Scaling-up Normative Change Interventions for Adolescent and Youth Reproductive Health: An Examination of the Evidence.

Adolescent and youth reproductive health (AYRH) outcomes are influenced by factors beyond individual control. Increasingly, interventions are seeking to influence community-level normative change to support healthy AYRH behaviors. While evidence is growing of the effectiveness of AYRH interventions that include normative change components, understanding on how to achieve scale-up and wider impact of these programs remains limited. We analyzed peer-reviewed and gray literature from 2000 to 2017 describing 42 AYRH interventions with community-based normative change components that have scaled-up in low/middle-income countries. Only 13 of 42 interventions had significant scale-up documentation. We compared scale-up strategies, scale-up facilitators and barriers, and identified recommendations for future programs. All 13 interventions addressed individual, interpersonal, and community-level outcomes, such as community attitudes and behaviors related to AYRH. Scale-up strategies included expansion via new organizations, adapting original intervention designs, and institutionalization of activities into public-sector and/or nongovernmental organization structures. Four overarching factors facilitated or inhibited scale-up processes: availability of financial and human resources, transferability of intervention designs and materials, substantive community and government-sector partnerships, and monitoring capacity. Scaling-up multifaceted normative change interventions is possible but not well documented. The global AYRH community should prioritize documentation of scale-up processes and measurement to build evidence and inform future programming.

Little evidence exists on how to integrate normative change efforts into AYRH programs at scale, outside of pilot efforts. This review examines the scale-up processes of 13 documented, successful AYRH interventions with significant normative change components and provides evidence and guidance for programmers seeking to scale-up these unique interventions.

Adolescents (aged 10–19 years) and youth (aged 15–24 years) make up one-quarter of the world's population and represent the largest cohort of young people in history [1]. Adolescents and youth (aged 10–24 years) face a multitude of reproductive health risks that, if not managed, will have consequences that follow them into adulthood. Early pregnancy and child marriage are a reality for millions, curtailing educational and vocational opportunities and contributing to intergenerational cycles of poverty [2]. Young people are still forming individual abilities, capacities, intentions, and agency. As such, they are susceptible to the influence of surrounding social systems that dictate social position and norms, the perceived attitudes or behaviors that are considered acceptable in a social group [3]. Structural barriers including lack of access to health services and economic assets further compound these factors [4].

Because adolescent and youth reproductive health (AYRH) outcomes and behaviors are influenced by the social norms outside of individual control, many believe that interventions to address AYRH must look beyond individual behavior change and seek to shift the negative normative environments that affect adolescents' well-being [5], [6]. For example, a prevalent social norm in many countries is that girls should leave school, get married, and have children early. Shifting or replacing these norms with norms that value gender equity and girls education is likely to enable girls to delay marriage and childbearing. The global health community is increasingly integrating these community-focused normative change activities, defined as “strategies designed to catalyze communities to challenge existing social norms” [7], into broader health programs. However, despite an increase in interest and implementation of these interventions, they are still a nascent area with little explicit guidance for program practice, measurement, and evaluation. As such, there remains a dearth of evidence on how best to foster normative change at scale to reach a larger population and thus achieve wider impact. Indeed, despite significant interest in scaling pilot interventions, little is known about how best to incorporate a norms focus into AYRH programs, demonstrate effectiveness of norms change or, subsequently, scale-up the normative change components of these interventions.

In 2015, the U.S. Agency for International Development awarded the Passages project to a consortium of organizations led by the Institute for Reproductive Health at Georgetown University to support development and testing of scalable approaches to foster social norms that support safe reproductive health (RH) behaviors among adolescents and youth, including delaying pregnancy and spacing subsequent births. To better understand the available evidence on the scale-up of normative change interventions for AYRH, members of the Passages project conducted a review of the current evidence base. The results are summarized here.

Literature Review

Methodology

In 2015, the Passages project Scale-up Task Team, led by Save the Children and Institute for Reproductive Health, conducted an exploratory literature review of the existing peer-reviewed and gray literature that describes the implementation or evaluation of AYRH interventions with normative change components that were in the process of, or had achieved, scale-up [7]. Combinations of terms, from three main domains—normative change, scale-up, and AYRH interventions—were used to identify relevant literature from three open-access, multidisciplinary research databases that provided access to a wide range of publications: Google Scholar, ScienceDirect, and JSTOR [6]. Relevant literature identified from these three databases was supplemented with additional literature identified through consultation with subject-matter experts. In total, 50 documents describing 42 interventions met the criteria for this initial broad review [7]. Findings from this initial review, summarized in a separate report [7], describe the community-based normative change interventions that were operating at scale to catalyze communities to challenge existing social norms that reinforce harmful attitudes and behaviors that lead to poor AYRH outcomes.

Subsequently, as described in Figure 1, we conducted a second “phase” of review, which focused on interventions from the initial 42 that could offer insight into scaling-up AYRH normative interventions. The conceptual understanding of scale-up applied throughout the initial report and this review was guided by the ExpandNet scale-up framework [8], the scale-up strategy that has been employed by the Passages project. ExpandNet defines scale-up as “deliberate efforts to increase the impact of successfully tested health innovations …to benefit more people and to foster policy and program development on a lasting basis [8]”. ExpandNet categorizes organizations with expertise in intervention implementation as “resource organizations” and the organizations that are expected to replicate the intervention at a larger scale as “user organizations.”

Figure 1

Process to identify interventions/documents for initial 2015 review and subsequent 2017 in-depth review of 42 normative interventions focused on AYRH going to scale. AYRH, adolescent and youth reproductive health.

