Cost–benefit analysis of adolescent HIV/STI prevention and comprehensive sexuality education interventions in Cambodia: a modelling study

Abstract

Background

Among youth in Cambodia, there is low awareness of HIV and sexually transmitted infection (STI) risk, and the prevalence of unintended pregnancies is high. We aimed to quantify the health and economic benefits of reaching the 2025 Joint United Nations Programme on HIV/AIDS (UNAIDS) target of 95% coverage of combination HIV prevention/testing packages for young key population groups as well as providing comprehensive sexuality education (CSE) for all youth.

Methods

The model considered 10–19-year olds in Cambodia, disaggregated by key population group (young men who have sex with men, young transgender persons, young people who use drugs, young female entertainment workers) and in-school/out-of-school status. Three scenarios were projected over 2023–2030: (1) a baseline scenario with fixed intervention coverage; (2) a scenario scaling up existing HIV/STI prevention and testing interventions for young key populations through multiple delivery modalities, plus in-school CSE, over 2024–2025 to reach 95% coverage (maintained until 2030); and (3) scenario 2 plus an additional out-of-school CSE programme for all being implemented and scaled to 95% coverage. HIV infections, unintended pregnancies, maternal deaths and stillbirths averted were converted to economic benefits.

Results

Scenario 2 costs US$10.6 million more than the baseline scenario over 2023–2030, and averted 1184 HIV infections, 10 172 unintended pregnancies, 16 maternal deaths and 106 stillbirths. This investment could generate US$36.1 million in economic benefits by 2050, with a benefit–cost ratio of 3.4. Scenario 3, including out-of-school CSE, achieved additional health benefits; however, it was less cost-effective than in-school CSE.

Conclusion

Investment in HIV prevention/testing packages for young key population groups and CSE programmes can have favourable returns in Cambodia. Strategies to maintain high levels of school enrolment can facilitate cost-effective ways to expand CSE coverage.

WHAT IS ALREADY KNOWN ON THIS TOPIC.

WHAT THIS STUDY ADDS

• Investments in HIV prevention/testing interventions for young key populations combined with comprehensive sexuality education programmes can generate significant health and economic benefits in Cambodia.

HOW THIS STUDY MIGHT AFFECT RESEARCH, PRACTICE OR POLICY

• Quality implementation of programmes through sustained financing avenues, improved education given to outreach workers, availability of sufficiently trained teachers and adequate and equitable supply of modern contraceptive commodities will be required to realise the benefits of these investments.

Introduction

Significant progress has been made in Cambodia towards the Joint United Nations Programme on HIV/AIDS (UNAIDS) FastTrack 95-95-95 targets, with an estimated 86% of people living with HIV diagnosed,¹ around 98% of people diagnosed on treatment, 98% of people on treatment virally suppressed in 2022, and there has been a 33% decline in new HIV infections since 2010.2,4 Despite the progress, Cambodia is still experiencing a concentrated HIV epidemic and, in 2022, there were an estimated 76 000 (63 000–83 000) people living with HIV and 1400 new HIV infections.³ Key populations disproportionately affected by HIV include men who have sex with men (3.6% prevalence in 2019)⁵; transgender persons (9.6% HIV prevalence in 2019)⁵; female entertainment workers (defined in the HIV Estimates Technical Working Group 2022 as ‘Non-freelance entertainment workers aged 15–19 years who work permanently or impermanently in massage parlour, beer garden, pub, clubs and other places for recent year with change in sex work dynamics due to COVID-19 and increased use of social media to find partners. Modelled as including females who inject drugs and females who use drugs as part of sexualised drug use, who are also entertainment workers, and potentially including sexually exploited children aged under 18 years (online supplemental appendix A)) (2.9% HIV prevalence in 2019)⁶; and people who inject drugs (12.8% HIV prevalence in 2018).⁷ A major concern is decreasing awareness of HIV and sexually transmitted infection (STI) risk among youth, leading to lower rates of testing and condom use. For example, while 86% of people living with HIV nationally are diagnosed, diagnosis among people aged 15–24 years is estimated to be only 66%, and as low as 20% among 15–19-year-old men who have sex with men and transgender persons.⁸ Recent HIV analyses have suggested that more focus on differentiated outreach to these younger key populations is needed.⁸

