JMIR Publications


We are scheduled to perform a server upgrade on Thursday, November 30, 2017 between 4 and 6 PM Eastern Time.

Please refrain from submitting support requests related to server downtime during this window.

JMIR Public Health and Surveillance

Advertisement

Citing this Article

Right click to copy or hit: ctrl+c (cmd+c on mac)

Published on 20.06.17 in Vol 3, No 2 (2017): Apr-Jun

This paper is in the following e-collection/theme issue:

    Viewpoint

    Ensuring Inclusion of Adolescent Key Populations at Higher Risk of HIV Exposure: Recommendations for Conducting Biological Behavioral Surveillance Surveys

    1Independent Consultant, UNICEF, East Asia and Pacific Regional Office, Valencia, Spain

    2UNESCO, Section of Education for Inclusion and Gender Equality, Paris, France

    3Youth Lead, Bangkok, Thailand

    4UNICEF, East Asia and Pacific Regional Office, Bangkok, Thailand

    *these authors contributed equally

    Corresponding Author:

    Lisa Grazina Johnston, MA, MPH, PhD

    Independent Consultant, UNICEF, East Asia and Pacific Regional Office

    14 Arzobispo Mayoral Puerta 5

    Valencia, 46002

    Spain

    Phone: 34 634913918

    Fax:34 634913918

    Email:


    ABSTRACT

    Ending acquired immune deficiency syndrome (AIDS) depends on greater efforts to reduce new human immunodeficiency virus (HIV) infections and prevent AIDS-related deaths among key populations at highest HIV risk, including males who have sex with males, sex workers, and people who inject drugs. Although adolescent key populations (AKP) are disproportionately affected by HIV, they have been largely ignored in HIV biological behavioral surveillance survey (BBSS) activities to date. This paper reviews current ethical and sampling challenges and provides suggestions to ensure AKP are included in surveillance activities, with the aim being to enhance evidence-informed, strategic, and targeted funding allocations and programs toward ending AIDS among AKP. HIV BBSS, conducted every few years worldwide among adult key populations, provide information on HIV and other infections’ prevalence, HIV testing, risk behaviors, program coverage, and when at least three of these surveys are conducted, trend data with which to evaluate progress. We provide suggestions and recommendations on how to make the case to ethical review boards to involve AKP in surveillance while assuring that AKP are properly protected. We also describe two widely used probability sampling methods, time location sampling and respondent driven sampling, and offer considerations of feature modifications when sampling AKP. Effectively responding to AKP’s HIV and sexual risks requires the inclusion of AKP in HIV BBSS activities. The implementation of strategies to overcome barriers to including AKP in HIV BBSS will result in more effective and targeted prevention and intervention programs directly suited to the needs of AKP.

    JMIR Public Health Surveill 2017;3(2):e40

    doi:10.2196/publichealth.7459

    KEYWORDS



    Introduction

    Ending human immunodeficiency virus (HIV) or acquired immune deficiency syndrome (AIDS) relies on greater efforts to reduce new HIV infections and prevent AIDS-related deaths among key populations at highest risk. Whereas data are limited, studies from low and concentrated epidemic countries suggest that HIV prevalence is disproportionately high among adolescents, aged 10-19 years, who sell sex, engage in same-sex relationships, and inject drugs [1]. Ignoring that these behaviors occur among adolescents places them at even higher risk of HIV infection and creates barriers to HIV testing, and, when needed, essential care and treatment [2]. Adolescents under the age of 18 years who sell sex and are likely to be defined as sexually exploited under human rights law [3,4], tend to be the most ignored and least protected group due to controversy and lack of clarity on how to meet their needs [5-7]. In addition, adolescent key populations (AKP) including adolescent males who have sex with males (MSM); male, female, and transgender adolescents who sell sex; and adolescents who inject drugs face numerous vulnerabilities including low education; low access to health care; and high levels of stigma, discrimination, and violence [2]. A punishing legal environment and severe taboos around same-sex relations, selling sex, and injecting drug use tend to drive these behaviors underground, reinforcing the exclusion of adolescents and perpetuating the infection cycle [2,8].