Since many of the identified interventions did not provide documentation of scale-up efforts, this second review retained only 13 interventions for further analysis. Interventions were included in this review if they documented both their normative change components and scale-up efforts, and if they fit the inclusion/exclusion criteria detailed in Figure 1. We then searched project websites between February 2017 and April 2017 to identify additional scale-up documentation on the 13 interventions. As a team, we reviewed the available literature and participated in meetings to reach consensus on common themes, and conclusions were extracted from the interventions. The documentation from both search phases was analyzed to (1) examine common characteristics of normative change interventions that were scaled-up; (2) explore scale-up processes employed; and (3) identify factors that facilitated or inhibited scale-up. While it was our original intent to document the intervention components that contributed to normative change, and how these specific components were scaled-up, it was often unclear in the documentation which components were explicitly responsible for normative change outcomes. In addition, it is unlikely that one component acts in isolation of other components to foster normative shifts. As a result, our analysis documents characteristics of and scale-up processes employed by interventions that included normative change components rather than characteristics directly attributable to normative change components.

Results

Intervention characteristics

Table 1 summarizes the 13 interventions included: a brief description of their primary target populations and outcomes and information about where, when, and with whom they were carried out. As shown in the table, most of the interventions were implemented in Africa (n=10), and most lasted longer than 5 years from initial implementation to scale-up (n=11). In fact, the Tostan program reported a time frame of nearly 30 years over which it has been scaling-up (and adapting) its initial intervention [38]. Given our selection criteria, all 13 interventions primarily targeted adolescents and youth and engaged secondary audiences in various normative change activities. These secondary audiences were the wider community (n=13), parents and family members (n=9), health providers (n=5), teachers (n=4), and cultural or community leaders (n=3). As all 13 interventions were focused on addressing AYRH, the most common target outcomes that the interventions sought to improve were RH knowledge, attitudes, and behaviors among adolescents. Although all the interventions were selected because of their community-level normative change components, only 11 of the 13 interventions explicitly mentioned measuring attitudes, beliefs, or behaviors among community members as one of their primary outcomes.

Table 1

Description of 13 AYRH interventions with normative components that included scale-up phases

Intervention	Region/country (bolded countries indicate pilot sites)	Time frame (from testing to scale-up phases)	Intervention description (primary population and outcomes targeted)	Secondary populations reacheda
1. African Youth Alliance (AYA) [9], [10], [11]	Botswana, Ghana, Tanzania, and Uganda	2000–2005	Primary population: in-school and out-of-school boys and girls (ages 10–24)Target outcomes: improve RH knowledge, attitudes, and behaviors (including modern contraceptive use and self-efficacy in negotiating condom use) and reduce STI and HIV/AIDS transmission	Parents, teachers, community and religious leaders, health providers, policymakers, and general community
2. Gender Roles Equality and Transformation (GREAT) [12], [13], [14], [15]	Uganda	2010–2017Scale-up: 3 years	Primary population: unmarried boys and girls (ages 10–19), newly married or parenting adolescents, and their communitiesTarget outcomes: improve RH knowledge, attitudes, and behaviors; promote gender-equitable attitudes and behaviors and reduce incidence of sexual and gender-based violence	Parents, health providers, community health workers, and general community
3. Geração Biz [16], [17]	Mozambique	1999–2010Scale-up: 10 years	Primary population:in-school and out-of-school youth (ages 10–24)Target outcomes: improve RH knowledge, attitudes, and behaviors (modern contraceptive use); reduce incidence of early or unintended pregnancies; and improve gender-equitable norms	Parents, teachers, health providers, and general community
4. Ishraq Program [18]	Egypt	2001–2013Scale-up: 9 years	Primary population: out-of-school girls (ages 12–15)Target outcomes: improve RH knowledge and behaviors; improve health-seeking behavior; increase rates of school enrollment and attainment; delay early marriage and childbearing; increase girls' self-confidence; and improve gender-equitable normsProgram later added a component targeting boys (ages 13–17) and a program for graduated girls (ages 18–28)	Parents of adolescent girls, general community, and teachers
5. Kenya Adolescent Reproductive Health Project [19], [20], [21]	Kenya	1999–2008Scale-up: 2 years	Primary population: in-school and out-of-school boys and girls (ages 10–19)Target outcomes: improve RH knowledge, attitudes, and behaviors; reduce school dropout rates; improve community and parental acceptance of AYSRH	Parents, teachers, health providers, government stakeholders, and general community
6. MEMA kwa Vijana [22], [23], [24]	Tanzania	1998–2008Scale-up: 4 years	Population: primary school (grades 5–7) students (ages 10–15)Target outcome: improve RH knowledge, attitudes, and behaviors; increase contraceptive and youth-friendly service use; and reduce STI/HIV incidence	Parents, teachers, government and ministry officials, and general community
7. PRACHAR [19], [25], [26]	India	2001–2012Scale-up: 7 years	Primary population: unmarried adolescent boys and girls, young married couples, and pregnant and postpartum women (ages 12–24)Targetoutcomes: improve RH knowledge and behaviors; delay age of marriage and age at first birth; increase contraception use and healthy birth spacing; and improve gender-equitable norms	Parents and in-laws of adolescents and young couples, community leaders, general community, and health providers
8. Program H & Program M [27], [28]	Brazil, Bolivia, Colombia, Mexico, Peru, Jamaica, Nicaragua, and India	1999–2010Scale-up: 7 years	Primary population: in-school and out-of-school youth; unmarried and married youth; and lesbian, gay, bisexual, transgender, or queer youth (ages 14–24)Target outcomes: improve RH knowledge, attitudes, behaviors; reduce incidence of gender-based violence; reduce drug use; improve couples' communication; and improve gender-equitable attitudesProgram M added to reach women in 2003, then Entre Nos multimedia campaign added to complement and reach wider community	General community
9. SASA! Raising Voices [29], [30], [31]	Uganda	2008–2012Scale-up: 3 years	Primary population: youth (ages 15–24) and adult women and menTarget outcomes: improve attitudes, behaviors, and norms related to gender inequality, gender-based violence, and HIV risk	Community leaders and general community
10. Sexto Sentido (part of Somos Diferentes, Somos Iguales) [32], [33], [34], [35]	Nicaragua, Bolivia, Costa Rica, El Salvador, Guatemala, Honduras, Mexico, and USA	2000–2005Scale-up: 3 years	Primary population: adolescents and youth (ages 13–24)Target outcomes: improve RH service utilization and RH knowledge, behaviors, and outcomes; reduce STI/HIV prevalence; improve couples communication; improve gender-equitable attitudes and gender norms; and reduce gender-based violence	General community
11. South Africa Regional SBC Communication Program [36], [37]	Malawi, Zambia, Zimbabwe, South Africa, Mozambique, Lesotho, Namibia, and Swaziland	2007–2011Scale-up: 4 years	Primary population: youth (ages 15–24)Target outcomes: improve RH knowledge, attitudes, and behaviors; reduce STI/HIV incidence and prevalence; reduce gender-based violence, and reduce stigma against people living with HIV	Health providers and general community
12. Tostan (Community Empowerment Program) [38], [39], [40]	Senegal, Burkina Faso, Djibouti, The Gambia, Guinea, Mali, Mauritania, Somalia, and Sudan	1988–present (ongoing)	Primary population: adolescents and adults (ages 13 and above)Target outcomes: improve RH knowledge, behaviors, and outcomes; improve the utilization of health services; reduce incidence of female genital mutilation; delay age at first birth and age at marriage; and improve gender norms	Parents of girls and general community
13. Young Empowered and Healthy Initiative (YEAH) (part of Health Communication Partnership) [41], [42]	Uganda	2004–2013	Primary population: adolescents and youth (ages 15–24)Target outcomes: improve RH knowledge, attitudes, behaviors, and outcomes; improve school enrollment; improve gender-equitable attitudes; and reduce incidence of gender-based violence	Adult men (ages 15–55), parents of adolescent girls, police force, and general community