Along with HIV/STI risks for key populations, adolescents in Cambodia are facing sexual and reproductive health concerns such as high rates of unintended pregnancies. Among 15–19-year olds in Cambodia in 2019, there were 52.9 births per 1000 girls (higher than the global average of 42)⁹ and a 19% unmet need for family planning.^{10 11} The central policy framework focused on the sexual and reproductive health of adolescents and youth in Cambodia is the Cambodia Ministry of Health (MoH) National Strategy for Sexual and Reproductive Health in Cambodia 2017–2020,¹² which includes comprehensive sexuality education (CSE) and defines related indicators and goals for monitoring progress in time. The aim of this CSE programme is to provide adolescents in Cambodia with an objective, rights-focused education on sexuality, as well as knowledge on gender equality and reproductive health.¹³ The CSE curriculum is implemented by the Ministry of Education, Youth and Sport in Cambodia as a compulsory subject for students in grades 5–12 (ages 9 or 10 up to 18 years). CSE delivery is in partnership with non-government organisations, and while there is currently no formal ‘out-of-school’ CSE programme, digital tools are used to improve engagement and access.¹³ Even though the in-school CSE programme has been started in Cambodia, the coverage is still not satisfactory,¹⁴ and with low secondary school enrolment rates, there is a gap in coverage for young people who are no longer in school.¹¹

CSE programmes have been shown to increase demand for sexual and reproductive health services among adolescents, leading to reduced unmet need for contraception, reduced unintended pregnancies and increased condom use and reduced risks of HIV/STIs.15,17 For example, previous studies in South Africa found that adolescent girls exposed to CSE classes in the previous 12 months, either in-school or out-of-school, were more likely to get tested for HIV¹⁶; and in the USA that funding for additional CSE programmes reduced country level births to adolescents by more than 3%.¹⁷ However, sexual maturity during adolescence can coincide with entry into key population groups, and these studies did not consider the combined impact of CSE when implemented alongside HIV prevention/testing programmes for adolescent key population groups. Moreover, these studies did not examine the cost-effectiveness and long-term economic benefits of CSE interventions.

The objective of this modelling study is to quantify the health and economic benefits in Cambodia of reaching the 2025 UNAIDS target of 95% coverage of combination HIV prevention/testing packages, including CSE, for young key population groups; and providing in-school CSE to all enrolled students and achieving 95% coverage of out-of-school CSE for all. This study was conducted in partnership with Cambodia United Nations Population Fund (UNFPA), and findings of this study can be used to inform evidence-based resource allocation decisions, particularly for increasing equity among those who are not adequately reached by existing HIV-related services.

Methodology

Model population

The model considers the population of adolescents aged 10–19 years over time, disaggregated based on HIV risk and potential to be reached by different programmes (online supplemental appendix A: table A.1, A.2 and figure A.1). Specifically, key population groups for 15–19-year-old men who have sex with men, transgender persons, female entertainment workers and people who inject drugs, as well as non-key population groups for 10–14-year olds in school/out-of-school, 15–17-year olds in school/out-of-school and the remainder of the population.

Each population group was defined by the following characteristics: population size (current and future projections collated from Cambodia national census report 2019 and AIDS Epidemic Model (AEM) modelling study)^{4 18}; prevalence of HIV and STIs⁴; testing rates for HIV and STIs⁴; rates of condom use⁴ and rates of unintended pregnancies (Cambodia Demographic and Health Survey 2021 and Cambodia youth situation analysis report 2020).^{10 11}

Further details and model inputs are provided in online supplemental appendix B (table B.1, B.2, B.3) and C (table C.1, C.2).

Interventions

Several combinations of HIV/STI prevention and testing interventions that included some form of CSE were modelled for each young key population group (aged 15–19 years), based on outreach through differentiated service delivery modalities: physical outreach (eg, venues), virtual outreach and mobile outreach (online supplemental appendix D: table D.1). Interventions and delivery modalities were based on those currently in operation and were parametrised by 2023 programmatic data.⁸

Each young key population group is assumed to be reached by a different combination of delivery modalities (determined through discussions with country teams), and each delivery modality can have different population group-specific effect sizes. For example, virtual outreach for young men who have sex with men may have different effects to virtual outreach for young female entertainment workers; and physical outreach may achieve higher coverage for young female entertainment workers than for young men who have sex with men.

Each intervention (ie, population group and delivery modality combination) was defined by estimates for baseline coverage; maximum achievable coverage; unit cost (cost per person reached per year)19,21; and effect sizes for increased condom use and increased STI/HIV testing.1522,31 The specific parameters and sources for each intervention are in online supplemental appendix D (table D.2, table D.3).