    Since early 2000, HIV biological behavioral surveillance surveys (BBSS) have been a key component of national responses to HIV [9,10]. BBSS, conducted every few years worldwide among adult key populations, is the routine collection of strategic information to measure HIV and other infections and biological and behavioral trends; to model the HIV epidemic; to determine allocation of limited resources and funding; to enhance efforts to more effectively respond to local, national, and global HIV prevention strategies; and to measure program coverage. For countries that have conducted at least three BBSS, trends provide information about different countries’ progress in responding to HIV. Many questions in the BBSS are sensitive, including those related to sexual behaviors; drug use; HIV testing history; visits to sexually transmitted infections services; and personal experiences of being arrested, raped, stigmatized, and discriminated against. Unfortunately, many of these surveillance activities do not collect data from persons under the age of 18 years. Some surveys have managed to collect data from participants as young as 15 years [11,12]; however, these data are not disaggregated (by age and sex), and the sample sizes comprising 15 to 19 year olds are often not sufficient for meaningful analysis [13,14]. To fill this gap, some countries have conducted separate studies on younger cohorts; however, many studies use non-probability sampling methods from which inferences about the population cannot be made [15,16].

    The lack of continuous, systematic collection, analysis, and interpretation of data on AKP has resulted in a dearth of strategic information needed for the planning, implementation, and evaluation of essential HIV programs. This paper discusses the ethical barriers related to sampling AKP, and suggestions on how to overcome them, and presents recommendations on how to include AKP in BBSS using time location sampling (TLS) and respondent driven sampling (RDS).


    Methods

    Ethical Barriers to Sampling Adolescent Key Populations (AKP)

    One of the barriers to AKP inclusion in BBSS is ethical constraints. In almost all countries, adolescents under 18 years are considered children and entitled to the protection of their rights under the Convention on Rights of the Child [17]. Compared with adults, children are more vulnerable to exploitation, abuse, violence, and other harmful outcomes of research, and therefore, require additional safeguards [18-20]. In order for children to participate in research, informed consent is usually required from both the child and the child’s parent or guardian. The failure of institutional review boards to approve self-consent and waive guardian permission in conducting HIV surveys is a significant barrier to AKP participation. AKP may fear being stigmatized, punished, or in some cases victimized by their families if guardian permission results in the disclosure of their behaviors, sexual orientation or gender identity, or HIV status [21]. As a result, surveys may end up with smaller samples and unrepresentative findings. Convincing governments to approve self or proxy (eg, trained social worker serving as a proxy guardian) consent may require building consensus of the value of AKP inclusion and promoting their right to participate, whereas at the same time ensuring their protection from harm. Key ethical parameters of data collection involving key populations under the age of 18 years include (1) informed consent; (2) domestic laws governing child protection; (3) identification of, and referral to, services for AKP; and (4) biological testing. Within each of these parameters, considerations and suggestions for supporting the inclusion of AKP in BBSS are presented in Table 1.

    Table 1. Ethical considerations and suggestions for conducting biological behavioral surveillance surveys (BBSS) among adolescent key populations (AKP) under the age of 18 years.
    View this table

    Informed consent for AKP under the age of 18 years should include considerations beyond general ethical assurances included in any protocol. Extra effort may be needed to ensure that adolescents understand all of the elements in a consent process, including the purpose of the research, the kinds of information to be collected, how confidentiality will be maintained, the interview procedure (in particular that the participant does not have to answer questions with which she or he does not feel comfortable, and that the interview can be stopped at any time), and a contact number for more information about the study or to make a complaint [22]. For instance, in the Philippines, AKP were required to understand and repeat back 4 key items of consent in order to participate in a BBSS using RDS and TLS: (1) participation is voluntary, (2) information is confidential (no one will know what you tell me), (3) participation involves an interview and HIV counseling and testing, and (4) participation will help improve services for adolescents [23]. Another safeguard has been to have on-site social or health workers as “parental proxies” to provide consent on behalf of or in addition to AKP under the age of 18 years [24].

    Some countries have specific regulations that disclosures of violence, abuse, neglect, or exploitation of a child override confidentiality and must be reported to relevant authorities. Government employees or particular professions (eg, social workers, health workers, and teachers) or any person aware of an incident must report it. If there are no exemptions for mandatory reporting for the research, then a waiver from an appropriate authority is needed so that interviewers are not required to report abuse disclosures without the adolescent’s approval. It is essential for survey planners to (1) review domestic mandatory reporting laws of disclosures of child abuse, neglect, violence, or exploitation and consider how reporting would affect the final research outcome; and (2) discuss options to waive mandatory reporting by adolescent protection officials, social workers, rights advocates, and partner agencies.

