Evidence gaps in economic evaluations of HIV interventions targeting young people: a systematic review

Abstract

Introduction

Young people living with HIV (YPLWH) face the burden of navigating the unique physical, psychological, and social challenges of adolescence while coping with a stigmatized infectious disease that requires lifelong care. Consequently, YPLWH experience worse HIV outcomes compared to children and adults. This systematic review seeks to collate evidence on the health and economic impact of HIV interventions targeting YPLWH and to identify gaps in the available evidence that may inform future economic evaluations of interventions for YPLWH.

Methods

We searched the MEDLINE, Embase, Scopus, and Global Index Medicus databases for peer-reviewed articles published through April 24, 2022 (PROSPERO ID: CRD42022356244). Our inclusion criteria encompassed economic evaluations of HIV interventions that report health and economic outcomes among individuals ages 10 to 24 years. Three investigators screened articles at the title, abstract, and full-text levels. Data was extracted in accordance with the Consolidated Health Economic Evaluation Reporting Standards 2022.

Results

Of the 3,735 unique articles retrieved through our search, 32 met our inclusion criteria. Of these 32 articles, 8 (25%) evaluated a behavioral, educational, or financial intervention, 6 (19%) voluntary medical male circumcision (VMMC), 5 (16%) HIV screening or testing, 4 (13%) pre-exposure prophylaxis, 3 (9%) a theoretical HIV vaccine, 2 (6%) antiretroviral therapy, 1 (3%) condom distribution, and 3 (9%) a combination of interventions. Twenty-two studies (69%) focused on Africa, 9 (28%) on North America, and 1 (3%) on Europe. Thirty studies (94%) were cost-effectiveness analyses and 2 (6%) were cost-utility analyses. Of the intervention types captured by this review, most were deemed cost-saving or cost-effective. Only two studies—one evaluating a financial intervention and one evaluating HIV testing—concluded that the intervention was not cost-effective.

Conclusions

Evidence presented by this review suggests that investments in HIV prevention and treatment for young people can be a cost-effective, and sometimes cost-saving, solution to combatting the global HIV epidemic. However, additional evaluations of HIV interventions targeting young people, and which adhere to standardized reporting practices, are needed to permit comparability of cost-effectiveness outcomes between interventions and settings.

INTRODUCTION

In 2014 the Joint United Nations Program on HIV/AIDS (UNAIDS) published the 95–95-95 by 2030 targets.¹ These global targets comprise the following achievements by the year 2030: 95% of all people living with HIV diagnosed, 95% of all people diagnosed with HIV on antiretroviral therapy, and 95% of all people diagnosed with HIV, who are receiving antiretroviral therapy, virally suppressed. Substantial progress has been made over the last decade in achieving these targets. UNAIDS reports that in 2021, of all people living with HIV worldwide, 85% knew their status, 75% of those who knew their status were accessing treatment, and 68% of those who were accessing treatment were virally suppressed.² However, despite this progress, millions of more people living with HIV must be diagnosed and placed on treatment before the 95–95-95 targets are met.

Young people living with HIV are one sub-group of the global population that must be considered if the 95–95-95 targets are to be achieved and sustained. Most recent estimates from the Institute for Health Metrics and Evaluation (IHME) show that young people aged 10 to 24 years comprise 10.3% of all people living with HIV, but account for 21.2% of new HIV cases annually.³ Not only do young people represent a large, and growing, proportion of the global HIV epidemic, but they also face the additional burden of navigating the unique physical, psychological, and social challenges of adolescence while coping with a stigmatized infectious disease that requires lifelong care.⁴ Such challenges may lead to lower rates of treatment adherence and viral suppression as well as higher rates of disease transmission in comparison to adults.⁵ Consequently, tailoring and scaling-up HIV interventions for young people living with HIV is crucial to eliminating the global HIV epidemic.

Health economic evaluations are an essential tool that can help guide the allocation of resources to minimize costs and optimize health outcomes. While numerous economic evaluations have been conducted on the costs and benefits of scaling-up HIV interventions to various whole and sub-populations, few have focused on the scale-up of interventions to young people living with HIV. The purpose of this study was to systematically review published model-based economic evaluations of the application of HIV interventions to young populations. In doing so, our aim was to collate existing evidence on the cost-effectiveness of HIV interventions when applied to young people and to identify gaps in this evidence that can be used to inform future economic evaluations of HIV interventions for young people living with or at-risk of contracting HIV.

METHODS

Protocol

This systematic review sought to identify published economic evaluations on interventions targeting young people living with HIV. This systematic review was conducted in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines.⁶ A completed 2020 PRISMA checklist can be found in the appendix pp 1–3. In addition, the protocol for this systematic review was registered with PROSPERO (ID: CRD42022356244).

Eligibility criteria

Our inclusion criteria were limited to economic evaluations, of any type, that use a quantitative model and which (1) include young people living with HIV, (2) evaluate any diagnostic, therapeutic, behavioral, educational, financial, or mental health intervention related to HIV, and (3) report cost and health outcomes among young people. As our search was conducted in April of 2022, our review only considered studies published up to the end of April 2022. Studies were excluded if they (1) were not economic evaluations, (2) did not include young people, (3) did not evaluate interventions related to HIV, (4) did not report cost or health outcomes among young people, or (5) were not original research articles. See the appendix p 4 for a population, intervention, comparison, outcomes, timing, and study types (PICOTS) table of our inclusion and exclusion criteria. For this systematic review, we defined young people as individuals ages 10 to 24 years. This definition aligns with the WHO’s definition of “young people” as individuals ages 10 to 24 years.⁷

Information sources and search strategy

We searched the MEDLINE, Embase, Scopus, and Global Index Medicus databases for relevant literature. All searches were conducted from April 21, 2022 to April 24, 2022. Search terms included key words related to young people, HIV, and economic evaluations. A complete list of the search terms used for each database can be found in the appendix pp 5–7.

Selection and data extraction

Two reviewers independently screened all articles for inclusion and exclusion criteria at both the abstract and full-text level. Disagreements between reviewers were resolved by a third reviewer. For all articles meeting the inclusion criteria at the full-text level, data was extracted by two reviewers. Data extraction was based on items included in the Consolidated Health Economic Evaluation Reporting Standards (CHEERS); a checklist of standardized components that should be reported in any health economic evaluation.⁸ Data extracted included information on the study population and setting, intervention and comparators evaluated, evaluation perspective, type of evaluation, type of model used, model validation procedures, time horizon used, currency reported, costs and health outcomes reported, measurement and valuation of costs and health outcomes, sensitivity and subgroup analyses performed, cost-effectiveness results reported, and main conclusions. Our full data extraction template can be found in the appendix pp 8–9.