AYSRH = adolescent youth sexual reproductive heath; RH = reproductive health; STI = sexually transmitted infection.

Populations noted in italics are those that are not explicitly stated as target populations in the documentation but are nevertheless referred to in the documentation as having benefitted from or been affected by the programs.

Table 2 provides a breakdown of the specific components implemented by each intervention as well as the type of key individual or community-level outcomes that each intervention measured. Community-level refers to activities that target groups in communities other than the primary target group of adolescents or youth whose behavior or health outcomes the program is seeking to shift. Thus, community-level outcomes measure change in secondary populations, such as parents, teachers, or health providers. All 13 interventions applied a multicomponent approach and implemented at least three intervention components. Many included curriculum-based family life education (n=4), peer education and support (n=5), or the creation of adolescent safe spaces (n=10). Social and behavior change communication activities, from interpersonal communication to mass media campaigns, were also popular (n=10). The mass media campaigns often accompanied community-level activities to expand an intervention's reach and mobilize action in communities [27], [32], [36], [41]. All 13 interventions encouraged community dialog on norms and AYRH topics through community group engagement activities. Many interventions also included structural components that aimed to strengthen youth-friendly health services (n=6) to improve access to quality services and to address policy through advocacy with government stakeholders (n=7). Some interventions also built the capacity of local partners (i.e., government and nongovernmental organizations [NGOs]) to manage intervention components, as part of their scale-up and sustainability efforts (n=9).

Table 2

Strategies utilized and key outcomes measured by included interventions

	Intervention components								Adolescent and youth outcomes			Secondary population attitudes, beliefs, or behaviors	Normative change findings or results
	FLE	Peer education and support	Adol. safe spaces	SBCC	CGE	HSS	Capacity-building of user orgs	Policy and advocacy	RH knowledge, attitudes, skills, or intentions	Behavior change	Biological health outcomes
1. AYAa[9], [10], [11]		✓	✓	✓	✓	✓	✓	✓	+	+		+	No explicit evaluation of norms. Implied change due to improved supportive ARH policies and support for ARH and YFHS among community members, parents, and AY.
2. GREAT [12], [13], [14], [15]			✓	✓	✓	✓	✓		+	+		+	Improved gender-equitable norms among community members, parents, and AY
3. Geração Biz [16], [17]	✓	✓	✓	✓	✓	✓	✓	✓	+			0	No explicit evaluation of norms. Implied change among health providers due to improved quality and use of YFHS. Implied gender norms did not significantly change among AY.
4. Ishraq Program [18]		✓	✓		✓		✓	✓	+	+		+	Improved gender-equitable norms among participants, parents, and community leaders.
5. Kenya ARH Project [19], [20], [21]	✓	✓	✓		✓	✓			+	+		+	Improved parent-child discussions on SRH and norms related to discussing ARH topics among community members.
6. MEMA kwa Vijana [22], [23], [24]	✓				✓	✓			+	+	0	+	Improved norms related to discussing SRH with AY among teachers and health workers. Implied change due to increased community support for FLE for unmarried AY.
7. PRACHAR [19], [25], [26]	✓		✓	✓	✓		✓	✓	+	+	+	+	Improved norms to delay child marriage and childbearing among AY and support from parents.
8. Program H & Program M [27], [28]				✓	✓		✓	✓	+	+		+	Improved gender-equitable norms among community members and AY.
9. SASA! Raising Voices [29], [30], [31]				✓	✓		✓	✓	+	+		+	No explicit evaluation of norms. Implied improved gender-equitable and SRH norms related to GBV among community members and AY due to reduction in GBV and more equitable behaviors and attitudes among community members.
10. Sexto Sentido [32], [33], [34], [35]			✓	✓	✓		✓		+	+		+	Improved gender-equitable norms and norms related to sexuality among community members and AY
11. South Africa Regional SBC Communication Program [36], [37]			✓	✓	✓	✓			+	+		+	Improved gender and SRH norms related to gender equity, GBV, and HIV
12. Tostan [38], [39], [40]		✓	✓	✓	✓		✓	✓	+	+	+	+	Improved gender norms related to FGM to reduce FGM prevalence among community members and parents
13. YEAH [41], [42]			✓	✓	✓				+	0		+	Improved gender and RH norms related to IPV and HIV among community members and AY