    Collecting biological specimens from AKP under the age of 18 years will normally require informed consent from parents and guardians. In many situations, parents and guardians will have access to the test results [25], which can be a strong impediment to AKP getting tested, especially given the sensitivity of an HIV result and the implications of sexual activity or injecting drug use [2]. In light of the increased risk of HIV infection faced by adolescents, some countries are reevaluating and adapting their current legal and policy constraints requiring parental consent for adolescents wanting confidential HIV counseling and testing [21,26,27]. Twelve countries in Asia and the Pacific now have HIV testing laws in place that allow people under 18 years to consent independently to HIV testing [2].

    Methods to Sample Adolescent Key Population (AKP)

    Much of our knowledge about adolescent health comes from household- and school-based surveillance, both of which rely on populations that have sampling frames [28]. However, these surveys typically miss populations at higher risk for HIV exposure, many of whom have unstable living environments and housing and prefer to remain “hidden” from law enforcement and other government authorities. Furthermore, these studies fail to capture strategic information on HIV, including HIV and other infections prevalence, program coverage, or specific risk behaviors. There are currently 2 probability-based sampling methods that have been used successfully to sample adult key populations without sampling frames for BBSS—TLS (also known as venue-day-time sampling) and RDS [10,29,30]. Both of these methods allow participants to remain anonymous (ie, no need to collect name, address, or other personal details that can be used to contact a participant) and are therefore, effective at sampling populations that practice risky behaviors and/or face stigma and discrimination. Knowledge about the target population is needed before deciding on which methodology to use. For instance, TLS requires that AKP congregate in identifiable and accessible locations such as street corners, markets, transportation centers, clubs, bars, or other areas [9,31,32], and RDS requires that the population be socially networked so that AKP can identify and recruit other eligible AKP.

    TLS was first used to estimate HIV seroprevalence among young MSM (15-22 years) in the United States [33]. The method entails identifying days and times when populations congregate at specific locations (ie, brothels, city blocks, bars, and so on), constructing a sampling frame of time and location units, randomly selecting and visiting time and location units (the primary sampling units), and systematically intercepting and collecting information from consenting individuals. The number of persons at each location provides a sampling weight that can be used a priori, to draw a self-weighting sample, or post priori, in analysis. Data collection takes place at the venue, if space (or venue owner) permits, or in a mobile site near the location, such as a van, or by making appointments to come to a designated study site. The major contribution of TLS over other cluster sampling methods is the ability to account for the fact that key populations are not statically associated with a particular location and often move between multiple locations during a single day. As such, TLS allows researchers to construct a sample with known properties, make statistical inference to the larger population of location visitors, and theorize about the introduction of biases that may limit generalization of results to the target population.

    RDS was first used to sample people who inject drugs in the United States [34]. Briefly, RDS begins with a handful of non-randomly selected participants (referred to as seeds) who enroll in the survey and upon completion, receive recruitment coupons that they use to recruit their peers [34-37]. Participants recruited by seeds make up the first wave of participants, who in turn recruit the second wave of participants. This recruitment process continues until the sample comprises multiple recruitment waves and ends once the desired sample size is reached. RDS provides a primary incentive for completing an interview and for recruiting peers. The use of uniquely numbered coupons identifies who recruited whom, while maintaining anonymity. Someone receiving a coupon can decide whether to enroll or not. The major contribution of RDS over other chain referral sampling methods is the ability to harness information about people’s social networks to make inferences about the population. Data on who recruited whom are used to adjust for differential recruitment, and the measurement of each participant’s social network size (the number of peers they know who also know them) is used to weigh the data to control for the overestimation of those with larger social networks and the underestimation of those with smaller social networks.

    Both methods can, and have been, used for HIV and non-HIV related surveys of AKP (10-19 years) [38,39]. For example, in Asia, TLS was used in a survey conducted among MSM, male sex workers, and transgender persons (15-24 years; 30.6% of which were between the ages of 15-19 years) in Chiang Mai and Phuket, Thailand (n=827) [40]. In this study, using oral testing for HIV, 13.1% of AKP were HIV seropositive [40]. In Cambodia, TLS was used to sample 1204 males and 1166 females (10-24 years; 52.4% of which were boys and 53.2% of which were girls between the ages of 15-19 years), who were unmarried, and considered high risk based on their visibility at high risk venues. The sample comprised 252 hotspots (ie, bars, night clubs, karaoke parlors, massage parlors, street corners, football fields, snooker clubs, and computer game shops) in the capital city and 7 other provinces [41]. Although information about HIV-related risk behaviors was captured, no biological testing was performed, and there is no indication of adjustments in the analysis for sizes of venues and frequency of visits. Another survey using TLS that directly targeted young key populations living on the streets in the Russian Federation and Ukraine, mapped metro and train stations, street markets, feeding centers or fast-food sites, parks, and computer clubs [42]. This survey used 2 mobile teams in vans in which participants consented to participate, were interviewed, and underwent pre- and post-test counseling and a rapid HIV test.

