The methods used in the two priority-setting exercises on adolescent health research were based on the CHNRI methodology for prioritization of health research [8,14-18]. The CHNRI methodology have been described in detail, including identification of participants, generation of research questions, and scoring protocols [3,4]. The five criteria for ranking the research questions were clarity, answerability, impact, implementation, and equity. These criteria were weighted equally in the adolescent sexual and reproductive health priority-setting exercise [3] whereas they were weighted based on published guidelines from CHNRI stakeholders [15] in the priority-setting exercise for the eight other areas of adolescent health [4]. These adolescent research priority-setting exercises identified 116 top-ranked research questions (36 in the seven domains of adolescent sexual and reproductive health [3], and 80 in the eight other areas of adolescent health [4]). All top-ranked questions were included because they had been identified as important to the field of adolescent health by participants. In this mixed-methods synthesis, we analyzed these 116 research questions at the individual word level using word cloud analysis and at the thematic level using content analysis.

Based on the word cloud analysis ( Figure 1 ), the most commonly repeated words were “effects” (n = 32 mentions), “intervention” (n = 25), “school” (n = 18), “LMIC” (n = 17), “program” (n = 14), “HIV” (n = 14), “treatment” (n = 13), “risk” (n = 13), “TB” (n = 11), “prevention” (n = 10), “cost” (n = 10), “physical activity” (n = 10), “pregnancy” (n = 10), “best” (n = 9), “behaviors” (n = 9), “community” (n = 9), “strategies” (n = 9), and “mental health” (n = 9). “Gender” was mentioned on six occasions and “sex” on four occasions. “Boys” received two mentions to “girls”’ eight. Words related to maternal care for adolescents, such as “birth,” “PMTCT (prevention of mother-to-child transmission of HIV),” and “skilled birth attendants,” received 2 mentions each.

Figure 1.  Word cloud of top-ranked questions from two adolescent research priority-setting exercises.

Six intervention delivery platforms were specified in the top-ranked questions: schools (n = 17), primary care (n = 12), community (n = 11), parenting (n = 6), virtual media (n = 5), and peers (n = 2) ( Table 1 ). The specific questions for each intervention delivery platform can be found in Table S1 in Online Supplementary Document(Online Supplementary Document) . School-based interventions included instruction in skills such as swimming and water survival, sexual education, and prevention activities related to sexually transmitted infections (STI), gender-based violence, mental health, obesity and physical activity. The importance of education as a determinant of adolescent health was also underscored by top-ranked questions that related to reaching out-of-school adolescents and on decreasing school attrition with conditional cash transfer programs.

Related to the community delivery platform or setting, two questions explored the potential of community health workers, asking 1) how acceptable and accessible they are to adolescents and 2) how they might be more effectively trained on adolescent-specific gender-based violence and sexual and reproductive health issues. Other questions raised the possibility of interventions on a range of health issues – obesity prevention, physical activity, substance use, gender-based violence, sexually transmitted infection counseling and testing, and human papilloma virus vaccination – that might be delivered in community settings. Finally, several questions sought to better characterize community-level risk factors for injuries, violence, drowning, and communicable diseases.

Related to the primary care platform or setting, several questions addressed the integration of health services, including physical, mental, and reproductive (including antenatal and postnatal) health services. Another theme coalesced around the identification of evidence-based screening and prevention interventions for alcohol abuse, mental health, intellectual disabilities, and neurodevelopmental disorders. Creation of adolescent-friendly health services was also a common theme, emphasizing confidentiality, anonymity, social services and counseling, and rapid results (ie, results of laboratory testing available during the same visit when possible, such as rapid HIV testing). Finally, a major theme concerned identifying barriers to adolescents accessing services for HIV, tuberculosis, and contraception.

The final three health platforms – parenting, virtual media, and peer education – appeared less frequently but nonetheless suggested important avenues for influencing adolescents. Questions explored the role of parents in addressing mental health, substance use, and physical inactivity at home. Questions related to virtual media suggested the need for investigation of how to best use virtual media to promote healthy choices among adolescents. Peer education was proposed as a strategy for addressing substance use and improving retention in HIV and tuberculosis care, among other conditions.

Across the top-ranked research questions, several vulnerable adolescent populations were identified ( Table 2 ). The most frequently mentioned vulnerable populations included adolescents living with HIV (n = 15), tuberculosis (n = 9), and mental illness (n = 8). Questions about other vulnerable populations included victims of gender-based violence (n = 7), out-of-school adolescents (n = 6), and adolescents living in resource-poor communities (n = 3). Questions that related specifically to adolescents at risk of suicide or self-harm behaviors; adolescents living with neurodevelopmental disorders, developmental delays, and intellectual disabilities; sex workers; refugees; slum dwellers; youth in armed conflict; and young people who inject drugs also appeared among the top-ranked questions.