Blank = not utilized or measured; + = positive significant change; 0 = no change in outcome.

FLE = curriculum-based reproductive health (RH) education for both in-school and out-of-school populations; Adol. safe spaces = any mention of the creation of a safe physical or emotional space for adolescents (both same-sex and mixed-sex groups) to congregate or discuss AYSRH topics; SBCC = individual-level counseling and education and mass media campaigns; CGE = activities to engage or mobilize communities in group dialogs and action to promote behavior and attitude changes [43]; HSS = strengthening of and community linkages to YFHS; policy and advocacy efforts = any efforts with government stakeholders to create enabling and supportive policies to support AYSRH and rights; RH knowledge, attitudes, skills, or intentions = changes in knowledge/attitudes/skills related to RH topics such as family planning methods, STI and HIV prevention, anatomy, and puberty; behavior change = changes in reported health behaviors such as family planning use, use of health services, partner violence, school attendance, early marriage rates, and couples communication, or decision-making; biological health outcomes = changes in rates of early pregnancies, STI prevalence, prevalence of female genital mutilation, and so forth; secondary population attitude/beliefs/behaviors = changes in attitudes or behaviors related to gender equity, gender-based violence, AYRH topics, parent-child communication, HIV stigma, early marriage and pregnancy, and so forth, among the secondary population.

ARH, adolescent reproductive health; AY = adolescents/youth; CBO = community-based organization; CGE = community group engagement; FGM = female genital mutilation; FLE = family life education; GBV = gender-based violence; HSS = health systems strengthening; IPV = intimate partner violence; SBCC = social and behavior change communication; SRH = sexual reproductive heath; YFHS = youth-friendly health services.

Intervention strategies varied across country programs. In Tanzania, the intervention included a component that integrated into livelihoods programs. In Uganda, the resource organization partnered with CBOs and religious institutions to implement various intervention strategies to support adolescents.

Significant improvements in RH knowledge, attitudes, and skills among adolescents were reported by all interventions, and improvements in behavior changes, such as use of contraception, school attendance, and marriage rates, were reported for 11 interventions. Documentation from MEMA kwa Vijana, PRACHAR, and Tostan measured changes in biological health outcomes, but only PRACHAR and Tostan demonstrated improvements in outcomes, such as reduced pregnancy rates and incidence of sexually transmitted infections [22], [25], [38].

Although only 11 of the interventions described efforts to target a normative outcome in their intervention design, all 13 interventions measured changes in the attitudes, beliefs, or behaviors of the secondary populations, thereby acknowledging the influence of the secondary populations' attitudes or norms on AYRH outcomes. Although not specified in Table 2, the interventions that measured the attitudes and beliefs of the secondary populations measured three types of change: attitudes toward AYRH topics and behaviors (n=11), gender-equitable attitudes (n=9), and acceptance or incidence of gender-based violence (n=7).

Scale-up strategies

Table 3 summarizes the application of frequently used scale-up strategies: (1) expanding to a larger geographic region, (2) expanding to more user organizations (to achieve greater reach or depth), (3) adapting program content and range of services offered, (4) adapting program design to reach new populations, and (5) institutionalizing the intervention into the public sector. As reflected in Table 3, 11 of the 13 interventions employed more than one of these strategies to achieve wider impact. One intervention, MEMA Kwa Vijana, lacked documentation related to the scale-up of its normative change components. However, the intervention was included in the review because it used a normative change strategy during pilot implementation and documented the scale-up of all activities besides the normative activity. Although the African Youth Alliance (AYA) intervention was implemented in four countries, the only scale-up documentation on the normative components available was specific to implementation in Uganda [12].

Table 3

Five types of strategies utilized to scale-up normative components in the 13 included interventions and prevalence of each

	Expanding to a larger geographic region in-country or replication in new countries	Expanding to more user organizations (e.g., local NGOs/community-based organizations, or international NGOs)	Adapting program design to increase depth and scope of the services offered	Adapting program design to reach new primary populations	Institutionalizing the intervention into the public sector
No. of interventions utilizing this strategy	11	12	5	2	7
Intervention name
1. AYAa[9], [10], [11]		✓
2. GREAT [12], [13], [14], [15]	✓Scale-up to new districts	✓			✓
3. Geração Biz [16], [17]	✓Nation-wide scale-up	✓	✓		✓
4. Ishraq Program [18]	✓Scale-up to new villages	✓	✓	✓	✓
5. Kenya ARH Project [19], [20], [21]	✓Scale-up to new provinces	✓	✓		✓
6. MEMA kwa Vijana [22], [23], [24]	While scaling-up, this program eliminated the normative component of the program due to challenges related to continuing community-level activities at wider scale.
7. PRACHAR [19], [25], [26]	✓Scale-up to new districts	✓			✓
8. Program H & Program M [27], [28]	✓Scale-up to new countries	✓		✓	✓
9. SASA! Raising Voices [29], [30], [31]	✓Scale-up to new countries	✓
10. Sexto Sentido [32], [33], [34], [35]	✓Scale-up to new countries	✓
11. South Africa Regional SBC Communication Program [36], [37]	✓Scale-up in multiple countries	✓
12. Tostan [38], [39], [40]	✓Scale-up to new countries	✓	✓
13. YEAH [41], [42]	✓Nation-wide scale-up	✓	✓		✓

ARH = adolescent reproductive health; AYA = African Youth Alliance; GREAT = Gender Roles, Equality and Transformation; SBC = social and behavior change; YEAH = Young Empowered and Healthy Initiative.