In addition, CSE for all was considered as a way of increasing condom and contraception use, with in-school (ie, Ministry of Education age-sequenced curriculum for 10–17 year-olds) and out-of-school delivery modalities (for 15–17-year olds). The prospective out-of-school CSE programme is not yet implemented in Cambodia. These CSE programmes were defined by estimates for baseline coverage; maximum coverage; unit cost (cost per person reached per year)19,21; and effect sizes for increased condom use and reduced unmet need for family planning.1522,31

Modelling framework

In this model, scaling up HIV/STI prevention and testing intervention coverage among young key populations increases the condom use and testing. This leads to reduced HIV infections, resulting in reduced treatment costs and disability-adjusted life years (DALYs). Scaling up CSE programmes (both in-school and out-of-school) increases condom use and modern contraception use. This leads to reduced unintended pregnancies, which in turn generates economic benefits through direct maternal health costs averted, increased education and workforce and social benefits (figure 1).

Figure 1. Modelled relationships between interventions and economic benefits. Top: relationship between (a) HIV/STI prevention and testing intervention coverage; (b) increased condom use and testing; (c) reduced HIV/STI infections; and (d) reduced treatment costs and disability-adjusted life years (DALYs). Bottom: relationship between (a) coverage of comprehensive sexuality education (CSE; in-school and out-of-school programmes); (b) increased condom and contraception use; (c) reduced adolescent unintended pregnancies; and (d) economic benefits through direct costs averted, increased education and workforce and social benefits. GDP, gross domestic product; STI, sexually transmitted infection.

Scenarios

Several scenarios were considered for the period 2023–2030:

1. Baseline. Assuming no changes to intervention coverages and behaviours.

2. Scale-up of existing programmes (young key population programmes+in-school CSE). By 2025, achieving:

   1. 95% coverage of combination HIV prevention/testing packages and out-of-school CSE for young key population groups (aged 15–19 years).

   2. Providing in-school CSE to enrolled 10–17-year olds (to reach enrolment rate of 46.5%) to increase contraceptive use (including condoms) and reduce unmet need for family planning.

3. Full scale-up. Scale-up of existing programmes plus by 2025 achieving 95% coverage of out-of-school CSE programmes for all out-of-school 15–17-year olds to increase contraceptive use (including condoms) and reduce unmet need for family planning.

Interventions and their delivery modalities start at baseline levels in 2023 (figure 2) and are scaled up over 2024–2025 and then maintained until 2030.

Figure 2. Baseline and scale-up scenario coverage of interventions. Key populations aged 15–19 years are reached by different modalities for HIV/STI prevention and testing, enrolled 10–17 year-olds can be reached by in-school CSE programmes, and all out-of-school 15–17 year-olds can be reached by out-of-school CSE programmes. Scale-up values are based on saturation values estimated for adult key populations in the ‘Focusing and sustaining the HIV response in Cambodia 2023’⁸ study, but there may be challenges in achieving such high coverage of programmes for young populations that would need to be overcome. CSE, comprehensive sexuality education; FEW, female entertainment workers; MSM, men who have sex with men; PWID, people who inject drugs; STI, sexually transmitted infection; TG, transgender persons.

Costs

Costs of each scenario were estimated from a government/health provider perspective over 2023–2030, and include annual costs of different outreach delivery modalities for key populations aged 15–19 years (including service delivery costs, condoms and testing packages taken from the Focusing and sustaining the HIV response in Cambodia 2023 report),⁸ costs of in-school CSE programmes for 10–17-year olds and out-of-school CSE programmes for 15–17-year olds, and the costs of contraception use among 15–19-year olds (including logistics, overheads and wastage). To estimate the worth of future costs as of the base year, discounting was used at a rate of 3% per annum. Costs are presented in 2023 US$.

Health outcomes

Health outcomes for each scenario were estimated including unintended pregnancies, maternal deaths, stillbirths, condom use and HIV/STI testing among 15–19-year olds over 2023–2030. Increases in condom use and HIV/STI testing rates among young key population groups were converted to HIV infections averted using the Cambodia Optima HIV model (online supplemental appendix E: figure E.1).⁸

Economic benefits

Total economic benefits were calculated based on the difference in outcomes between scale-up and baseline scenarios. For young key population interventions, these are direct HIV treatment costs averted and DALYs averted (converted to economic benefits based on value of a statistical life year). For CSE interventions, these are direct maternal health costs averted; DALYs averted from maternal mortality (a small percentage of unintended pregnancies); for unintended pregnancies averted among girls 15–17 years, increased schooling years leading to increased earnings on entering the workforce (assumed 18 years, adjusted for female participation rates); for unintended pregnancies averted among women 18–19 years (as a result of receiving CSE before leaving school), increased workforce participation due to no longer requiring maternity leave. Additional details are in online supplemental appendix F (table F.1).