Of the top-ranked questions, 108 aligned with one or a combination of the Global Strategy’s Survive, Thrive or Transform agendas ( Figure 2 ). Thirty nine related to Survive; 67 to Thrive; and 28 to Transform (Table S2 in Online Supplementary Document(Online Supplementary Document) ). Those addressing the Survive agenda primarily concerned defining the most effective prevention, monitoring, and treatment strategies for communicable and non-communicable diseases. Those addressing the Thrive agenda chiefly focused on leveraging key communication channels – including peers, parents and virtual media – to promote healthy behavior. And those addressing the Transform agenda explored the impact of non-health sectors on adolescent health. Additionally, nine questions related to both the Survive and Thrive agendas; seven to both Survive and Transform; and seven to both Thrive and Transform. Questions addressing both Survive and Thrive were united by their focus on prevention, particularly in vulnerable populations, to promote wellness. Those addressing both Survive and Transform sought to identify more scalable touchpoints, from schools to primary care, for engaging adolescents in their own health. And those addressing both Thrive and Transform asked how existing health delivery platforms might be reimagined to appeal more effectively to adolescents. Three questions – seeking interventions that holistically addressed adolescent health needs – fulfilled all three agendas.

Figure 2.  Venn diagram of adolescent health research priorities placed in the United Nations Global Strategy for Women’s, Children’s, and Adolescents’ Health “Survive, Thrive, and Transform” framework.

We present the results of the first mixed-methods synthesis of multiple research priority-setting exercises based on the CHNRI methodology. We combined the top-ranked questions identified in two separate adolescent health research priority-setting exercises that comprehensively spanned multiple areas of adolescent health research in LMICs to identify key words and themes. Key areas of analysis included delivery platforms for adolescent health interventions; vulnerable adolescent populations; and alignment with the Survive, Thrive, and Transform framework of the United Nations Global Strategy for Women’s, Children’s and Adolescents’ Health.

In the word cloud analysis, the dominance of the words “effects” and “intervention” suggests that a specific type of research was prioritized, namely studies to evaluate the effectiveness of interventions. Also “schools” were frequently mentioned, suggesting an emphasis on education and on schools as a key intervention delivery platform.

Questions related to gender and sex featured prominently in the research priority-setting exercises. Several questions addressed the feasibility and impact of gender-based violence interventions in community settings and in schools. Overall, girls were mentioned four times as frequently as boys in the world cloud analysis. Questions related to adolescent girls focused on interventions to keep girls in school and to improve their antenatal, perinatal, and postnatal care during pregnancy. In addition, several questions addressed specific health issues in adolescent girls, including anemia, self-harm and suicide, and burn injuries. Only one question, related to modalities for delivering integrated HIV and family planning services, focused exclusively on boys. These questions reflect a corresponding emphasis on adolescent girls in the global health research agenda [23].

Analysis across multiple health areas and specialties allowed for the identification of themes that transcend vertical programming structures. In particular, integration of services across subspecialties and across delivery platforms was a major focus of the questions. For instance, primary care was frequently positioned as a delivery platform where mental health and sexual and reproductive health, including antenatal and postnatal care, could be addressed in integrated programs. Similarly, analysis revealed opportunities to investigate the effectiveness of collaborations between community health workers, teachers, and school health workers. Virtual media technologies were also suggested as a tool to investigate novel adolescent-focused interventions related to health education, behavior change, referral, and treatment.

Furthermore, schools emerged as an important platform to research health interventions for adolescents. To this end, top-ranked research questions addressed incentives such as conditional cash transfer programs to keep girls in schools. Other questions explored potential health topics for schools, including sexual health education, STI prevention, gender-based violence reduction, and water survival and safety. Specific interventions for research evaluation in schools included physical education classes to promote moderate to vigorous physical activity and swimming classes to prevent drowning.

Many questions were moreover united by an intent to address how best to reduce inequalities in global adolescent health. Reducing inequalities remains a major global health priority as evidenced as its inclusion as one of the United Nations’ core Sustainable Development Goals [24]. Furthermore, the Lancet Commission on Investing in Health identifies the elimination of health inequities within a generation as a means of achieving the most dramatic gains in global health by 2035, a phenomenon referred to as the “grand convergence” [25]. Critical to this ambition is addressing the health needs of the most vulnerable populations, many of which are identified through analysis of the research priorities questions. A top research priority should therefore be to assess equity of access to health interventions for these vulnerable populations.

Finally, the meta-synthesis allowed for analysis of adolescent health research through a life course approach. For instance, the current health condition of adolescents will affect their future adult health, in addition to the next generation. This life course model reflects the new Global Strategy, which for the first time includes adolescents and redefines the right to health as a three-part quest: 1) to Survive, by ending preventable causes of death; 2) to Thrive, by ensuring health and well-being; and 3) to Transform, by expanding enabling environments in partnership with multiple sectors to sustainably improve inter-generational health. Examining both research priority exercises through this lens invites researchers to imagine adolescent health as a hierarchy of needs — from medical treatment, to community wellness, to health systems sustainability — and to frame their own questions accordingly. Current questions reflect strong interest in Survive and Thrive, with perhaps less developed recognition of the utility or relevance of the Transform objective [26].