Risk of bias

The quality of each article was assessed by two reviewers. Quality was assessed using the CHEERS checklist. The CHEERS checklist contains 26 standardized items that should be reported in any health economic evaluation. We assessed each article for the presence or absence of each of these 26 items. For each article, these 26 items were marked as being included, partially included, or not included. Articles were given a score of 1 for each included item, a score of 0.5 for each partially included item, and a score of 0 for each non included item. Articles were then assigned an overall quality score equal to the mean score across all 26 items. Thus, a mean score across all 26 items of 0, 0.5, or 1, for example, indicates that a study is 0%, 50%, or 100% adherent to the CHEERS checklist, respectively.

Presentation of cost-effectiveness outcomes

For each article, we present the cost-effectiveness ratios or incremental cost-effectiveness ratios (ICERs) reported in the study. All cost parameters are presented with the currency and currency year reported in the study. For studies that do not report the currency year, we assume a currency year equal to the year of publication.

RESULTS

Overview

Our search returned 5,214 articles (Figure 1). After the removal of duplicate studies, 3,735 studies were screened for eligibility at the abstract level, and 197 studies were screened for eligibility at the full-text level. In total, 32 studies met the inclusion criteria and were therefore eligible for data extraction. Table 1 provides an overview of the key characteristics of the included studies. Of the 32 economic evaluations included in this analysis, 8 (25%) evaluated behavioral, educational, or financial interventions, 6 (19%) voluntary medical male circumcision (VMMC), 5 (16%) HIV testing, 4 (13%) pre-exposure prophylaxis (PrEP), 3 (9%) a hypothetical HIV vaccine, 2 (6%) antiretroviral therapy (ART), 1 (3%) condom distribution, and 3 (9%) combinations of expanded PrEP, ART, HIV testing, male circumcision, condom usage, and other HIV interventions. While all 32 (100%) evaluations included individuals aged 10–24 years, eight (25%) also included individuals aged 0–9 years, and 18 (56%) also included individuals aged 25+ years. Compartmental models were most common (25%), followed by Markov models (9%), microsimulations (16%), and other mathematical models (50%), which included dynamic networks and stochastic simulations. Twenty-five (78%) studies adopted a health system perspective while eight (25%) adopted a societal perspective. Twenty-two (69%) studies focused on African countries, nine (28%) on North American countries, and one (3%) on European countries. Seventeen (53%) studies used a time horizon of five or more years, nine (28%) a time horizon of less than five years, and seven (22%) a lifetime time horizon. In terms of health outcomes, 21 (66%) studies report HIV infections averted, nine (28%) QALYs gained, five (16%) DALYs averted, three (9%) life years gained, two (6%) deaths averted, and seven (22%) report other health outcomes.

Figure 1.

Flow diagram of our systematic review.

Table 1.

Study characteristics.

Characteristic	Number of studies (n=32)*
Setting
Africa	22 (69%)
North America	9 (28%)
Europe	1 (3%)
Population **
Individuals ages 0–9 years	8 (25%)
Individuals ages 10–24 years	32 (100%)
Individuals ages 25+ years	18 (56%)
Study design ***
Cost-effectiveness analysis	30 (94%)
Cost-utility analysis	2 (6%)
Study perspective ****
Health system perspective	25 (78%)
Societal perspective	8 (25%)
Type of model employed
Compartmental	8 (25%)
Markov	3 (9%)
Microsimulation	5 (16%)
Other mathematical model	16 (50%)
Intervention evaluated
Behavioral, educational, or financial intervention	8 (25%)
Male circumcision	6 (19%)
HIV testing	5 (16%)
Pre-exposure prophylaxis	4 (13%)
HIV vaccine	3 (9%)
Antiretroviral therapy	2 (6%)
Condom distribution	1 (3%)
Combination of HIV related interventions	3 (9%)
Time horizon
Less than 5 years	9 (28%)
5 or more years	17 (53%)
Lifetime	7 (22%)
Health outcomes
DALYs averted	5 (16%)
Deaths averted	2 (6%)
HIV infections averted	21 (66%)
QALYs gained	9 (28%)
Life years gained	3 (9%)
Other	7 (22%)

^*Percentages may not add to 100 because a study may be represented multiple times.

^**Appendix pp 10–11 shows the percentage of each study's target population that is in the age range 10–24 years.

^***Cost-utility analysis is a type of cost-effectiveness analysis in which health outcomes are measured in terms of quality-adjusted life years.

^****The study perspective indicates the point of view adopted with regard to the costs and effects included in the analysis. A health system perspective only considers costs and effects within a health system. A societal perspective considers all costs and effects regardless of who bears them.

Table 2 summarizes the results of included studies by intervention type. Behavioral, educational, and financial interventions were found to be cost-saving in one study⁹, cost-effective in six studies^10–15, and not cost-effective in one study.¹⁶ VMMC was found to be cost-saving in two studies^17,18 and cost-effective in four studies.^19–22 HIV testing was found to be cost-saving in one study²³, cost-effective in three studies^24–26, and not cost-effective in one study.²⁷ PrEP was found to be cost-saving in two studies^28,29 and cost-effective in two studies.^30,31 A hypothetical HIV vaccine was found to be cost-effective in three studies.^32–34 ART was found to be cost-effective in two studies.^35,36 Condom distribution was found to be cost-effective in one study.³⁷ Lastly, combination interventions were found to be cost-saving in one study³⁸ and cost-effective in two studies.^39,40 Table 3 provides a detailed breakdown of the data extracted from each of the 32 economic evaluations included in this analysis. The mean quality assessment score across all 32 studies was 0.80 (range: 0.69–0.90) (Table 4).

Table 2.

