HIV and syphilis by district.

Fig 2 shows that Mkushi and Mungu districts had the highest HIV prevalence of 641 (6.7%) and 640 (6.7%)% respectively, followed by Mwense 691 (5.6%), Kabwe 655 (5.2%) in that order. Districts that reported least prevalence of HIV were Chinsali 705 (1.8%), Chembe 701 (1.9%), Mungwi (1.9%) and Mwinilunga 723 (2.1%). For syphilis, the districts that reported highest prevalence were Mkushi 641 (7.3%), Mwense 691 (5.9%), Mwinilunga 723 (5.5%) and Chembe 701 (5.4%). Least prevalence of syphilis was reported from Ndola 652 (1.2%), Chililabombwe 636 (1.4%), Limulunga 639 (1.4%) and Chisamba 703 (1.65%). The overall prevalence of syphilis was 12, 813 (3.2%).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 2. Prevalence of HIV and syphilis by districts.

https://doi.org/10.1371/journal.pone.0294545.g002

Factors associated with HIV infection.

In the bivariable logistic regression analysis for factors associated with HIV infection among AGYW in selected districts of Zambia: age, ever married, school status, ever had sex, knowledge of HIV prevention and vulnerability were associated with HIV infection. However, in the multivariable regression, only age, ever had sex and school status were significantly associated with HIV infection. AGYW who were aged 15–19 years were 2.25 (1.17–4.31) times the odds of HIV infection compared to their counterparts who were aged 10–14 years. Similarly, those who were aged 20–24 years were 3.67 (1.89–7.11) times the odds of HIV infection compared to those who were aged 10–14 years. Being out-school was associated with 1.40 (1.10–1.78) times the odds of HIV infection compared to those who were in-school. Also, this study examined that the likelihood of having HIV infection was 1.84 fold (1.34–2.51) higher among AGYW who reported ‘ever to have sex’ compared to those who had ‘never experienced sex’ (Table 10).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 10. Bivariable and multivariable logistic regression analysis for factors associated with HIV infection among AGYW in selected districts of Zambia, 2021.

https://doi.org/10.1371/journal.pone.0294545.t010

Factors associated with syphilis infection.

Regarding syphilis infection, the bivariable logistic regression analysis showed age, ever married, school status, ever had sex, knowledge of HIV prevention and vulnerability were associated with syphilis infection.

However, in the multivariable regression only, age, ever married and ever had sex were significantly associated with syphilis infection. Participants who were aged 20–24 years were 3.11 (95% CI,1.41–6.73) times greater odds of having syphilis infection compared to those aged 10–14 years. For marriage, those who reported that they were ever married were 1.26 (95% CI, 1.01–1.63) times the odds of having syphilis infection compared to those who reported that they were never married. Similarly, participants who indicated that they had ever experienced sexual intercourse were 3.41(95% CI, 2.23–5.38) times the odds of syphilis infection compared to their counterparts who were reported not to have ever experienced sexual intercourse (Table 11).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 11. Bivariable and multivariable logistic regression analysis for factors associated with syphilis infection among AGYW in selected districts of Zambia, 2021.

https://doi.org/10.1371/journal.pone.0294545.t011

Behavioural characteristics of AGYW

Our findings in this study suggest that several sexual and health behavioural challenges and needs of AGYW need to be addressed by different players at different levels of society. It’s important to note that in some of the behavioural aspects reported in this study the out-of-school AGYW are much more disadvantaged than their in-school AGYW. For example, the proportion of AGYW that had never had sex were 87.4% among the in school and 40.7% among the out of school AGYW. Also concerning willingness to have sex at the first time, 71.9% out of school compared to 64.2% in school AGYW were very willing to have sex. This finding is in the same direction as found by Matovu et al., (2021) [11]. In addition, 40.2% out-of-school AGWY reported having sex within a week before the survey compared to only 13.4% of the in-school. Also, concerning use of condoms, 45.9% in-school AGYW used a condom the last time they had sex compared to 23.6% of the out-of-school. However, consistent use of condoms in the past 12 months was very low in both groups; 10.1% and 11.2% in-school and out-of-school AGYW respectively. Concerted effort is needed on the out-of-school AGYW since they seem to be much more disadvantaged than the in-school AGYW. Stakeholders at different levels who are currently working with AGYW need to continue with and increase the STI behavioural education that promote behavioural change in this particular group, especially among the out-of-school AGYW.

Prevalence of HIV and syphilis and associated factors

The overall HIV prevalence was 3.7%, with a higher prevalence among out-of-school adolescents (as compared to the in-school adolescents (5.5%) compared to the in-school adolescents (2.0%). This finding is consistent with other studies [8, 12, 28–30]. Similarly, the prevalence of syphilis was higher in the out-of school than the in-school participants (4.1% vs 1.5%). The study also revealed some gaps in knowledge that existed among the AGYW. For example, 23% of AGYW did not know that a healthy-looking person can have HIV; 40% did not know that HIV can be transmitted from mother to child during pregnancy; and 47% thought if a man/woman had HIV the partner will also have HIV. Therefore, there is a need to scale up efforts that target AGYW especially the out of school AGYW with interventions to address HIV and Syphilis amongst this group. Health promotion and education interventions can be used to enhance targeted messages to address knowledge gaps about HIV. In addition, proven high impact biomedical interventions such as PreP remain key to addressing the HIV risk at individual level for the AGYWs.

Vulnerability levels of households and individual AGYW

Findings show that at individual level the majority of out-of-school girls (71.6%) were categorized as highly vulnerable compared to the in-school girls (69.7%). Having ever given birth, first sex before the age of 15 years and having multiple sexual partners were indicative of very high vulnerability among the girls. At household level, all indicators of vulnerability had high proportions of girls in the high tertile. Community vulnerability showed a similar trend. Drivers of vulnerability include living with parents, pregnant or given birth, having sex before 15 years and with multiple sexual partners, at individual level. At household level, the head of the household without education and lack of access to improved sanitation or water were indicative of high vulnerability. Similarly, at community level AGYW who were HIV positive, and knowledge about HIV were indicative of high vulnerability. The high levels of individual, household and community-level vulnerabilities suggest that interventions are needed to reduce the levels of vulnerabilities.

The vulnerabilities of AGYW observed also contributes to the observed high prevalence in AGYW compared to their men counterpart. In addition, girls’ intimate relationships and sexual debut is earlier than their male peers; and secondly, they have partners older than them, some of whom emanate from age groups/brackets with higher HIV incidence. The power dynamics in these relationships usually leave the female folk vulnerable, especially amidst questionable condom use and/or its absence and lack of other personalized, and female tailored HIV prevention interventions such as the Cabotegravir injectable and Dapivirine ring, respectively. Both of which are only being introduced now. There is also a biological part which makes it easier for HIV acquisition in these young females. Of course, in our setting, there is the issue of Sexual Gender Based Violence (SGBV) which is a big problem in Zambia. Our study showed that all these factors contributed to the fact that HIV prevalence was high among AGYW compared to men.

Sexual reproductive services in Zambia

The SRH services available in Zambia are: Family planning (contraceptive) services, antenatal, postnatal, delivery, safe and post abortion care, cervical cancer screening, HIV testing and treatment services, Human Papilloma Virus Vaccination for pre and adolescent girls aged between 9 and 14, menstrual hygiene and management. These services are available and accessible at different levels of the health care system from primary to tertiary level and ranging from information only, counselling, and actual services.

However, the SRH services availability can be improved by ensuring availability of SRH commodities and equipment, trained personnel, and right infrastructure for provision of comprehensive SRH services.