    RDS was used in several provinces of Thailand among young (15-24 years) females who exchange sex for money and goods, MSM, young non-Thai migrants, and transgender youth [43]. Although RDS recruitment worked well, the sample sizes were not sufficient enough to capture many adolescents, and although information about HIV-related risk behaviors and testing was captured, no biological testing was performed. These limitations were also found in RDS surveys conducted among young MSM in Monywa and Yangon, Myanmar [44].

    Web-based RDS has been developed to sample hard to reach populations through messaging and emails [45-47]. Although this method does not allow for the collection of biological information since there is no face-to-face contact between research staff and participants, it can be useful and efficient for collecting behavioral information from AKP. For instance, in China, an RDS survey is being planned for adolescent males who have sex with males using a Web-based application. A working group of young MSM and representatives from community-based networks are involved in the questionnaire and survey design and will help in the selection of seeds. Although the inclusion of young key populations, most of whom do not have research skills, involves additional time, their input into the questionnaire has resulted in the prioritization of interview questions, the inclusion of adolescent appropriate language, and important local terms.

    When using TLS and RDS to conduct research on AKP, it is important to consider how the methodology should be modified in order for it to be accepted by, and appropriate for, adolescents. When selecting time periods for sampling AKP, one TLS survey of adolescents who use drugs adjusted their data collection activities to avoid hours when activities such as school, work, or chores were most likely to occur. When selecting a data collection site, an RDS survey of street youth in Albania anticipated that waiting space would be needed for older siblings, parents, or other caretakers accompanying an adolescent participant or for younger siblings who were in the care of an adolescent participant [39]. An extra staff person was hired to engage younger siblings with toys, puzzles, games, and other activities as they waited for their adolescent caretaker to finish the survey process. In place of cash incentives, surveys of adolescents have used food items, soap, clothing, games, and other items [39]. Surveys of AKP conducted in Thailand and Myanmar used phone credit as incentives [43,44]. When asking AKP about social network sizes, some RDS surveys have used pictures and counting techniques to help adolescents think about size differentials. For instance, one survey used several different pictures of circles of increasing sizes to help adolescents visualize their most accurate social network size [39]—the smallest circle indicated the smallest social network size (ie, one or two), and the largest size indicated the largest network size (ie, up to 100). Tables 2 and 3 provide an overview of the features specific to each of the sampling methodologies, how each feature is generally implemented, and considerations on how they can be modified for AKP.

    Table 2. Time location sampling (TLS) methodology features, description of those features, and considerations for applying those features when conducting surveys of adolescent key populations (AKP).
    View this table
    Table 3. Respondent driven sampling (RDS) methodology features, description of those features, and considerations for applying those features when conducting surveys of adolescent key populations (AKP).
    View this table

    Both TLS and RDS require additional considerations when collecting biological and sensitive behavioral data from AKP (Table 4). Pre-survey assessments, data collection forms, interviewing, biological specimen procedures techniques, and eligibility descriptions may need to be modified when sampling youth populations. Additional staff may be needed to address the specific needs of AKP, especially those who have been in abusive situations. For instance, in some RDS surveys of adolescents, trained social workers were hired to respond to any participant needing assistance, feeling distress as a result of answering sensitive questions, or expressing that they were in a harmful or abusive situation [23,39]. Once data are analyzed, important findings should be shared with the population and validated. For instance, in the BBSS RDS surveys conducted in the Philippines, the surprising finding that a large percentage of adolescent female sex workers and males who have sex with males were attending school resulted in a validation process [48]. Qualitative research conducted by the Department of Health in schools indicated that even though HIV education was part of the curriculum, few students received adequate HIV information in school. Identified reasons included that teachers were not receiving formal HIV training; that textbooks had outdated HIV information; and that some teachers were not comfortable discussing HIV and skipped lessons about HIV. This exercise resulted in plans to strengthen HIV education, thereby, reaching school-enrolled AKP and building upon HIV BBSS findings.