Summary of included studies, by intervention type

Intervention	Countries	Focus populations	Study designs	Model types	Time horizons	Outcomes assessed	Results
Behavioral, educational, or financial intervention 9–16	United States, Nigeria, South Africa, Uganda	Individuals ages 10–27 years	Cost-effectiveness analysis, cost-utility analysis	Mathematical model	3 months to 20 years	HIV infections averted, QALYs gained, additional patients virally suppressed	USD$6,900-$748,312 per HIA, USD$6,180-$57,372 per QALY gained, USD$970-$1,419 per additional patient virally suppressed
Male circumcision 17–22	Botswana, Ethiopia, Eswatini, Kenya, Lesotho, Malawi, Mozambique, Namibia, Rwanda, South Africa, Tanzania, Uganda, Zambia, Zimbabwe	Newborn males, adolescent males, adult males	Cost-effectiveness analysis	Compartmental model, mathematical model	11 years to lifetime	HIV infections averted	Cost-saving to USD$6,800 per HIA
HIV testing 23–27	Kenya, Malawi, Uganda, United States, Zambia, Zimbabwe	Children, youth ages 13–24 years, HIV infected children and adolescents, men who have sex with men age 15 years, adults ages 15–64 years, pregnant women	Cost-effectiveness analysis	Mathematical model, Markov model, microsimulation model, stochastic simulation model	31 months to lifetime	HIV infections averted, DALYs averted, QALYs gained, life expectancy	Cost-saving per HIA, USD$1,100-$2,000 per DALY averted, USD$769-$123,400 per QALY gained, USD$62,500-$96,000 per year of life saved
Pre-exposure prophylaxis 28–31	South Africa, United States	Adolescents, young adults, pregnant women, female sex workers, men who have sex with men	Cost-effectiveness analysis	Compartmental model, dynamic network model, microsimulation model, stochastic simulation model	5 years to lifetime	HIV infections averted, DALYs averted, QALYs gained	Cost-saving to USD$7,715 per HIA, cost-saving per DALY averted, USD$33,064-$427,788 per QALY gained
HIV vaccine 32–34	South Africa	Adolescents age 9 years, adolescents age 10 years, adults ages 15–49 years	Cost-effectiveness analysis	Markov model, microsimulation model	20 years to lifetime	QALYs gained, life years gained	USD$43 per QALYS gained, USD$5 per life year gained
Antiretroviral therapy 35,36	Kenya, United States	Individuals ages 0–59 years, adolescents living with HIV ages 13–24 years	Cost-effectiveness analysis	Compartmental model, microsimulation model	1–10 years	HIV infections averted, deaths averted, DALYs averted, QALYs gained	USD$7,900 per QALY gained
Condom distribution 37	England	Young people ages 13 to 24 years	Cost-effectiveness analysis	Mathematical model	Not reported	QALYs gained, sexually transmitted infections averted	GBP£17,411 per QALY gained
Combination of HIV related interventions 38–40	Kenya, South Africa	Adolescents ages 15–24 years, adults ages 15+ years, individuals ages 15–49 years, young women ages 16–24 years, adult women ages 25–49 years, female sex workers	Cost-effectiveness analysis	Compartmental model, mathematical model	5 years to lifetime	HIV infections averted, DALYs averted	Cost-saving to USD$20,000 per HIA, cost-saving to $1,706 per DALY averted

Table 3.

