Introduction

Globally, suicide is the second leading cause of death for 15–29-year olds (World Health Organization, 2014), with an estimated 100,000 adolescent deaths and 4 million suicide attempts annually (World Health Organization, 2008). One of the key strategies to address this is the increased implementation of school-based suicide prevention programs (Cusimano & Sameem, 2011). This has resulted in such a broad range of school-based adolescent programs that identifying effective components and recommendations for future initiatives has become a significant challenge.

School has been identified as the ideal location in which to address adolescent suicide as it is regarded as a nexus for teen life (Cooper, Clements & Holt, 2011), students are a captive audience whose interactions can be mobilized around a common theme (Miller, 2014), and school personnel are increasingly cognisant of the need to identify and address the link between youth mental health problems and suicidal behavior (Davidson & Linnoila, 2013; Lake & Gould, 2011). Consequently, school-based suicide prevention programs increased rapidly during the 1980s to counter the significant rising trend in suicide rates among 15–19-year olds in many developed countries (White, Morris, & Hinbest, 2012).

A recent systematic review highlighted five distinct types of school-based suicide prevention programs: education or awareness; gatekeeper; peer leadership; skills training; and screening or assessment programs (Katz et al., 2013). Education or awareness programs familiarize students with the signs and symptoms of suicide in themselves and others. Gatekeeper training teaches natural helpers (i.e., teachers, school personnel, etc.) to recognize signs and symptoms in students and how to react effectively. Since students are more likely to confide in their peers, peer leadership training enables students to help by training them to respond appropriately and refer those of concern to a trusted adult. Skills training programs aim to indirectly prevent suicidal behavior by increasing protective factors such as coping, problem solving, decision making, and cognitive skills. Screening or assessment programs involve screening all students, identifying those at increased risk, and then recommending further treatment. Most of these programs and associated research have been designed for high school- or middle school-aged students (11–18 years old) as this is recognized as a vulnerable time that can result in mental health and academic difficulties, increased risk of suicidal ideation (Nadeem et al., 2011), and health problems in later life (Patton et al., 2012).

A review of these approaches, however, has identified a distinct lack of consensus in relation to the effectiveness across these program types (Robinson et al., 2014a, b), due in part to the complexity of suicide prevention, and to the sheer volume of such programs (Balaguru, Sharma, & Waheed, 2013; Pirruccello, 2010). In general, suicides are rare and hard to predict, and consequently it is difficult to measure the impact of programs on the prevention of suicide. Thus, many studies have focused on proximal outcomes such as knowledge and attitudes, which have unspecified relationships with actual suicidal behavior. Additionally, programs have been developed without reference to, or knowledge of, preceding interventions, resulting in a disparate field of frequently conflicting research that is of limited value to school personnel and designers concerned with implementing an effective suicide prevention program. The aim of this research was to review school-based programs to identify research gaps and best practices, in order to generate a series of key recommendations to inform the development of more effective suicide prevention programs.

Methods

This study employed a scoping review process, a method that enables the effective deconstruction of large or complex issues to promote comprehension and ease of interpretation (Arksey & O’Malley, 2005). Scoping reviews differ from systematic reviews, since the quality of included studies is typically not assessed. They also differ from narrative or the literature reviews since the scoping process requires analytical reinterpretation of the literature (Levac, Colquhoun, & O’Brien, 2010). The goal of this study was to identify research gaps and best practices in the existing literature regarding suicide prevention programs in schools, to inform recommendations for future programs. Accordingly, this study followed the five-phase framework set out by Arksey and O’Malley (2005) for this type of scoping review: identify the research question; identify relevant studies; study selection and criteria; chart the data; collate, summarize, and report the results. A range of study designs are incorporated in the review, addressing questions beyond those related to intervention effectiveness.

Identify the Research Question

The research aimed to answer the question ‘What is known about contemporary suicide prevention in schools and how can this inform future programs?’ Levac et al. (2010) recommend combining a broad research question with a clearly articulated scope of inquiry. Within this study, this refers to any school-based program relating to the prevention of suicide among school-attending adolescents (ages 11–18) and where details of the program were published in an international peer-reviewed journal.