    Table 4. Special considerations and suggestions when adapting biological behavioral surveillance surveys (BBSS) among adolescent key populations (AKP) in time location sampling (TLS) and respondent driven sampling (RDS).
    View this table

    Discussion

    This paper is a product of an international meeting convened to share lessons learned, improvements and innovations, and outputs on HIV surveillance activities. This was the third such meeting (2004 in Addis Ababa, Ethiopia; 2009 and 2015 in Bangkok, Thailand) but the first to include a focus on AKP, reflecting recent concerns about their absence in HIV BBSS [49]. Data from surveys conducted among AKP thus far have provided essential information to respond to the specific needs of AKP, including overlapping risks (ie, adolescents who inject drugs and sell sex), sexual behaviors, barriers to service uptake, violence, as well as HIV- and non-HIV-related (ie, reproductive and mental) health needs [14,48,50].

    We note that there are several ethical barriers to sampling adolescents; however, using the strategies proposed here, AKP under the age of 18 years can be involved in research and still be protected from undue harm. Countries enacting laws allowing adolescents to access HIV testing without parental consent may serve as leverage to argue the benefits of conducting research among AKP without parental consent, as long as there are sufficient ethical protections and access to needed care and treatment.

    During the past decade or more, HIV prevalence and associated risk behaviors among adult key populations worldwide have been successfully measured using innovative sampling methods including TLS and RDS. However, it is unclear why the lessons learned from surveys of adult key populations are not being harnessed to capture strategic information about AKP in existing BBSS or in surveys targeting AKP. In some cases, more time and expense may be needed to capture AKP. Given that survey designs need to be tailored to reach AKP, they comprise a small part of the key populations and may be more hidden than their adult counterparts. Many surveys of adult key populations include adolescents as young as 15 years [11,12]; however, disaggregated analysis that might inform donors and governments about 15-19 year olds, including disaggregation by sex, is rarely presented [14]. Other important information including the age of initiation of drug use or sexual behavior from surveys of adult key populations have been useful for advocating new policies to benefit AKP.

    As in most surveys of adult key populations, research among AKP should include the involvement of the research population in the survey planning, implementation, analysis, and dissemination. Researchers should endeavor to provide avenues to involve adolescents, especially AKP, in all phases of research as such involvement will increase their ownership of the data collected and a higher likelihood that the findings are used on their behalf to develop more useful national HIV testing, treatment, care, and support strategies. When designing research, adolescents may have insight into current behavior trends such as which drugs are being used and which venues are being frequented by AKP, and which tools and language will be most effective. In addition, older adolescents can be included as part of the data collection and analysis team. Future directions for sampling AKP could include developing novel data collection strategies such as using game interfaces with colors, sounds, and levels of earning points when responding to questionnaires in an effort to keep adolescents engaged. As more BBSS’ are conducted among AKP using TLS and RDS, it is hoped that additional lessons learned will be shared to ensure the most optimal implementation and adaptation of these methods to sample these underserved high risk populations.

    Conclusions

    Despite the fact that deaths due to AIDS continue to increase for adolescents, while they have decreased for all other age groups since 2010 [13], little data are available to ensure that adolescents are provided adequate prevention, intervention, care, and treatment services. Although there are some ethical barriers to conducting surveillance among AKP, with effort and collective action, many of these can be overcome. There are currently effective methods for collecting and using data from surveys of adult key populations that could easily be used in AKP. At a minimum, existing data of adolescents in adult BBSS should be disaggregated into adolescent age groupings. It is necessary to take the extra time and resources to effectively collect meaningful data that will provide needed strategies to address the health needs of AKP.

    Conflicts of Interest

    None declared.