Data extracted from included studies

Reference	Study design	Model type	Setting	Population	Perspective - time horizon	Intervention (I) and comparators (C)	Currency and year	Health outcome measure	Cost-effectiveness outcomes
Alsallaq et al. 2017	Cost-effectiveness analysis	Deterministic compartmental model	Kenya	Adolescents ages 15–24 years and adults ages 15+ years	Health provider perspective - 5, 20, and 35 years	(I) 80% HIV testing coverage among youth + treatment at diagnosis for youth + 10% increased condom use among HIV positive diagnosed youth + male circumcision for HIV negative young men + PREP for HIV negative females ages 20–24 years + cash transfers for HIV negative girls ages 15–19 years (C) 70% HIV testing coverage with condoms for adults ages 15+ years + treatment at diagnosis for adults ages 15+ years	2012 USD	DALYs averted, HIV infections averted	80% HIV testing coverage among youth + treatment at diagnosis for youth + 10% increased condom use among HIV positive youth + high male circumcision coverage among HIV negative youth averted 94,000 more infections, 5.0 million more DALYs, and saved $46.0 million over 35 years in comparison to 70% HIV testing coverage among adults + increased condom use among adults + treatment at diagnosis for adults.
Awad et al. 2015	Cost-effectiveness analysis	Deterministic compartmental model	Zambia	Males ages 15–49 years	Health system perspective - 35 years	(I) 80% VMMC coverage among males ages 15–49 years; 80% VMMC coverage across different 5-year age groups from 0–99 years (C) Current VMMC coverage in Zambia (12.85%)	2015 USD	HIV infections averted	$1,089/HIA for 80% VMMC coverage among 15–49 year olds; $888/HIA for 80% VMMC coverage among 20–24 year olds; $3,300/HIA for 80% VMMC coverage among 45–49 year olds
Binagwaho et al. 2010	Cost-effectiveness analysis	Mathematical model	Rwanda	Newborns, adolescents, and adult men	Health care payer perspective - lifetime	(I) 70% MC coverage (C) Current MC coverage in Rwanda (15%)	2008 USD	HIV infections averted	Infant MC is cost saving at $3,233 saved/HIA; adolescent MC costs $3,932/HIA; adult MC costs $4,949/HIA
Cambiano et al. 2019	Cost-effectiveness analysis	Stochastic simulation model	Malawi, Zimbabwe, and Zambia	Adults ages 15–64 years	Health provider perspective - 50 years	(I) Community based HIV self-testing (CB-HIVST) (C) Status quo	2016 USD	DALYs averted	Targeting young people with CB-HIVST averts 1,490 infections per year at a cost $1,100–2,000/DALY averted; targeting adults with CB-HIVST averts 1,500 infections per year at a cost of $520–880/DALY averted
Culhane et al. 2020	Cost-effectiveness analysis	Dynamic compartmental model	Kenya	Individuals ages 0–59 years	Health system perspective - 10 years	(I) Long-acting ART (C) ART	2017 USD	HIV infections averted, DALYs averted, deaths averted	85% of adolescents and young adults switching from oral ART to long-acting ART can prevent 40,540 infections and 20,480 deaths over 10 years; to be cost-effective long-acting ART must be less than twice the annual per person cost of oral ART
deNecker et al. 2019	Cost-effectiveness analysis	Markov model	Kenya	Non-pregnant adults, pregnant women, and children	Health system perspective - 5 years for adults, 31 months for pregnant women	(I) Incorporation of HIV viral load test modules into existing GeneXpert infrastructure. (C) HIV viral load tests performed in centralized laboratories	2018 USD	HIV infections averted	The GeneXpert scenario averted 117 infections and 393 opportunistic infections with cost savings of $21,978,755 among non-pregnant and pregnant adults and $22,808,533 among non-pregnant adults, pregnant adults, and children
Ekwunife et al. 2021	Cost-effectiveness analysis	Mathematical model	Nigeria	Adolescents ages 15–24 years	Health provider perspective - 1 year	(I) Conditional economic incentives + motivational interviewing (C) Routine HIV care	2019 USD	Undetectable viral load	$1,419/additional patient with undetected viral load in comparison to routine HIV care
Fatti et al. 2018	Cost-effectiveness analysis	Mathematical model	South Africa	Adolescents ages 10–24 years	Health provider perspective - unclear	(I) Community based support (adherence counseling + psychosocial support + sexual and reproductive health education and family planning) + standard HIV care (C) Standard HIV care	2012 USD	Time to all-cause mortality after starting ART, time to loss to follow-up after starting ART	The intervention reduced patient attrition by 42.2% after one year and 35.9% after five years; the ICER was $600, $776, and $1,149/patient loss averted (defined as death averted or loss-to-follow-up averted) after one, two, and five years respectively.
Haacker et al. 2016	Cost-effectiveness analysis	Compartmental model	South Africa	Males of all ages	Health system perspective - 55 years for males age 0 years and 45 years for males of all other ages	(I) VMMC (C) No VMMC	2013 USD	HIV infections averted	Circumcision of one male age 20 years averts 0.227 infections, costs $659/HIA, and results in net financial savings of $617 over 45 years; targeting ages 0–15 averts 0.218–0.236 infections/MC at a cost of $859–1,157/HIA with net financial savings of $301–475/MC over 45 years; targeting ages 30–55 averts 0.005–0.079 infections/MC at a cost of $1,749–24,157/HIA with net losses of $2 to $95 over 45 years
Hontelez et al. 2011	Cost-effectiveness analysis	Microsimulation model	South Africa	Adults ages 15–49 years	Health system perspective - 20 years	(I) HIV vaccine (C) No HIV vaccine	2009 USD	HIV infections averted, life years gained	60% vaccine coverage of 15–29 year olds prevents 12% of new infections at a cost of $73/vaccine; 60% vaccine coverage of 15–49 year olds prevents 32% of new infections at a cost of $104/vaccine. Both strategies assume revaccination every two years
Jamieson et al. 2020	Cost-effectiveness analysis	Compartmental model	South Africa	Adolescents, young adults, pregnant women, female sex workers, and men who have sex with men	Health provider perspective - 20 years	(I) Introduction of PrEP to different sub-populations (C) Currently available interventions in the South African public sector HIV programme	2019 USD	HIV infections averted	ICERs for high-risk vs. all-risk sub-populations was $507/HIA vs. $4,537/HIA within female adolescents, $2,108 vs $5,637/HIA within male adolescents, $1,592/HIA vs. $10,323/HIA within young women, and $2,605/HIA vs. $7,715/HIA within young men
Kahn et al. 2001	Cost-effectiveness analysis	Mathematical model	United States	Gay men ages 18–27 years	Health provider and societal perspective - 1, 5, and 20 years	(I) Peer outreach to promote safe sex + peer led group sessions with discussion on HIV prevention + publicity campaign to increase awareness and acceptance of the intervention (C) No intervention	2000 USD	HIV infections averted, QALYs gained	The societal cost per HIA ranged from $39,300-$46,400 over 1 year, $14,600-$18,300 over 5 years, and $6,900-$9,900 over 20 years
Kripke et al. 2016a	Cost-effectiveness analysis	Compartmental model (DMPPT 2.1)	Botswana, Kenya, Lesotho, Malawi, Mozambique, Namibia, Rwanda, South Africa, Swaziland, Tanzania, Uganda, Zambia, and Zimbabwe	Men ages 15–49 years	Health system perspective - 11 years	(I) 80% VMMC coverage among men ages 15 to 49 (C) Status quo	2016 USD	HIV infections averted	80% VMMC coverage by 2015 averted 1.1 million infections through 2025 (860,000 more than the status quo) at a cost of $4,400/HIA
Kripke et al. 2016b	Cost-effectiveness analysis	Compartmental model (DMPPT 2.0)	Tanzania	Men ages 10–34 years	Health payer perspective - 36 years	(I) 80% VMMC coverage (C) Status quo	2014 USD	HIV infections averted	80% VMMC coverage among 10–24 year olds had the greatest impact on HIV incidence; 80% VMMC coverage among 15–34 year olds had the lowest cost per HIA; both strategies are cost saving
Lee et al. 2005	Cost-utility analysis	Mathematical model	United States	Young people living with HIV	Societal perspective - 3 months and 1 year	(I) Behavioral intervention to reduce HIV transmissions and improve quality of life (C) Standard of care	Not indicated	HIV infections averted	The intervention costs $103,366/HIA and averts 2.02 infections per 1,000 young people living with HIV