Identify Relevant Studies

The period under review extends from January 2010 to July 2015, inclusive. This timeline was chosen to ensure that searches identified studies that reflected the most recent and contemporary approaches to suicide prevention. Online international databases searched included PsycINFO, MEDLINE, CINAHL, Cochrane Library, Google Scholar, British Education Index, Education, ERIC, OmniFile, PsycARTICLES, Sage, PubMed, Social Sciences, and relevant journals relating to suicide prevention and mental health promotion in schools. Keyword combinations (‘Suicide’ or ‘suicidal’] and [‘school’ or ‘school-based’] and [‘program’ or ‘program’ or ‘prevention’ or ‘intervention’) were used. Reference lists of relevant articles were reviewed and key articles searched for studies focusing on suicide prevention in schools.

Study Selection and Criteria

While systematic reviews develop inclusion and exclusion criteria at the outset of projects, criteria for scoping reviews are usually devised post hoc, based on increasing familiarity with the literature, and then applied to all the citations to determine relevance (Arksey & O’Malley, 2005). A total of 397 abstracts from initial searches were reviewed iteratively based upon their relevance to suicide prevention in schools. In particular, review articles were identified and analyzed in depth, allowing for the mapping of broad issues that could potentially guide a thematic analysis. Relevant studies cited within review articles were also included; reviews were excluded from final analysis to minimize bias. PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses), a tool to improve the reporting of systematic reviews and meta-analyses (Moher, Liberati, Tetzlaff, Altman, & The PRISMA Group, 2009), was used for this process.

Of the initial 397 studies identified, 324 were removed because they were not specifically relevant to suicide prevention and/or students aged 11–18 years old. The remaining 73 full articles were read and categorized based on program focus and/or type of study. Inclusion and exclusion criteria were then decided upon and applied to all citations.

Studies had to fulfill five criteria to be included in the review. Firstly, studies had to have a school-based suicide prevention program as their main focus. This narrowed the focus of the review since studies concerning general mental health programs were excluded. Secondly, the target population of programs had be adolescents (11–18 years old) attending schools in secondary education, regardless of school type (e.g., military, public, private, etc.). Thirdly, to maximize study relevance only studies from 2010 onwards were acquired. Fourthly, studies had to be reported in English or sufficiently translated. Finally, studies had to be completed.

Twenty studies met these inclusion criteria. Figure 1 shows the four-phase flow diagram that depicts the different phases of the review process and the number of records identified, included, and excluded at each stage.

Fig. 1
figure 1

PRISMA flow diagram for study inclusion in scoping review

Full size image

Chart the Data

Data were extracted from studies using a descriptive analytical method employed for scoping reviews (Levac et al., 2010). This involved synthesizing process information onto a data charting form using an Excel spreadsheet. Data extracted from studies included author(s), origin, design and results, limitations, and recommendations. Data relating to specific programs were also extracted including program name, type of program, target population, duration and frequency, requirements, delivery, delivered by, focus, and expected outcomes.

Collate, Summarize, and Report the Results

Studies were not distinguished by methodological criteria or design, nor were relative weights attributed to their data (Arksey & O’Malley, 2005). Drawing from the data charting form and the operational definitions of the five types of programs identified by Katz et al. (2013), all papers were reviewed, with a focus on program implementation. Features of programs including type of program, target population, duration and frequency, delivery, deliverers, focus, expected outcomes, origin, and studies are presented in Table 1.

Table 1 Summary of articles on secondary school-level suicide intervention programs published between 2010 and 2015 eligible for inclusion in scoping review
Full size table

Results

There were 20 studies that met inclusion criteria which included 13 distinct programs. Nine programs were universal (were for all students in a given population such as grade level, school, or district) while four were selective (were specifically developed for at-risk students). Information for each program category is discussed below in terms of implementation with reference to specific studies.

Education or Awareness Programs

Seven programs included the implementation of awareness education into curricula and demonstrated mixed results in terms of effectiveness. Only the Youth Aware of Mental Health Program (YAM) was shown to reduce suicide attempts and suicidal ideation (Wasserman et al., 2015). This program was specifically developed for the SEYLE project which aimed to investigate the efficacy of three preventative interventions for 11,110 students in 168 schools across Europe. It was facilitated in five 1-h sessions across 4 weeks focusing on raising awareness about the risk and protective factors associated with suicide, including knowledge about depression and anxiety, and skills enhancement for adverse life events, stress, and behaviors. Two programs reported improved awareness of factors associated with suicide and suicide prevention. ‘Signs of Suicide’ utilized video and guided classroom discussions over 2 days for military middle school students (Schilling, Lawless, Buchanan, & Aseltine, 2014), and ‘Surviving the Teens’ entailed four 50-min sessions over 4 days to educate students on the signs of depression and suicide through observational videos, lectures, interactive activities, and role-play. The programs also incorporate elements of screening and gatekeeper components and were feasible with support from school personnel.