    References

    1. Idele P, Gillespie A, Porth T, Suzuki C, Mahy M, Kasedde S, et al. Epidemiology of HIV and AIDS among adolescents: current status, inequities, and data gaps. J Acquir Immune Defic Syndr 2014 Jul 01;66(Suppl 2):S144-S153. [CrossRef] [Medline]
    2. UNICEF. UNICEF. 2015. Adolescents under the radar in the Asia-Pacific AIDS response   URL: http://www.unicef.org/eapro/Adolescents_Under_the_Radar_final.pdf [accessed 2017-06-04] [WebCite Cache]
    3. United Nations General Assembly. Protocol to prevent, suppress and punish trafficking in persons, especially women and children, supplementing the United Nations convention against transnational organized crime. In: Tully S, editor. International Documents On Corporate Responsibility. New York: Edward Elgar Publishing; 2000.
    4. McClure C, Chandler C, Bissell S. Responses to HIV in sexually exploited children or adolescents who sell sex. Lancet 2015 Jan 10;385(9963):97-99. [CrossRef] [Medline]
    5. Interagency Working Group on Key Populations. UNAIDS. 2014. HIV and young people who sell sex: a technical brief   URL: http://www.unaids.org/sites/default/files/media_asset/2015_young_people_who_sell_sex_en.pdf [accessed 2017-06-04] [WebCite Cache]
    6. Busza J, Mtetwa S, Mapfumo R, Hanisch D, Wong-Gruenwald R, Cowan F. Underage and underserved: reaching young women who sell sex in Zimbabwe. AIDS Care 2016 Mar;28(Suppl 2):14-20 [FREE Full text] [CrossRef] [Medline]
    7. Traore IT, Meda N, Hema NM, Ouedraogo D, Some F, Some R, et al. HIV prevention and care services for female sex workers: efficacy of a targeted community-based intervention in Burkina Faso. J Int AIDS Soc 2015;18:20088 [FREE Full text] [Medline]
    8. Curth NK, Hansson LN, Storm F, Lazarus JV. Select barriers to harm-reduction services for IDUs in eastern Europe. Cent Eur J Public Health 2009 Dec;17(4):191-197 [FREE Full text] [Medline]
    9. Family Health International. WHO. Guidelines for repeated behavioral surveys in populations at risk of HIV   URL: http://www.who.int/hiv/strategic/pubbss/en/ [accessed 2017-06-03] [WebCite Cache]
    10. UNAIDS. 2011. Guidelines on surveillance among populations most at risk for HIV   URL: http:/​/www.​unaids.org/​sites/​default/​files/​en/​media/​unaids/​contentassets/​documents/​epidemiology/​2011/​20110518_Surveillance_among_most_at_risk.​pdf [accessed 2017-06-04] [WebCite Cache]
    11. Malekinejad M, Johnston LG, Kendall C, Kerr LR, Rifkin MR, Rutherford GW. Using respondent-driven sampling methodology for HIV biological and behavioral surveillance in international settings: a systematic review. AIDS Behav 2008 Jul;12(4 Suppl):S105-S130. [CrossRef] [Medline]
    12. Montealegre JR, Johnston LG, Murrill C, Monterroso E. Respondent driven sampling for HIV biological and behavioral surveillance in Latin America and the Caribbean. AIDS Behav 2013 Sep;17(7):2313-2340. [CrossRef] [Medline]
    13. UNICEF. UNICEF. 2016. For every child, end AIDS: seventh stocktaking report, 2016   URL: https:/​/www.​unicef.org/​publications/​files/​Children_and_AIDS_Seventh_Stocktaking_Report_2016_EN.​pdf.​pdf [accessed 2017-06-04] [WebCite Cache]
    14. Dayton R, Nary P, Cunningham J, Plourde KF, Green K, Wambugu S, et al. How can we better serve adolescent key populations? strategies to encourage and inform future data collection, analysis, and use. Glob Public Health 2016 Apr 24:1-13 Epub ahead of print. [CrossRef] [Medline]
    15. UNESCO, UNFPA. UNICEF. 2013. Young key populations at higher risk of HIV in Asia and the Pacific: making the case with strategic information   URL: http://www.unicef.org/eapro/Young_key_populations_at_high_risk_of_HIV_in_Asia_Pacific.pdf [accessed 2017-06-03] [WebCite Cache]
    16. Barrett D, Hunt N, Stoicescu C. HRI. 2013. Injecting drug use among under 18s: a snapshot of available data   URL: https://www.hri.global/files/2014/08/06/injecting_among_under_18s_snapshot_WEB.pdf [accessed 2017-06-03] [WebCite Cache]
    17. UNICEF. 2007. Implementation handbook for the convention on the rights of the child: fully revised, third edition   URL: http://www.unicef.org/french/publications/index_43110.html [accessed 2017-06-03] [WebCite Cache]
    18. Powell MA, Taylor N, Fitzgerald R. UNICEF. 2013. Ethical research involving children   URL: https://www.unicef-irc.org/publications/706/ [accessed 2017-06-03] [WebCite Cache]