Moodley et al. 2016a	Cost-effectiveness analysis	Semi-Markov model	South Africa	Adolescents ages 10 years	Health provider perspective - lifetime	(I) HIV vaccine + standard of care (C) Standard of care	2012 USD	Life years gained	60% vaccination coverage at $12/dose yields an ICER of $5/life year gained in comparison to the status quo; national implementation targeting 10–19 year olds could result in 23.6 million life years gained
Moodley et al. 2016b	Cost-effectiveness analysis	Semi-Markov model	South Africa	Children ages 9 years	Health provider perspective - lifetime	(I) School based HIV vaccine campaign (C) Status quo	2012 USD	QALYs gained	At a cost of $12/dose the ICER is $43/QALY gained in comparison to the status quo
Neilan et al. 2018	Cost-effectiveness analysis	Microsimulation model (CEPAC)	United States	Youth ages 13–24 years	Health system perspective - lifetime	(I) One-time HIV screening in youth ages 13–24 + current HIV screening and testing practices (C) Current HIV screening and testing practices	2013 USD	CD4 cell count, life expectancy	ICER for screening at age 30 in comparison to current practice was $62,500/year of life saved; ICER for screening at age 25 compared to screening at age 30 was $96,000/life year saved; screens at ages 15, 18, and 21 were dominated
Neilan et al. 2021a	Cost-effectiveness analysis	Microsimulation model (CEPAC)	United States	Young people with HIV ages 13–24 years	Health payer perspective - 1, 5, and 10 years	(I) Interactive smartphone HIV treatment adherence reminders (C) Standard of care	Not indicated	HIV infections averted, deaths averted, life expectancy, QALYs gained	In comparison to the standard of care, the intervention reduced HIV infections by 15% and deaths by 12% yielding an ICER of $7,900/QALY gained
Neilan et al. 2021b	Cost-effectiveness analysis	Microsimulation model (CEPAC)	United States	HIV uninfected men who have sex with men ages 15 years	Healthcare sector perspective - lifetime	(I) Yearly, 6-month, and 3-month HIV screening among individuals ages 15–30 years (C) Current patterns of HIV screening	2018 USD	HIV infections averted, QALYs gained	In comparison to current practice screening yearly yields an ICER of $123,400/QALY gained, screening 6-monthly yields an ICER of $110,200/QALY gained, and screening 3-monthly yields an ICER of $70,900/QALY gained
Phillips et al. 2021	Cost-effectiveness analysis	Stochastic simulation model	South Africa	Adolescent girls and young women ages 15–24 years, female sex workers ages 1564 years, and all other individuals ages 15–64 years	Health provider perspective - 50 years	(I) PrEP scale-up (C) Current levels of PrEP uptake	2017 USD	DALYs averted	In comparison to the status quo, scaling-up PrEP to men and women aged 15–64 years and scaling up PrEP to women aged 15–24 years + female sex worker aged 15–64 years were both cost-saving strategies that reduced HIV incidence by 33% and 12% respectively
Pinkerton et al. 2000	Cost-utility analysis	Mathematical model	United States	African American male adolescents	Societal perspective - 3 months	(I) Educational intervention to reduce risky sexual behavior and increase knowledge of HIV/AIDS (C) Career opportunities workshop similar in design and duration to the intervention	1997 USD	HIV infections averted, QALYs gained	In comparison to the control, the intervention had an ICER of $57,327/QALY gained and $748,312/HIA
Pretorius et al. 2010	Cost-effectiveness analysis	Mathematical model	South Africa	Individuals ages 15–49 years	Societal perspective - 11 years	(I) PrEP + accelerated expansion of testing and treatment (C) Expansion of testing and treatment at current rate	2010 USD	HIV infections averted	Scaling-up the intervention to women ages 15–35 averts 10–25% of infections in this age group and 512% of infections in all age groups at a cost of $12,500-$20,000/HIA
Quaife et al. 2018	Cost-effectiveness analysis	Mathematical model	South Africa	Young women ages 16–24 years, adult women ages 25–49 years, and female sex workers	Health system perspective - lifetime	(I) Roll-out of oral PrEP + intravaginal rings + injectable antiretrovirals + microbicide gels + diaphragms (C) Current rates of male condom usage	2018 USD	DALYs averted, HIV infections averted	Among women ages 15–24 years: oral PrEP costs $810/DALY averted, oral PrEP + intravaginal ring costs $808/DALY averted, and all five intervention products together costs $700/DALY averted. The minimum cost per DALY averted for any combination of interventions among 25–49 year olds was $1706
Revill et al. 2015	Cost-effectiveness analysis	Mathematical model	Uganda and Zimbabwe	HIV infected children and adolescents	Health system perspective - 4.4 and 2.0 years	(I) Laboratory and clinical monitoring/continuing cotrimoxazole prophylaxis once children are stabilized on ART. (C) Clinically driven monitoring/stopping cotrimoxazole prophylaxis once children are stabilized on ART.	2012 USD	QALYs gained	Laboratory and clinical monitoring was not cost-effective in comparison to clinically driven monitoring. Continuing cotrimoxazole once stabilized on ART was cost-effective in comparison to stopping cotrimoxazole once stabilized on ART
Sadler et al. 2017	Cost-effectiveness analysis	Mathematical model	United Kingdom	Young people ages 13–24 years	Health system perspective - not reported	(I) Condom distribution program (C) Status quo	2015 GBP	Sexually transmitted infections averted, QALYs gained	In comparison to the status quo, the intervention yielded an ICER of £17,411/QALY gained
Stegman et al. 2019	Cost-effectiveness analysis	Compartmental model (DMPPT 2.1)	Namibia	Population of Namibia	Societal perspective - 15 years	(I) VMMC scale-up (C) Current VMMC levels	2011 USD	HIV infections averted	Scale-up of VMMC to males aged 15–29 years was the most cost-effective strategy with an ICER of $6,200/HIA; strategies targeting 15–24 year olds and 15–49 year olds were only slightly more expensive at $6,700/HIA and $6,800/HIA respectively
Tao et al. 1998	Cost-effectiveness analysis	Mathematical model	United States	Men ages 13–21 years who identify as gay, bisexual, or men who have sex with men	Societal perspective - 10 years	(I) Individual risk assessment + risk reduction counseling + peer education + HIV antibody testing + referral to medical or psychosocial services if needed + longitudinal follow-up (C) No intervention	1994 USD	HIV infections averted, QALYs gained	In comparison to the status quo, the intervention yielded an ICER of $6,180/QALY gained
Tozan et al. 2021	Cost-effectiveness analysis	Mathematical model	Uganda	Adolescents living with HIV ages 10–16 years	Health system perspective - 2 years	(I) Counselling sessions on HIV and ART adherence + mentorship from peers + asset building and financial planning workshops + bolstered standard of care (C) Bolstered standard of care	2015 USD	Viral suppression (HIV RNA viral load <40 copies/ml)	In comparison to the bolstered standard of care, the intervention yielded an ICER of $970/additional patient virally suppressed
Walensky et al. 2016	Cost-effectiveness analysis	Microsimulation model (CEPAC)	South Africa	Women ages 1825 years	HIV program perspective - 5 years	(I) Standard PrEP and long-acting PrEP (C) No PrEP	2014 USD	HIV infections averted	In comparison to no PrEP, and over 5 years, standard PrEP and long-acting PrEP cost $580 and $870 more per woman and averted 127 and 156 more infections per 1000 woman respectively; over the lifetime both PrEP and long-acting PrEP were cost saving in comparison to no PrEP
Wang et al. 2000	Cost-effectiveness analysis	Mathematical model	United States	Highschool students	Societal perspective - 1 year	(I) Sexual and reproductive health education intervention (C) Standard, information-based HIV prevention curriculum	1994 USD	Sexually transmitted disease infections averted, pregnancies averted	With a total cost of $105,243 the intervention prevented 0.12 cases of HIV, 24.37 cases of chlamydia, 2.77 cases of gonorrhea, 5.86 cases of pelvic inflammatory disease, and 18.5 pregnancies; the intervention saved $2.65 in societal costs for each dollar invested in the program
Wang et al. 2020	Cost-effectiveness analysis	Stochastic, dynamic network model	United States	Adolescent sexual minority males (ASMM)	Societal perspective - lifetime	(I) PrEP use for 10 years (C) No PrEP use	2017 USD	HIV infections averted, QALYs gained	In comparison to no PrEP, 10-year PrEP use yielded an ICER of $33,064/QALY gained among black ASSM and $427,788/QALY gained among white ASMM