Benefits were calculated for those receiving the interventions over 2023–2030, including direct maternal health costs averted up to 2030 and other benefits up to 2050.

Sensitivity analyses

Three sensitivity analyses were run to explore the assumptions in calculation of the benefit–cost ratio: (a) discounting rate of 3% per annum versus 1% or 5% per annum (b) 1.5 years of education gain from unintended pregnancy averted versus 1 year gain or 2 years gain (c) 0.5 times the gross domestic product (GDP) per capita for statistical value of a life year versus 0.25 times or 1.0 times.

Results

Baseline scenario

Maintaining baseline coverage of interventions over 2023–2030 was estimated to cost US$23.8 million (discounted), comprising US$2.6 million for out-of-school young key population interventions (15–19 years) through different delivery modalities, and US$21.1 million for in-school CSE programmes and contraception.

Scale-up of existing programmes

Scaling up to achieve 95% coverage of out-of-school young key population interventions and in-school CSE to reach maximum school enrolment rate (46.5%) by 2025 required an additional US$10.6 million (discounted) over 2023–2030 (figure 3).

Figure 3. Annual cost of combination HIV prevention/testing packages for young key population groups (15–19-year olds), in-school CSE programmes for 10–17-year olds, and contraception for 15–19-year olds (scenario 2). Programme coverages were scaled up starting from 2023 to reach maximum reachable coverage by 2025, maintained to 2030. Total costs are in 2023 US$, discounted at 3% per annum and include overhead logistics and wastage costs. Note that annual costs decline over time, even with fixed programme coverage (%), due to projected decreasing adolescent population size. CSE, comprehensive sexuality education; FEW, female entertainment workers; MSM, men who have sex with men; PWID, people who inject drugs; TG, transgender persons.

Scaling-up interventions to achieve 95% coverage by 2025 could have significant health benefits for 15–19-year old key population groups. HIV testing increased by 13.7 percentage points among young men who have sex with men, 0.5 percentage points among young female entertainment workers, 15.1 percentage points among young transgender persons and 4.6 percentage points among young people who inject drugs. Condom use increased by 16 percentage points among young men who have sex with men, 1 percentage point among young female entertainment workers, 14 percentage points among young transgender persons and 18 percentage points among young people who inject drugs (figure 4). These increases in HIV testing and condom use could avert 1184 HIV infections over 2023–2030 (estimated using Optima HIV Cambodia model; details in online supplemental appendix E).⁸ In addition, the model estimated that 10 172 unintended pregnancies, 16 maternal deaths and 106 stillbirths could be averted among women and girls aged 15–19 years over 2023–2030 (based on estimated maternal mortality ratio of 154 per 100 000 live births and 10.4 stillbirths per 1000 births).¹⁰

Figure 4. Health outcomes among adolescents for the intervention scale-up scenario (scenario 2). Top left: condom use prevalence among key population groups aged 15–19 years. Top right: HIV testing rates among key population groups aged 15–19 years. Bottom left: STIs testing rates among key population groups aged 15–19 years. Bottom right: annual unintended pregnancies for baseline and scale-up scenarios among 15–19-year olds. FEW, female entertainment workers; MSM, men who have sex with men; PWID, people who inject drugs; STI, sexually transmitted infection; TG, transgender persons.

With 3% annual discounting, the scale-up of existing programmes was estimated to lead to US$36.1 million in economic benefits by 2050 due to unintended pregnancies averted, maternal deaths and stillbirths averted, and HIV infections averted in adolescents. This gave the intervention scale-up (scenario 2) a benefit–cost ratio of 3.4 by 2050. That is, for every $1 invested, there could be $3.4 worth of benefit to the economy.

Full scale-up scenario

In addition to scaling up the existing programmes, achieving 95% coverage of a prospective out-of-school CSE programme for all was estimated to cost an extra US$39.8 million over 2023–2030 (online supplemental appendix G: figure G.1), but could avert an additional 23 250 unintended pregnancies, 37 maternal deaths and 242 stillbirths among adolescent girls. Based on these health impacts, the out-of-school CSE programme was estimated to generate an additional US$76.1 million in economic benefits by 2050 with a benefit–cost ratio of 1.9 (table 1).