    19. DiClemente RJ, Ruiz MS, Sales JM. Barriers to adolescents' participation in HIV biomedical prevention research. J Acquir Immune Defic Syndr 2010 Jul;54(Suppl 1):S12-S17 [FREE Full text] [CrossRef] [Medline]
    20. Nelson RM, Lewis LL, Struble K, Wood SF. Ethical and regulatory considerations for the inclusion of adolescents in HIV biomedical prevention research. J Acquir Immune Defic Syndr 2010 Jul;54(Suppl 1):S18-S24. [CrossRef] [Medline]
    21. Fisher CB, Arbeit MR, Dumont MS, Macapagal K, Mustanski B. Self-consent for HIV prevention research involving sexual and gender minority youth: reducing barriers through evidence-based ethics. J Empir Res Hum Res Ethics 2016 Feb;11(1):3-14. [CrossRef] [Medline]
    22. Gilbert AL, Knopf AS, Fortenberry JD, Hosek SG, Kapogiannis BG, Zimet GD. Adolescent self-consent for biomedical human immunodeficiency virus prevention research. J Adolesc Health 2015 Jul;57(1):113-119 [FREE Full text] [CrossRef] [Medline]
    23. Samonte G. Capturing younger cohorts through surveillance systems: Philippines. 2012 Presented at: Methodologies for Obtaining Strategic Information on Young People at Higher Risk of HIV Exposure; September 3-5, 2012; Bangkok, Thailand.
    24. Worku EB, Davis AM, Morrow B. A critical review of health research ethical guidelines regarding caregiver consent for HIV research involving minors in South Africa: Ethical and legal issues. S Afr J BL 2016 Oct 18;9(2):78. [CrossRef]
    25. Ho WW, Brandfield J, Retkin R, Laraque D. Complexities in HIV consent in adolescents. Clin Pediatr (Phila) 2005;44(6):473-478. [Medline]
    26. UNESCO. 2013. Young people and the law in Asia and the Pacific: a review of laws and policies affecting young people?s access to sexual and reproductive health and HIV services   URL: http://unesdoc.unesco.org/images/0022/002247/224782E.pdf [accessed 2017-06-03] [WebCite Cache]
    27. World Health Organization. Youngpeopleandhiv. 2013. HIV and adolescents: guidance for HIV testing and counselling and care for adolescents living with HIV? recommendations for a public health approach and considerations for policy-makers and managers   URL: http://www.youngpeopleandhiv.org/files/HIV_Testing_guideline.pdf [accessed 2017-06-03] [WebCite Cache]
    28. Patton GC, Coffey C, Cappa C, Currie D, Riley L, Gore F, et al. Health of the world's adolescents: a synthesis of internationally comparable data. Lancet 2012 Apr 28;379(9826):1665-1675. [CrossRef] [Medline]
    29. Johnston LG, Sabin K, Prybylski D. Update for sampling most-at-risk and hidden populations for HIV biological and behavioral surveillance. jHASE J HIV/AIDS Surveill Epidemiol 2010;2(1):2 [FREE Full text]
    30. Sabin KM, Johnston LG. Epidemiological challenges to the assessment of HIV burdens among key populations: respondent-driven sampling, time-location sampling and demographic and health surveys. Curr Opin HIV AIDS 2014 Mar;9(2):101-106. [CrossRef] [Medline]
    31. Karon JM, Wejnert C. Statistical methods for the analysis of time-location sampling data. J Urban Health 2012 Jun;89(3):565-586 [FREE Full text] [CrossRef] [Medline]
    32. Globalhealthsciences. 2010. Resource guide: time location sampling   URL: http:/​/globalhealthsciences.​ucsf.edu/​sites/​default/​files/​content/​pphg/​surveillance/​modules/​global-trainings/​tls-res-guide-2nd-edition.​pdf [accessed 2017-06-04] [WebCite Cache]
    33. MacKellar D, Valleroy L, Karon J, Lemp G, Janssen R. The Young Men's Survey: methods for estimating HIV seroprevalence and risk factors among young men who have sex with men. Public Health Rep 1996;111(Suppl 1):138-144 [FREE Full text] [Medline]
    34. Heckathorn DD. Respondent-driven sampling: a new approach to the study of hidden populations. Social Problems 1997 May;44(2):174-199. [CrossRef]
    35. Heckathorn DD. Extensions of respondent-driven sampling: analyzing continuous variables and controlling for differential recruitment. ‎Sociol Methodol 2007 Aug;37(1):151-208. [CrossRef]
    36. Johnston LG. WHO. Geneva, Switzerland: World Health Organization; 2013. Introduction to respondent-driven sampling   URL: http://applications.emro.who.int/dsaf/EMRPUB_2013_EN_1539.pdf [accessed 2017-06-04] [WebCite Cache]
    37. Gile KJ, Handcock MS. Respondent-driven sampling: an assessment of current methodology. Sociol Methodol 2010 Aug;40(1):285-327 [FREE Full text] [CrossRef] [Medline]