Table 4.

Quality assessments.

	1	2	3	4	5	6	7	8	9	10	11	12	13	14	15	16	17	18	19	20	21	22	23	24	25	26	Mean
Alsallaq 2017	1	0	1	1	1	1	1	0.5	1	1	1	1	1	1	1	1	1	1	1	1	1	1	0	1	1	1	0.90
Awad 2015	1	0	1	1	1	0.5	1	0.5	1	1	1	1	0.5	1	1	1	1	1	0	1	1	1	0	1	1	1	0.83
Binagwaho 2010	1	0	1	1	1	1	1	1	1	1	1	1	1	1	1	1	0	1	0	1	1	1	0	1	1	1	0.85
Cambiano 2019	1	0	1	1	1	1	1	0	1	1	1	1	1	1	1	0	0	1	0	1	1	1	0	1	1	1	0.77
Culhane 2020	1	0	1	1	1	1	1	1	1	1	1	1	1	1	1	0	0	1	0	1	1	1	0	1	1	1	0.81
deNecker 2019	1	0	1	1	1	1	1	0	1	1	1	1	1	1	1	1	1	1	0	1	1	1	0	1	1	1	0.85
Ekwunife 2021	1	0	1	1	1	1	1	0	1	1	1	1	1	1	1	0	0	1	0	1	1	1	1	1	1	1	0.81
Fatti 2018	1	1	1	1	1	1	1	1	1	1	1	1	1	1	1	0	0	0	0	1	1	0	1	1	1	1	0.81
Haacker 2016	1	0	1	0.5	1	0	1	1	1	1	1	1	1	1	1	1	1	1	0	1	1	1	0	1	1	0	0.79
Hontelez 2011	1	0	1	1	1	1	1	1	1	1	1	1	1	1	1	1	1	1	0	1	1	1	0	1	0	0	0.81
Jamieson 2020	1	0	1	1	1	1	1	1	1	1	1	1	1	1	1	1	1	1	0	1	1	1	0	1	1	1	0.88
Kahn 2001	1	0	1	1	1	1	1	1	1	1	1	1	1	1	1	0	0	1	0	1	1	1	0	1	1	0	0.77
Kripke 2016a	1	0	1	1	1	0	1	0.5	1	1	1	1	0.5	1	1	1	1	1	0	1	1	1	0	1	1	1	0.81
Kripke 2016b	1	0	1	1	1	0	0.5	0.5	1	1	1	1	1	1	1	1	1	1	0	1	1	1	0	1	1	1	0.81
Lee 2005	1	0	1	0.5	1	1	1	0.5	1	1	1	1	0.5	1	1	0	0	1	0	1	1	1	0	1	1	1	0.75
Moodley 2016	1	0	1	1	1	1	1	0	1	1	0	1	1	1	1	0	0	0	0	1	1	1	0	1	1	1	0.69
Moodley 2016	1	0	1	1	1	1	1	1	1	1	1	1	1	1	1	0	0	1	0	1	1	1	0	1	1	1	0.81
Neilan 2018	1	0	1	1	1	1	1	1	1	1	0	1	1	1	1	0	0	1	0	1	1	1	0	1	1	1	0.77
Neilan 2021a	1	0	1	0	1	0.5	1	0.5	1	1	1	1	0	1	1	0	0	1	0	1	1	1	0	1	1	1	0.69
Neilan 2021b	1	1	1	1	1	1	1	1	1	1	1	1	1	1	1	0	0	1	0	1	1	1	0	1	1	1	0.85
Phillips 2021	1	0	1	0.5	1	1	1	0.5	1	1	1	1	1	1	1	1	1	1	0	1	1	1	1	1	1	1	0.88
Pinkerton 2000	1	0	1	1	1	1	1	1	1	1	1	1	1	1	1	1	1	1	0	1	1	1	0	1	1	0	0.85
Pretorius 2010	1	0	1	1	1	1	1	0	1	1	0	1	0	1	1	0	0	1	0	1	1	1	0	1	1	1	0.69
Quaife 2018	1	0	1	1	1	1	1	1	1	1	1	1	1	1	1	1	1	1	0	1	1	1	0	1	1	1	0.88
Revill 2015	1	0	1	1	1	1	1	0	1	1	0	1	1	1	1	1	1	1	0	1	1	1	0	1	1	1	0.81
Sadler 2017	1	0	1	1	1	1	0	1	1	1	1	1	1	1	1	1	1	1	0	1	1	1	0	1	1	1	0.85
Stegman 2019	1	0	1	1	1	1	1	0	1	1	0	1	1	1	1	1	1	1	0	1	1	1	0	1	1	0	0.77
Tao 1998	1	0	1	1	1	1	1	0	1	1	1	1	1	1	1	1	1	1	0	1	1	1	0	1	1	0	0.81
Tozan 2021	1	0	1	1	1	1	1	0	1	1	0	1	1	1	1	0	0	1	0	1	1	1	0	1	1	1	0.73
Walensky 2016	1	0	1	1	1	1	1	1	1	1	0	1	1	1	1	0	0	1	0	1	1	1	0	1	1	1	0.77
Wang 2000	1	0	1	1	1	0.5	0.5	0.5	1	1	1	1	1	1	1	0	0	1	0	1	1	1	0	1	1	0	0.71
Wang 2020	1	0	1	1	1	1	1	1	1	1	1	1	1	1	1	0	0	1	0	1	1	1	0	1	0	1	0.77

Abbreviations: Y=item completely fulfilled, P=item partially fulfilled, N=item not fulfilled.

Items: 1=study objectives, 2=health economic analysis plan, 3=population, 4=setting, 5=intervention and comparators, 6=perspective, 7=time horizon, 8=discount rate, 9=outcomes, 10=measurement of outcomes, 11=valuation of outcomes, 12=measurement and valuation of resources and costs, 13=currency and currency year, 14=model description, 15=analytics and assumptions, 16=characterization of heterogeneity, 17=characterization of distributional effects, 18=characterization of uncertainty, 19=approach to engagement with patients or others affected by study, 20=study parameters, 21=summary of main results, 22=effect of uncertainty, 23=effect of engagement with patients or others effects by study, 24=discussion, 25=source of financing, 26=conflict of interest statement. Mean = mean quality score across all 26 items.

Evaluations of behavioral, educational, and financial interventions

Eight studies evaluated behavioral, educational, or financial interventions for individuals living with, or at-risk of contracting, HIV.^9–16 Five of these evaluations were conducted in the United States, one in Nigeria, one in South Africa, and one in Uganda. All eight studies focused exclusively on young people in the age range of 10–27 years. Each of the interventions evaluated comprised at least one of the following elements: motivational interviewing, treatment adherence counseling, financial incentives, sexual and reproductive health education, or risk-reduction counseling. These interventions were found to be cost-saving in one study⁹, cost-effective in six studies^10–15, and not cost-effective in one study.¹⁶ In seven of the eight studies the control was standard HIV care. One study evaluating a health education workshop used a career opportunities workshop similar in design to the intervention as a control.¹³ Across the five studies conducted in the United States, cost-effectiveness measures ranged from USD$6,900-$877,025 per HIV infection averted (HIA) and USD$6,180-$57,327 per quality-adjusted life year (QALY) gained. The study in Nigeria reported a cost of USD$1,419 per additional patient with undetected viral load, the South Africa study reported USD$600-$1,149 per patient loss to follow-up averted, and the Uganda study reported USD$970 per additional patient virally suppressed.

Evaluations of VMMC

Six studies evaluated VMMC, all of which were conducted in African countries (Botswana, Ethiopia, Eswatini, Kenya, Lesotho, Malawi, Mozambique, Namibia, Rwanda, South Africa, Tanzania, Uganda, Zambia, and Zimbabwe).^17–22 One study found that increasing VMMC coverage among infants would prevent HIV infections when they become adolescents and over their lifetime and was cost-saving at USD$3,233 saved per HIA.¹⁷ Another study reported costs of USD$859-$1,157/HIA for increasing VMMC coverage among 0–15-year-olds.²⁰ Five studies reported increased VMMC coverage among 10–24-year-olds with costs ranging from cost saving to USD$6,700/HIA.^17–20,22 Two studies reported increased VMMC coverage among individuals 25 years or older with costs of USD$1,749-$24,157/HIA.^19,20 Five studies modeled increased VMMC coverage among 15–49-year-olds with costs ranging from cost saving to USD$6,800/HIA.^{17–19,21,22} In all, VMMC was found to be cost-saving in two studies^17,18 and cost-effective in four studies.^19–22