Blank = available program documentation did not mention the category as a scale-up strategy utilized.

Available documentation specific to scale-up experience in Uganda.

Interventions expanded geographically, either in the same country where the pilot occurred or through replication in new countries. The extent of geographic expansion varied. Four interventions (i.e., Gender Roles, Equality and Transformation [GREAT]; Ishraq; Kenya Adolescent Reproductive Health Project [KARHP]; and PRACHAR) were each piloted and scaled-up to additional districts within the same country [13], [18], [19], [25], [39]. Six interventions scaled-up internationally. Program H and Sexto Sentido were piloted in one country and then replicated in new countries by user organizations [27], [32]; two interventions expanded within the pilot country first, and then replicated in new countries by new user organizations (i.e., SASA! Raising Voices, Tostan) [29], [38]; and two were implemented as pilot interventions in multiple countries from the beginning, scaling-up through different partners in each country (i.e., AYA, Southern African Regional Social and Behavior Change Communication Project) [20], [36].

The most common scale-up strategy was to increase the number and capacity of user organizations during and after pilot implementation to create wider networks of organizations with the capacity for broad reach and impact. For example, PRACHAR built the capacity of local NGO and government partners as user organizations at every stage of its planned expansion through a “learning-by-doing” approach, building skills beyond those related to reproductive health topics. The resource organization held classroom orientations, conducted field exposure visits, and provided support to build the competencies of the user organizations' staff in project management, budgeting, and monitoring [25].

To facilitate scale-up in new contexts, five interventions adapted activities, increasing the depth of services provided, to better meet needs of new user organizations and communities. The KARHP, for example, tested a streamlined package of activities that included successful components of its intervention during adaptation, and then added additional community outreach and income-generation activities to better support its target populations, based on feedback from the community [19]. Two interventions also adapted activities to reach new primary populations. In the case of Ishraq, the resource organization added two additional program components requested by the communities; a parallel life skills and sexual reproductive heath program for adolescent boys and services for older female graduates who needed support to transition into the formal schooling system [18].

Seven of the interventions documented efforts to work with government ministries to be institutionalized into the public sector. The implementing organizations did so by delegating responsibilities for managing at least one component of the intervention to a relevant government entity. In these cases, to ensure the impact observed during the pilot could be replicated and sustained through the public sector, the resource organizations worked with government stakeholders to build their capacity to plan, budget, and manage implementation of activities under existing government initiatives. For example, Geração Biz was designed and implemented with the Mozambican Ministries of Health, Education, and Youth and Sports. The resource organization worked closely with the Ministry of Health and the Ministry of Education to manage the youth-friendly health services and in-school family life education components of the intervention. The Ministry of Youth and Sports, responsible for managing the community-based peer education activities, received support to develop the peer education materials and manage community group events. The resource organization also supported capacity building of relevant provincial-level offices within each ministry and facilitated intersectoral meetings across the ministries to ensure that activities were integrated into each ministry's operating budget [16].

The five interventions that did not attempt institutionalization within the public sector focused instead on building the capacity of NGO user organizations to budget for, manage, and monitor activities. SASA! Raising Voices (SASA!) and Sexto Sentido, for example, both identified and built networks of local community-based organization and NGOs that could lead implementation of the community-level activities, and they both provided ongoing technical assistance (e.g., trainings, guidance on monitoring tool development) to support the user organizations to work with youth and address RH issues in their own communities throughout the pilot and during scale-up [39], [41]. For example, in lieu of public-sector program standards to guide quality implementation, SASA! also developed an implementation toolkit to help NGO user organizations manage program activities [39].

Factors facilitating and challenging scale-up processes

Across the reviewed literature, we identified four categories of factors that were mentioned as facilitators or challenges to scale-up success: (1) resource needs, (2) intervention design, (3) partnerships for sustainability, and (4) monitoring and evaluation (M&E) systems and data. Distribution of the documentation of these factors across interventions is depicted in Table 4. In some cases, it was difficult to distinguish whether the facilitators and challenges discussed were related to the success of implementation generally or to the scale-up process more specifically.

Table 4

Factors identified as facilitators or challenges to scale-up efforts of normative strategies of each of the 13 included interventions