Table 1. Costs and economic benefits for the baseline and intervention scale-up scenario to reach the 2025 UNAIDS target of 95% coverage of combination HIV prevention/testing packages for young key populations, in-school CSE programmes for 10–17-year olds and out-of-school CSE for 15–17-year olds in Cambodia (2023–2030).

	Baseline cost 2023–2030	Scale-up cost 2023–2030	Additional cost 2023–2030	Economic benefits 2023–2050	Benefit–cost ratio (BCR)
Interventions for men who have sex with men aged 15–19 years	$1 215 000	$2 456 000	$1 241 000	$2 851 151	1.7
Interventions for transgender persons aged 15–19 years	$205 000	$415 000	$210 000
Interventions for female entertainment workers aged 15–19 years	$1 148 000	$1 268 000	$120 000
Interventions for people who inject drugs aged 15–19 years	$50 000	$177 000	$127 000
In-school CSE interventions for 10–17-year olds+contraception use of 15–19-year olds	$21 140 000	$29 113 000	$8 860 800	$33 283 292	3.8
Out-of-school CSE interventions for 15–17-year olds+additional contraception use among 15–19-year olds	$0	$39 823 000	$39 823 000	$76 071 002	1.9

Costs and benefits are in 2023 US$, discounted at 3% per annum.

Economic benefits and the benefit–cost ratio for key population interventions are calculated as a combined package, as epidemiological impacts were modelled in aggregate.

CSE, comprehensive sexuality education; UNAIDS, United Nations Programme on HIV/AIDS.

Results of the sensitivity analyses

The benefit–cost ratio of the existing programme scale-up scenario (scenario 2; benefit–cost ratio of 3.4 in the main analysis): increased to 4.4 when 1% per annum discounting was used; decreased to 2.7 if 5% per annum discounting was used; decreased to 2.6 if 1 year of education gain per unintended pregnancy averted was assumed; increased to 4.3 if 2 years of education gain per unintended pregnancy averted was assumed; remained around 3.4 if 0.25 times GDP per capita was used for statistical value of a life year; increased to 4.3 if 1.0 times GDP per capita for statistical value of a life year was used (online supplemental appendix G: figure G.2).

Discussion

This study found that in Cambodia, scaling up HIV prevention/testing packages for young key populations to reach 95% coverage and school-based CSE programmes to reach maximum enrolment rate by 2025 could cost US$10.6 million more than the baseline scenario over 2023–2030, but could avert 1184 HIV infections, 10 172 unintended pregnancies, 16 maternal deaths and 106 stillbirths. This investment could generate US$36.1 million in economic benefits by 2050, with an overall benefit–cost ratio of 3.4.

Achieving 95% coverage of combination, HIV prevention/testing packages for young key population groups could be challenging in practice. Young people generally prefer to access sexual health services through specialised, youth-favourable locations and a focus on differentiated outreach through trusted access platforms (a trusted access platform helps all key populations access services, accommodating universal needs regardless of HIV status or other specific healthcare needs)¹¹ will be necessary to reduce stigma and create an enabling environment. It is possible that this will require new modalities or modification to those currently being delivered, and this may require additional investments, which will have both additional costs and additional benefits not captured in this study.

The scale-up scenario in our study assumed that analogous outreach programmes designed and tailored for young key populations would have similar cost and effectiveness to the versions currently being delivered for adult key populations. This is because data are limited, which prevented us from differentiating their actual cost and effectiveness among 15–19-year olds compared with among all adults. Also, age-specific challenges faced by young people compared with adults, such as legal dependency and lack of sexuality education, could result in the different levels of effectiveness of HIV prevention/testing interventions.^{32 33} Any such differences would impact the total costs, estimated health outcomes and economic benefits. In addition, youth-focused programmes might initially be more expensive due to the need for more intensive trust-building and outreach.³² However, such programmes could have the potential to establish lifelong health-seeking behaviours; therefore, it could be possible that the long-term benefit–cost ratio could be even higher than estimated.

Health impacts from scaling up interventions, such as estimated increase in HIV/STI testing and condom use, were different among different young key population groups. This is due to differences in baseline testing rates and condom use, baseline intervention coverages, the extent to which intervention coverage could be increased and measures of intervention effectiveness for each group. For example, young female entertainment workers had a relatively higher baseline coverage of the physical outreach modality (69%) compared with young men who have sex with men (26%), and so reduced the capacity to scale up further. In addition, the physical outreach modality was estimated to lead to a 20 percentage point increase in condom use among young female entertainment workers compared with a 40 percentage point increase among young men who have sex with men (online supplemental appendix D: table D.2).