Evaluations of HIV testing

Five studies evaluated some form of HIV testing.^23–27 Three of these studies focused on African countries (Kenya, Malawi, Uganda, Zambia, and Zimbabwe)^23,26,27 and two on the United States.^24,25 Neilan et al. 2018 found that one-time HIV screenings in the United States at age 30 years yields an ICER of USD$62,500/life year saved in comparison to current age-specific HIV screening probabilities and is more cost-effective than one-time screenings at ages 15, 18, and 21 years.²⁴ Similarly, Neilan et al. 2021a found that one-time HIV screenings in the United States at age 15 years among HIV uninfected men who have sex with men yields an ICER of USD$123,400/QALY gained in comparison to current age-specific HIV screening probabilities.²⁵ Cambiano et al. 2019 evaluated a community-based HIV self-testing program in sub-Saharan Africa and found that the program costs USD$1,100-$2,000/DALY averted, in comparison to the status quo, when targeted at young people ages 15 to 24 years and USD$520-$880/DALY averted when targeted at adult men ages 25 to 49 years.²⁷ DeNecker et al. 2019 compared HIV viral load testing with GeneXpert infrastructure to laboratory-based HIV viral load testing in Kenya and found that the GeneXpert approach saves upwards of USD$20 million in antiretroviral, laboratory test, clinic visit, and opportunistic infection costs among pregnant adults, non-pregnant adults, and children (for reference, total costs in the laboratory-based HIV viral load testing scenario amounted to just over USD$400 million).²³ Revill et al. 2015 compared laboratory and clinical monitoring of HIV viral load to clinically driven monitoring of HIV viral load in sub-Saharan Africa and found clinically driven monitoring to be more cost-effective.²⁶ In all, revised or expanded HIV testing strategies were found to be cost-saving in one study²³, cost-effective in three studies^24–26, and not cost-effective in one study.²⁷

Evaluations of Pre-Exposure Prophylaxis

Four studies evaluated the expansion of PrEP coverage.^28–31 One study was conducted in the United States and three in South Africa. In the United States, 10-year PrEP use among adolescent sexual minority males was found to cost USD$33,064-USD$427,788/QALY gained from a societal perspective in comparison to no PrEP use and was therefore deemed potentially cost-effective.³¹ In South Africa, Phillips et al. 2021 found that expanding PrEP coverage among all individuals ages 15–64 years or only among women ages 15–24 years and female sex workers ages 15–64 years were both cost-saving strategies from a health systems perspective in comparison to no PrEP coverage in the population.²⁸ Similarly, Walensky et al. 2016 found that expanding PrEP coverage in South Africa to women ages 18–25 years was cost-saving from an HIV program perspective in comparison to no PrEP coverage.²⁹ Jamieson et al. 2020 compared the cost of expanding PrEP coverage among various sub-populations in South Africa from a provider perspective to no PrEP coverage and estimated the cost of coverage to be USD$4,537/HIA among female adolescents, USD$5,637/HIA among male adolescents, USD$10,323/HIA among young women, and USD$7,715/HIA among young men.³⁰ Jamieson et al. 2020 concluded that targeting specific high-risk groups is more cost-effective than targeting the general population.

Evaluations of an HIV vaccine

Three studies evaluated the cost-effectiveness of rolling-out a hypothetical HIV vaccine, all of which were conducted in South Africa.^32–34 Hontelez et al. 2011 report that 60% vaccination coverage among 15–29-year-olds, with an assumed vaccine efficacy of 31.2%, can prevent 12% of new infections at a cost of USD$73 per vaccine while 60% vaccination coverage among 15–49-year-olds can prevent 32% of infections at a cost of USD$104/vaccine.³² Moodley et al. 2016a report that 60% vaccination coverage among 10-year-olds, with an assumed vaccine efficacy of 50%, would only cost USD$5/life-year gained in comparison to no vaccination.³⁴ Similarly, Moodley et al. 2016b reports that scaling-up vaccination coverage to 9-year-olds, using an assumed vaccine efficacy of 50%, would cost USD$43/QALY gained in comparison to no vaccination.³³ All three studies conclude that an HIV vaccine with modest efficacy would be cost-effective in comparison to no vaccination.

Evaluations of ART

Two studies evaluated ART related interventions.^35,36 One study conducted in Kenya compared long-acting ART to oral ART and found that scaling up long-acting ART to adolescents and young adults can prevent upwards of 20,000 deaths over 10 years but must be less than twice the annual per person cost of oral ART to be cost-effective.³⁵ Another study from the United States compared mobile ART adherence reminders for youth living with HIV to the standard of care and found that the intervention, from a healthcare payer perspective, costs USD$7,900/QALY gained and is cost-effective.³⁶

Evaluations of condom distribution

One study compared national scale-up of a condom distribution program targeting all young people ages 13 to 24 years in England to current levels of condom usage among this age group and found that the intervention costs £17,411/QALY gained and is cost-effective from a health system perspective.³⁷

Evaluations of combination interventions

Three studies evaluated the simultaneous scale-up of multiple HIV interventions.^38–40 Alsallaq et al. 2017 compared expanded testing coverage, treatment coverage, condom use, PrEP coverage, and VMMC coverage among adolescents ages 15–24 years in Kenya to expanded testing coverage, treatment coverage, and condom use among all adults ages 15+ years and found the adolescent focused strategy to be cost-saving from a health provider perspective in comparison to the adult focused strategy.³⁸ Pretorius et al. 2010 compared expanded PrEP coverage, testing coverage, and treatment coverage among women ages 15–35 years in South Africa to the status quo and found that the intervention costs USD$12,500-$20,000 per HIA from a societal perspective and is cost-effective.³⁹ Quaife et al. 2018 compared the roll-out of various combinations of expanded PrEP, intravaginal ring, long-acting ART, microbicide gel, and diaphragm coverage to current rates of male condom use in South Africa and calculated health system perspective intervention ICERs of USD$700-$810/DALY averted among females ages 15–24 years and USD$1,706/DALY averted among 25–49-year-olds which were deemed cost-effective.⁴⁰

Quality assessment

Table 4 shows the results of our quality assessment for each of the articles included in our analysis. Elements of the CHEERS checklist that were most commonly not reported included engagement with patients or others affected by the study (96.9%), a health economic analysis plan (93.8%), characterization of distributional effects (53.1%), characterization of heterogeneity (50.0%), discount rate (28.1%), valuation of outcomes (21.9%), and conflict of interest statements (21.9%). Three studies (9.4%) did not report the perspective, one study (3.1%) did not report the time horizon, and two studies (6.3%) did not report the currency or currency year. Only two studies (6.3%) did not report sensitivity analyses. The mean quality score across all thirty-two articles was 0.80. The minimum quality score was 0.69 and the maximum was 0.90.

DISCUSSION

This systematic review summarizes evidence from 32 model-based economic evaluations on the economic and health benefits of targeting HIV interventions to young people. Interventions examined in this review include behavioral, educational, and financial interventions, VMMC, HIV testing, PrEP, HIV vaccines, ART, and condom distribution. Results from this review compare the cost-effectiveness of these interventions and highlight evidence gaps in current economic analyses addressing young people living with HIV.