	Resource needs		Intervention design		Partnerships for sustainability		Monitoring and evaluation systems and data
	Financial resources	Human resources	Content and structure	Adaptability of programming	Community support and engagement	Government support and ownership
No. interventions that cited a facilitating factor	7	4	7	3	10	9	7
No. interventions that cited a challenging factor	5	5	4	1	3	3	2
1. AYAa[9], [10], [11]
Facilitators	Advocacy and partnerships with Uganda Kingdoms led to select Kingdoms securing financial resources to take on project initiatives				Communities (including religious institutions) participated in all stages of programming, building capacity to analyze and address AYRH issues	Policymakers involved in all stages of programming, and partnerships with Uganda Kingdoms created supportive AYRH policies
Challenges	No challenges to scale-up documented
GREAT [12], [13], [14], [15]
Facilitators	Used a “low-investment approach” design and user organizations could leverage financial resources to integrate GREAT components into existing programming	Building capacity of staff to understand own gender norms supported community-level work, building sustainability of activities. Resource organization prepared for transition as implementer to capacity builder, provided mentoring to user organizations to lead activities	Conceptualized with “scale in mind”; developed a toolkit with guides that can be easily used by user organizations; worked through existing community mechanisms		Received positive support from community members; active and early engagement with potential user organizations helped build local ownership and sustainability of GREAT components	Assigned scale-up coordination responsibilities to MOH and district stakeholders, thus ensuring ownership of scale-up	Partnered with user organizations and stakeholders to develop monitoring, evaluation, and learning system and indicators in line with district databases and M&E systems
Challenges	The Community Action Cycle component was difficult for user organizations to understand and required repeated trainings and capacity-building initiatives	Existing village health teams were overworked and resource organizations experienced high staff turnover			Not enough community participation necessary to achieve wide diffusion and reach the tipping point for social normative change		User organizations needed capacity building from the resource organization to support M&E system
2. Geração Biz [16], [17]
Facilitators	User organizations could continue activities through integrating program costs into operating budgets				Local user organizations expressed interest and could integrate program costs into operating budgets	Government showed commitment and ministries were involved in development and implementation of intervention	Availability of M&E data helped adapt activities and developed M&E system to be adaptable for user organizations
Challenges	Costs to implement across sectors and at various administrative levels were substantial	High staff turnover, requiring follow-up and additional technical assistance from the resource organization. Gender inequity among peer educators and inadequate gender sensitivity training may have affected program effect on social normative change					M&E systems were inconsistent across provinces, requiring significant time and support from resource organization
3. Ishraq Program [18]
Facilitators			Created steps to integrate graduates into formal schooling and existing systems	Activities easily fit into government systems and initiatives	Local communities maintained support and demand for project to continue and were very involved in community activities	Government ministries involved in design and implementation; increased attention to improving AYRH	Rigorous M&E system allowed for effective learning and implementation of adjustments to streamline activities
Challenges	Cost of providing continued support to graduates needed to be raised from local funds		Graduates aged out of formal program and required additional support			Lack of government legal records and documentation for graduated girls made it difficult to access public services
4. Kenya ARH Project [19], [20], [21]
Facilitators	Costing activities helped to identify essential program components for replication and MOH could leverage resources to integrate activities in existing initiatives		Availability of implementation tools and guidance documents facilitated transition to user organizations		Local community expressed high demand and was very engaged with community activities	Supportive government policies brought attention to project and integration of various intervention components into MOH initiatives	Strong pilot data and dissemination showcased evidence and generated buy-in to adapt and refine for scale-up
Challenges	Lack of sufficient resources for all components	High turnover of relevant staff required high level of continued external technical assistance and additional retraining				Integrating activities into ministries was difficult due to the complex government systems
5. MEMA kwa Vijana [22], [23], [24]	Scale-up of normative components not documented
6. PRACHAR [19], [25], [26]
Facilitators		Building capacity of local NGO staff and community members who led activities to understand own norms and internalize their role as change agents enhanced performance		Adaptable activities and systems to respond to the needs of community and user organizations	Communities were engaged in activities; consistent partnerships with local user organizations from the start fostered commitment		Rigorous M&E data showed evidence of project impact, which generated local support and demand
Challenges			Multiple components were too large for public sector, requiring refinement/adaptation
7. Program H & Program M [27], [28]
Facilitators	Resource organization budgeted for capacity building of user organizations as part of scale-up efforts and made materials available at no cost		Developed materials for user organizations to adopt and made them readily available		Communities showed strong interest and engagement and built capacity of user organizations as part of activities and program costs	Initiated early engagement with government stakeholders and supported government to integrate project activities into ongoing initiatives	Rigorous data and results from adaptations in multiple countries demonstrated programs' effectiveness
Challenges					Recruitment and commitment of participants due to competing priorities was difficult
8. SASA! Raising Voices [29], [30], [31]
Facilitators	Discussion leaders were unpaid volunteers but still showed commitment and engagement; the resource organization made online trainings and program materials available to user organizations at no cost	Program addressed social norms of staff and volunteers first, empowering them to take action and building their commitment to community mobilization activities	Intervention focused on empowerment rather than negative behaviors	Developed an open-source toolkit that is publicly available and freely distributes supplementary materials and online trainings	Messages diffused outside of target population showing strong interest among participants; community advocacy activities built support among user organizations	Fostering relationships with and support from local government leaders built interest and support of activities	M&E tools developed are easy to use and strong impact demonstrated
Challenges	Short-term donor cycles cited as a barrier to achieving the long-term normative change necessary to replicate impact at scale	Difficult to monitor use of freely available materials to ensure fidelity to core components	Community mobilization process can be difficult and costly
9. Sexto Sentido [32], [33], [34], [35]
Facilitators			Availability of telenovela episodes and group discussion materials for user organizations		Strong partnership and support from civil society organizations that became user organizations; target populations generated demand for program	Supportive policy environment with government ownership
Challenges	No challenges to scale-up documented
10. South Africa Regional SBC Communication Program [36], [37]	Facilitators and challenges noted were related to pilot implementation and not specifically to scale-up efforts.
11. Tostan [38], [39], [40]
Facilitators	Resource organization accounted for costs related to capacity building and mentoring of user organization staff	Resource organization mentored and built capacity of user organizations to manage program and understand underlying norms	Content avoided focus of negative behavior; focus on noncombative manner reinforced women's empowerment messages		Community showed enthusiasm for activities; inclusion of capacity-building activities with local user organizations built local ownership	Eventually gained support from government bodies that made public declarations to end female genital cutting
Challenges		Difficult to find local residents to serve as facilitators, increasing program costs	Some content was too difficult for facilitators to discuss, leading to changes in core program components and messages	The complexity of female genital mutilation norms in countries where practice is universal made it difficult to initiate behavior change	Opposition from some community and religious leaders; lack of community participation without tangible incentives	Some countries faced challenges gaining support from government stakeholders at start
12. YEAH [41], [42]	Facilitators and challenges noted were related to pilot implementation and not specifically to scale-up efforts.