    38. Decker MR, Marshall BD, Emerson M, Kalamar A, Covarrubias L, Astone N, et al. Respondent-driven sampling for an adolescent health study in vulnerable urban settings: a multi-country study. J Adolesc Health 2014 Dec;55(6 Suppl):S6-S12 [FREE Full text] [CrossRef] [Medline]
    39. Johnston LG, Thurman TR, Mock N, Nano L, Carcani V. Respondent-driven sampling: a new method for studying street children with findings from Albania. Vulnerable Child Youth Stud 2010 Apr 07;5(1):1-11. [CrossRef]
    40. Chemnasiri T, Netwong T, Visarutratana S, Varangrat A, Li A, Phanuphak P, et al. Inconsistent condom use among young men who have sex with men, male sex workers, and transgenders in Thailand. AIDS Educ Prev 2010 Apr;22(2):100-109. [CrossRef] [Medline]
    41. Yi S, Tuot S, Yung K, Kim S, Chhea C, Saphonn V. Factors associated with risky sexual behavior among unmarried most-at-risk young people in Cambodia. AJPHR 2014 Jan 7;2(5):211-220. [CrossRef]
    42. Kissin D. Slideserve. 2011. Using time-location sampling for HIV surveillance in street youth powerpoint presentation   URL: http://www.slideserve.com/Patman/using-time-location-sampling-for-hiv-surveillance-in-street-youth [accessed 2017-06-03] [WebCite Cache]
    43. UNICEF. UNICEF. 2014. Situational analysis of youth people at high risk for HIV exposure   URL: http://www.unicef.org/thailand/UNICEF_study_on_HIV_infection_among_young_peole_FINAL_ENGLISH.pdf [accessed 2017-06-03] [WebCite Cache]
    44. Johnston LG, Mon MM, Steinhaus M, Sass J. Correlates of forced sex among young men who have sex with men in Yangon and Monywa, Myanmar. Arch Sex Behav 2016;46(4):1001-1010. [CrossRef] [Medline]
    45. Bauermeister JA, Zimmerman MA, Johns MM, Glowacki P, Stoddard S, Volz E. Innovative recruitment using online networks: lessons learned from an online study of alcohol and other drug use utilizing a web-based, respondent-driven sampling (webRDS) strategy. J Stud Alcohol Drugs 2012 Sep;73(5):834-838 [FREE Full text] [Medline]
    46. Bengtsson L, Lu X, Nguyen QC, Camitz M, Hoang NL, Nguyen TA, et al. Implementation of web-based respondent-driven sampling among men who have sex with men in Vietnam. PLoS One 2012 Nov;7(11):e49417 [FREE Full text] [CrossRef] [Medline]
    47. Wejnert C, Heckathorn D. Web-based network sampling. ‎Socio Meth Res 2008 Aug;37(1):105-134. [CrossRef]
    48. Mirano J. UNAIDS. 2016. Enhancing interventions for young key affected populations in selected cities in the Philippines - ALL in to #End Adolescent AIDS   URL: http://www.unaids.org/sites/default/files/media_asset/20150217_ALL_IN_brochure.pdf [accessed 2017-06-09] [WebCite Cache]
    49. Lall P, Lim SH, Khairuddin N, Kamarulzaman A. Review: an urgent need for research on factors impacting adherence to and retention in care among HIV-positive youth and adolescents from key populations. J Int AIDS Soc 2015;18(2 Suppl 1):19393 [FREE Full text] [Medline]
    50. Delany-Moretlwe S, Cowan FM, Busza J, Bolton-Moore C, Kelley K, Fairlie L. Providing comprehensive health services for young key populations: needs, barriers and gaps. J Int AIDS Soc 2015;18(2 Suppl 1):19833 [FREE Full text] [Medline]


    Abbreviations

    AIDS: acquired immune deficiency syndrome
    AKP: adolescent key populations
    BBSS: biological behavioral surveillance surveys
    HIV: human immunodeficiency virus
    MSM: males who have sex with males
    RDS: respondent driven sampling
    TLS: time location sampling


    Edited by P Mee; submitted 07.02.17; peer-reviewed by B Hensen, J Busza; comments to author 09.03.17; revised version received 30.03.17; accepted 27.04.17; published 20.06.17

    ©Lisa Grazina Johnston, Justine Sass, Jeffry Acaba, Wing-Sie Cheng, Shirley Mark Prabhu. Originally published in JMIR Public Health and Surveillance (http://publichealth.jmir.org), 20.06.2017.

    This is an open-access article distributed under the terms of the Creative Commons Attribution License (https://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work, first published in JMIR Public Health and Surveillance, is properly cited. The complete bibliographic information, a link to the original publication on http://publichealth.jmir.org, as well as this copyright and license information must be included.