With an estimated ~4 35 000 15–17-year olds out of school, there were high costs ($14.56 per person) associated with reaching 95% coverage of the unfocused out-of-school CSE programme. While this prospective programme was estimated to have a benefit–cost ratio of greater than 1, it had a lower cost-effectiveness than the in-school programme, highlighting the importance of maintaining a high level of school enrolment. The benefit–cost ratio would be improved if the programme was directed towards higher risk populations (eg, low socioeconomic status, external and internal migrants, indigenous communities and young key populations) rather than extending the coverage uniformly across all adolescents. Also, there may be additional benefits of the out-of-school CSE programme not captured in this analysis; for example, the scale-up scenario includes highly aspirational coverages of interventions directed towards young key populations, and out-of-school CSE for all may provide a necessary platform for achieving these targets.

The benefits of the CSE interventions are likely to have been underestimated in this analysis. The effects of CSE in the model are to increase contraception use (including condoms) and reduce unmet need among people aged 15–17 years, which results in unintended pregnancies averted and associated economic benefits. However, CSE for adolescents has additional time-lagged benefits, including for HIV/STI prevention among key populations, which are not captured in this study because quantitative effect sizes are limited. For example, by increasing sexual health awareness among young people before they enter key populations (eg, among 10–14 year-olds), CSE has the potential to reduce risk-taking, increase condom use and increase testing for key populations aged 15–19 years. Similarly, CSE for 10–14-year olds in school can increase knowledge among 15–19-year olds who have dropped out of school, leading to increased contraception prevalence use, reduced STI transmission and further benefits among this harder to reach group. Because these benefits could not be captured due to limited data, this means CSE is likely to have a higher benefit–cost ratio than estimated. Benefits relating to STI cases averted were not modelled and these provide potential further improved health outcomes and reduced treatment costs.

There are some important limitations to this study. First, exact numbers for the 15–19-year-old HIV key population sizes were not available; therefore, they were prorated by age from total key population sizes. Second, epidemiological model inputs for 15–19-year-old key populations were estimated largely from surveys of 18–24-year-olds. Also, baseline coverage estimates for interventions were available for key population groups in total (ie, all ages) but had to be estimated for 15–19-year-old key population groups. Third, unit costs for key population interventions were estimated using a programme experience approach (total spending/number of people reached). It is not clear whether these would be higher or lower for young people compared with an all-age average. Fourth, unit costs for modern contraception methods were estimated using an ingredient-based approach (ie, each element needed for the method is identified and priced)³⁴ and are unvalidated. Fifth, the effect sizes of key population interventions were based on programme data for the proportion of people reached who were tested and reported using condoms; however, people may still access testing and condoms through private means and so this may not fully capture the benefits of the interventions. Sixth, the effect of the CSE interventions on contraception prevalence use was based on global literature, and the actual impact will depend on the quality of education and access/availability of contraceptives that is implemented in Cambodia, which is difficult to measure. Seventh, benefits relating to STI cases averted could not be modelled, and nor could uptake of other interventions (eg, PrEP) as a result of CSE, and thus these estimated economic benefits are conservative.

This study is, to our knowledge, the first to quantify the health outcomes, costs and economic benefit of scaling up HIV/STI prevention and testing interventions for adolescent key populations alongside CSE programmes. Both in-school and out-of-school CSE programmes were found to have a positive return on investment; however, a larger unit cost for out-of-school CSE meant it was less cost-effective than in-school CSE. This supports other research highlighting the benefits of maintaining high school enrolment rates, making it possible for students to attend the in-school CSE programmes.¹⁶

Conclusion and next steps

Investments in HIV prevention and testing interventions for young key populations combined with CSE programmes for adolescents can generate favourable health and economic benefits in Cambodia. To achieve the benefits of these investments in practice will require collaborative actions to maintain high school enrolment rates, quality implementation of programmes, relevant education, a sufficient number of trained teachers and outreach workers to deliver programmes, equitable coverage across the country as well as adequate procurement and last-mile distribution of preventive commodities such as condoms and other modern contraceptives.

UNFPA alongside its partners in the Ministry of Health and Education Youth and Sports are exploring ways to further finance the full rollout of the CSE curriculum. UNFPA is also working with UNICEF, WHO, UNDP and UNAIDS towards a range of financing for health and education options including sustainable financing mechanisms through taxation, blended financing, public private partnership and other modalities.

Data availability statement

All data relevant to the study are included in the article or uploaded as supplementary information.