Blank = available program documentation did not mention the category as a facilitator of their scale-up effort.

ARH = adolescent reproductive health; AYA = African Youth Alliance; AYRH = adolescent and youth reproductive health; GREAT = Gender Roles, Equality and Transformation; M&E = monitoring and evaluation; MOH = Ministry of Health; SBC = social and behavior change; YEAH = Young Empowered and Healthy Initiative.

Available documentation specific to project scale-up experience in Uganda.

Resource needs

Eight of the interventions identified the need for financial and human resources to support scale-up. In most instances, the documentation referenced financial resources as a facilitator to scale-up as resource organizations supported user organizations to incorporate activities into their operational budgets. The development of low-cost materials was also cited as a facilitator to scale-up. For example, SASA! was able to reduce the financial burden among user organizations by providing free online program materials and tools, but because it did not have staff who could monitor the use of materials, the organization could not ensure fidelity to the intervention's core components [39]. GREAT's “low-investment approach” design of a toolkit and activities that could be adopted by existing community groups limited the financial burden placed on user organizations to replicate activities, enabling them to leverage their own resources to integrate GREAT's activities into existing initiatives [12].

Human resources were most often identified as facilitators when resource organizations successfully built the capacity of user organization staff to ensure program fidelity and manage activities. Tostan, for example, mentors and trains user organizations to manage activities in their own communities to support scale-up [39]. In addition to ensuring the technical capacity of staff in user organizations, many resource organizations supported staff to examine and reflect upon their own values and norms [39]. For instance, documentation of the PRACHAR project stressed the importance of training project staff and volunteers to reflect on their own norms, to be empowered to take action and build their commitment to the project's objectives [22]. Not surprisingly, a lack of financial or human resources was often cited as a challenge to scale-up efforts. High staff turnover among user organizations, mentioned by three interventions, required resource organizations to invest additional time and financial resources to train new staff [12], [16], [18].

Intervention design

Intervention design appeared to play a key role as both a facilitator and an impediment to scale-up. Seven interventions noted the content or structure of their intervention as a facilitator of scale-up. The KARHP and GREAT are two such examples. Both planned for and developed implementation toolkits or guidance materials during pilot implementation, which later facilitated scale-up through user organizations [12], [18].

Four interventions cited a need to further modify intervention design during the scale-up phase as a challenge. In some cases, the intervention activities implemented during the pilot phase were too costly for user organizations to continue or replicate. For example, in response to concerns that the SASA!'s community mobilization process was difficult and costly, it created nonmonetary incentives to engage volunteers for the project. The nonmonetary incentives took the form of capacity-building opportunities as well as the ability to obtain recognition from peers for serving as change agents in the community [39]. Such adaptability of intervention activities was also mentioned by three interventions as a facilitating factor to scale-up. PRACHAR and Ishraq both noted that the ability to adapt activities, either to better address community needs or to simplify processes, facilitated scale-up [13], [22].

Partnerships for sustainability

Partnerships with and support from community groups and government stakeholders were the most frequently mentioned facilitators of scale-up. Documentation from 10 interventions noted that community engagement and support for the AYRH interventions facilitated not only pilot implementation efforts but also the scale-up process by fostering trust and ownership among the organizations that would become user organizations in the expansion process. Tostan, for example, noted that their success in building the necessary critical mass needed for social change was achieved through capacity building of and support from the local committees through conducting village meetings during pilot implementation [8]. Other activities to engage community stakeholders included participatory activities to identify and address local AYRH issues, consultations with stakeholders to inform intervention design, and establishing mechanisms to receive and share data with communities about ongoing activities.

Advocacy and partnerships with government stakeholders were also noted by implementers as a facilitator to scale-up. Documentation from nine interventions mentioned that partnerships with government ministries at local or national levels and capacity-building activities with government partners supported pilot implementation and eventual scale-up efforts. AYA, Geraçao Biz, Ishraq, and Project H either implemented through government partners from the start, or began collaborating with them early on during implementation, to integrate activities into government systems and ensure activities aligned with government strategic priorities. In turn, this built ownership of programming and intervention results before “handover” to the government. AYA in Uganda engaged policymakers and community leaders, including representatives of four Kingdoms, in all phases of intervention design. The project also partnered with the Kingdoms to implement the community-level activities, which translated into the Kingdoms adopting supportive AYRH policies in their agendas and securing funding to continue the initiatives started by the resource organization [25].

Notably, interventions that mentioned community or government support as a facilitator to scale-up also mentioned lack of support as a challenge to scale-up. Five of the interventions identified lack of community or government support or adaptability as a challenge. For instance, the KARHP noted that despite substantial interest from the government to adopt supportive AYRH policies, because of the complex budgeting and planning process, the resource organization still struggled to integrate activities into the government system and had to adapt activities to accommodate systems, underscoring that supportive policy environments alone do not facilitate the sustainability of intervention impact [18].