Among the interventions captured by this review, VMMC appears to be the most consistently cost-effective HIV prevention strategy for young people. This finding aligns with a previously conducted systematic review on the cost-effectiveness of HIV prevention interventions in sub-Saharan Africa.⁴¹ The review by Sarkar et al. 2019 found VMMC to be among the most cost-effective interventions evaluated across the 60 studies included in the review. Also consistent with our findings, the authors of Sakar et al. 2019 report VMMC coverage among adolescents to be cost saving in comparison to coverage among adult populations.⁴¹ Our review also suggests that an HIV vaccine with only 31.2% or 50% efficacy may be among the most cost-effective HIV prevention interventions for young people, with ICERs as low as USD$5 per year of life gained in comparison to a no vaccine scenario. However, it must be noted that no such vaccine is currently accessible.⁴² While the remaining intervention categories (behavioral, educational, and financial interventions, HIV testing, PrEP, ART, condom distribution, and combination interventions) all demonstrated cost-effectiveness, their degree of cost-effectiveness varied substantially across studies with some interventions being cost-saving or cost-effective in some studies yet not cost-effective in other studies.

Some of the interventions captured by this review also demonstrated a clear trend within or across studies of being more cost-effective when targeted to young people in comparison to adults. Seventy percent VMMC coverage in Rwanda, for example, had a cost of USD$3,932 per HIA when applied to adolescents and a cost of USD$4,949 when applied to adults, relative to current national VMMC rates.¹⁷ As another example, expanded PrEP coverage in South Africa, relative to current HIV programs, had a cost of USD$507 per HIA among female adolescents but a cost of USD$1,592 per HIA among young adult women.³⁰ Similarly, a financial intervention targeting adolescents ages 10–16 years living with HIV in Uganda had a cost of USD$970 per additional patient virally suppressed¹⁵ whereas a similar intervention targeting adolescents ages 15–24 years living with HIV in Nigeria had a cost of USD$1,419 per additional patient virally suppressed.¹⁶ Although based on a small body of evidence, these findings hint at the potential health and economic advantages of prioritizing HIV care for young people.

Twenty-five percent of the studies included in our review evaluated interventions designed to help young people cope with, rather than prevent or treat, HIV. More specifically, eight of the thirty-two studies included evaluated behavioral, educational, or financial interventions comprising motivational interviewing, treatment adherence counseling, financial incentives, sexual and reproductive health education, and or risk-reduction counseling. Young people living with HIV must cope with unique challenges including biological changes following puberty, transitions from dependency on parents to independence and self-reliance, stigma, peer pressure, and rejection by peers, all while managing a chronic illness. These biological and psychosocial challenges often lead to worse clinical and mental health outcomes in comparison to both HIV-negative young people and HIV-positive adults.^43,44 Understanding the economic and health benefits of interventions designed to address these challenges is therefore vital to informing appropriate HIV health policies in regions where HIV prevalence is high among young people. The studies identified in this review, and which evaluate behavioral, educational, or financial interventions, present a wide range of cost-effectiveness outcomes and suggest additional studies are needed to better understand the benefits of such interventions. In addition, none of the studies identified by this review evaluated mental health interventions as a main focus. Mental health is an essential component of HIV care for young people transitioning into adulthood, yet few mental health interventions for young people living with HIV (in both HICs and LMICs) have been studied and the cost-effectiveness of these interventions remain unknown.⁴⁵ Evaluating the economic and health impact of mental health interventions for young people living with HIV will be crucial to integrating these services into health systems to ensure a comprehensive approach to HIV care for this population.

The eight studies identified in this review which evaluate interventions designed to help young people cope with HIV occurred predominately in HICs. Only three of these eight studies occurred in LMICs. A past review by Mellins et al. 2013 found a similar trend in which a majority of studies evaluating mental health challenges and outcomes among adolescents living with HIV occurred in HICs.⁴⁶ A lack of data on mental health outcomes and mental health interventions for young people living with HIV in LMICs may be driven by resource limitations that prevent the expansion of clinical services specific to HIV infected young people, shortages of clinical staff trained in pediatric HIV care and mental health, or a lack of research to culturally adapt and validate instruments that measure mental health outcomes, but which were developed in HICs.⁴⁵ Given that most of the world’s HIV infected young people live in LMICs, additional evaluations are needed to inform HIV health policies that provide appropriate care to young people in LMICs. One possible solution to generating additional evidence is to pair randomized controlled trials (RCTs) with economic evaluations. Mavhu et al. 2020, for example, conducted a costing analysis alongside an RCT of enhanced HIV care support for adolescents aged 13–19 years in Zimbabwe.⁴⁷ The authors found that the peer-supported community-based intervention significantly improved virological suppression at an annual cost of USD$1,340 and USD$450 per virally suppressed adolescent in the intervention and control group, respectively.⁴⁷ Conducting economic evaluations alongside RCTs can reduce the time between evidence generation and policy impact and ensure that cost-effective interventions reach those most in need as quickly as possible.

While the models identified in this review are of high quality as indicated by an average quality assessment score of 0.80 (Table 4), there is substantial variation between models that makes comparison difficult. Across all models, for example, the definition of “young people” varies from 10–24 years to 13–24 years to 15–24 years. Such variation exists even among models evaluating the same type of intervention. For example, among models evaluating HIV testing, the time horizons ranged from 31 months to the lifetime of individuals. As another example, models evaluating a behavioral, educational, or financial intervention used outcome measures that varied between HIV infections averted, QALYs gained, and additional patients virally suppressed. In addition, the perspectives modeled across all studies varied between a health system, program, and societal perspective (Table 3). Future economic models that quantify the economic and health effects of interventions targeting young people living with HIV may benefit from adhering to a standardized analysis framework. Under such circumstances researchers could present results using standardized definitions (ex. for age group, perspective, time horizon, health outcome measure, cost-effectiveness outcome, and cost-effectiveness threshold) in addition to results using non-standard definitions that may be better suited to the study question or context. Inclusion of results from a standardized analysis would ensure direct comparability between studies evaluating the same, or different, interventions.

There are several limitations to the evidence presented in this review that must be considered. First, few studies are presented for each of the interventions analyzed in this review. Basing decisions on such a small body of cost-effectiveness evidence may lead to undesirable outcomes. Second, the intervention comparators, time horizons, perspectives, discount rates, and health outcomes employed vary widely across the economic evaluations included in this review. Consequently, comparing results between studies evaluating the same, or different, interventions is difficult and sometimes not possible. Despite these limitations, this review summarizes available evidence on the economic and health benefits of interventions for young people at-risk or living with HIV and highlights areas where further research is required.

Overall, this review has important policy and research implications. To our knowledge, this is the first systematic review to document economic evaluations of interventions for young people living with HIV. Evidence presented in this review may help policy makers emphasize the value of investing resources in intervention packages designed specifically for HIV infected young people. In addition, this review offers recommendations for future research, namely that future economic evaluations of interventions for young people living with HIV should focus on strategies to improve mental and behavioral health outcomes in LMICs and that such evaluations should adhere to standardized practices to allow for comparability of cost-effectiveness outcomes between interventions and settings.

Implications and Contribution.

This review highlights the value of investing in interventions designed specifically for young people living with HIV. Evidence suggests that such interventions are a cost-effective, approach to combatting the global HIV epidemic. Future economic evaluations of HIV interventions targeting young people should adhere to standardized practices to allow for comparability.

Data sharing statement

All data used for this study is available within the manuscript or attached appendices.