Monitoring and evaluation systems and data

The monitoring and use of data was mentioned as a facilitating factor for scale-up by seven of the interventions. Notably, the interventions that identified M&E capacity as a facilitator were often interventions that demonstrated evidence of impact during the pilot phase. PRACHAR and the KARHP noted that the availability of evaluation data showing intervention impact from pilot implementation, particularly regarding the importance of the normative components, provided the evidence needed to generate buy-in from user organizations [19], [20]. Several interventions faced challenges related to the M&E capacities of user organizations, which in some cases were unable to replicate the M&E systems developed by the resource organizations. For instance, Geraçao Biz found that the M&E capacity across scale-up locations was inconsistent. To address this, the intervention conducted periodic evaluations to improve the M&E system, which was designed to be adaptable and thus could easily be integrated into the systems of varying capacity [16].

Discussion

This exploratory review of peer-reviewed and gray literature pertaining to the scale-up of normative change interventions for AYRH identified only 13 interventions that both met our definitions of normative change and scale-up and provided documentation of their scale-up efforts. The 13 interventions we analyzed used a variety of scale-up strategies across diverse contexts and time frames with different scale-up goals. As the language on scale-up and normative change varied, and the interventions were multifaceted, it was difficult to separate which components contributed specifically to normative change and to assess how normative change outcomes were evaluated. Despite these limitations, we discerned many elements common to scale-up success and several unique considerations for the scale-up of normative change interventions.

Many of the interventions planned for scale-up during the pilot phase, citing early preparation as a critical factor in their success for later expansion and institutionalization. This preparation took many forms. Some resource organizations developed a strategic scale-up plan, while others sought to ensure community and government stakeholder buy-in through advocacy and early engagement. Many interventions incorporated measures to align intervention components with government policies, systems, or NGO platforms so that the interventions could be easily integrated into existing programs. Other organizations budgeted capacity-building activities for user organizations to independently implement the interventions over time. Unique to normative change interventions, working with staff to identify and clarify their own norms and roles as change agents was emphasized by many as a critical component to successful implementation and scale-up.

Social norms are highly contextual. Thus, program adaptability was highlighted across the reviewed literature as a facilitator of scale-up. Interventions identified for this review were almost always adapted when scaled-up in new contexts. Guidance documents and tools, combined with capacity building of user organizations, were identified as critical supports to maintain fidelity of the core normative change components during scale-up.

The complexity and use of M&E systems was also cited by about half of the interventions as important for guiding scale-up efforts. When scale-up involved cross-organizational monitoring, it was useful for multiple organizations to share core indicators. Often, however, monitoring systems developed for pilot implementation needed to be adapted or simplified to accommodate new contexts and organizational systems during scale-up.

We note that the development of tools and the adaptation of systems often require significant initial investment and is likely to add to intervention timelines. Indeed, since changing social norms requires changing the beliefs of many individuals, the time frame for reaching tipping points and demonstrating effectiveness of normative change efforts is likely to require longer than the standard three- to five-year time frames of most health-focused projects. Advocacy is needed to increase awareness of these longer term resource needs, especially if normative change at scale is the ultimate goal.

Although all but one of the interventions documented a change in the attitudes, beliefs, or behaviors of individuals in the larger community, clear measures of normative change outcomes were notably lacking from the documents reviewed. In most cases, documentation alluded to the assumption that the mechanisms of norms change were effective due to changes in health outcomes, but norms change itself was rarely evaluated. Documentation of efforts to confirm findings and assess normative change with communities was lacking as well. The articulation and measurement of social norms need to be given considerably more attention. This includes being more explicit about what norms are expected to change and about how norms change will be monitored over time during pilot implementation and under scale-up conditions. To this latter point, careful M&E is needed to ensure that the normative change mechanism inherent in norms-focused activities continues to operate at scale. Currently, for instance, it is unclear how much interventions can be adapted before effectiveness must be re-evaluated. Unfortunately, documentation of how to monitor the normative change components of AYRH interventions is sorely lacking. Simple indicators and adaptable approaches to measuring normative change at scale are needed [3], [44].

We note the following limitations. Our review did not include information about on-going scale-up activities. The lack of a common language when referring to normative interventions or scale-up made it difficult to ascertain if challenges or facilitators were related to scale-up or to the implementation of the pilot itself. As mentioned previously, documentation of the scale-up process of normative change strategies, even among the interventions included in our review, was limited. Most of the reviewed documentation of challenges or facilitators focused on describing elements of the strength of the intervention itself and whether impact was achieved. Projects tended to provide minimal description or analysis of their scale-up experiences, and even less description of the process of scaling-up their normative change strategies specifically. Documentation regarding measurement and evaluation of normative outcomes, whether during pilot or while operating at-scale, was also lacking.

Summary and Implications

The ability to scale-up community-based normative change interventions is commonly questioned. However, the 13 interventions included in this review demonstrate that the scale-up of multicomponent community-based normative change interventions is feasible. They also show that the scale-up of such interventions requires planning and considerations that are distinct from those required for the scale-up of traditional behavior change approaches because they seek to influence change at both individual and community level and are highly contextual to complex social environments. The success of scaling normative change interventions is facilitated by planning for scale-up from the beginning, even before evidence of effectiveness indicates an intervention is worthy of going to scale.

The few interventions included in this review and the scant documentation of scale-up processes highlight the need for more research and evaluation, as well as better articulation and documentation of scale-up and lessons learned. In addition, greater shared learning across the many organizations that are implementing normative change interventions is needed to improve measurement and analysis of normative change and scale-up, to ultimately ensure sustained impact of these initiatives. Building the evidence base for effective approaches for shifting social norms and creating enabling environments for behavior change at scale is crucial if the field is to meet the large and growing RH needs of adolescents. Despite the need for more evidence, the insights gleaned from this review provide an important starting point to inform future normative change programming for AYRH and have broad applicability to other health sectors.