An underlying assumption of education and awareness programs is that awareness of suicide is sufficient to prevent suicidal behavior. Although these programs were designed to discourage suicide and destigmatize the use of mental health services (Freedenthal, 2010; Schmidt, Iachini, George, Koller, & Weist, 2015), knowledge and attitude changes did not necessarily correlate with changes in behavior, indicating a limitation of this design (White et al., 2012). Few studies included information on specific protocols for responding to sensitive issues or in-class crises, or on the sociopolitical contexts underlying implementation (such as relations among involved teachers, counselors, community educators, and funders).

Gatekeeper Training

Four programs involving gatekeeper training were explored in ten studies (Cross et al., 2011; Freedenthal, 2010; Johnson & Parsons, 2012; Nadeem et al., 2011; Petrova, Wyman, Schmeelk‐Cone, & Pisani, 2015; Schmidt et al., 2015; Stein et al., 2010; Tompkins, Witt, & Abraibesh, 2010; Wasserman et al., 2015; Wyman et al., 2010).

Two qualitative studies (Nadeem et al., 2011; Stein et al., 2010) were conducted on the ‘Youth Suicide Prevention Program,’ an intervention delivered to almost 688,000 students in 900 schools. These identified challenges around adequate training for events during (warning signs, classroom behavior interventions, crisis management) and after the program (post-crisis challenges, limited post-referral communication), and the need for regular refresher training and information on external resources.

Mixed results were reported for the Question, Persuade, Refer program. The positive post-training findings on attitudes, knowledge, and beliefs regarding suicide were found to be moderated by a number of factors, including age, professional role, prior training, and contact with suicidal youths (Tompkins et al., 2010), while Wasserman et al. (2015) did not find any post-training reduction in the number of suicide attempts.

It was suggested that gatekeepers would benefit from additional training in identifying and responding to distressed students, and from clear and collaborative procedural guidelines for referral and follow-up (Nadeem et al., 2011).

Peer Leadership

Research into programs such as ‘Sources of Strength’ (Wyman et al., 2010) and ‘Surviving the Teens’ (Strunk, King, Vidourek, & Sorter, 2014) reported improved perceptions of adult support for suicidal youths and the acceptability of seeking help, and improved adaptive norms in relation to suicide. One factor contributing to the success of the former was the duration of the intervention, which involved biweekly 30–60-min peer-supervisor meetings over a 4 months period.

Peer support components in other suicide prevention programs have been associated with positive outcomes (Strunk et al., 2014; Wyman et al., 2010), including improved self-efficacy for students in need and their supporting peers (Miller, 2014), more positive coping norms and the ability to respond appropriately and associate with a trusted adult (Katz et al., 2013). Petrova et al. (2015) report that friends of youths who completed suicide demonstrate a unique awareness of risk factors, and posit that positive peer modeling is a promising alternative to communications that habitually focus on negative consequences and directives. However, the identification, selection, and retention of peer leaders, particularly from high-risk groups, need to be considered in future program design and studies.

Skills Training

Seven programs used a skills training approach for reducing risk factors and increasing protective factors (Hooven, Herting, & Snedker, 2010; Hooven, Walsh, Pike, & Herting, 2012; Jegannathan, Dahlblom, & Kullgren, 2014; Landgrave & Gomez-Maqueo, 2011; Schmidt et al., 2015; Strunk et al., 2014; Wasserman et al., 2015; Wyman et al., 2010). Although this approach did not directly target suicide, the goal is to prevent the development of suicidal behavior by targeting risk factors and by giving youth important skills.

Several used a longer-term intervention to indirectly influence attitudes toward suicide by promoting positive mental health in a variety of contexts. One employed a multi-disciplinary team to deliver the program over six weekly 100-min interactive and engaging sessions involving discussions, activities, and home assignments (Jegannathan et al., 2014), while another involved eight weekly modules that integrated a cognitive behavior therapy computer program alongside face-to-face assessments (Hetrick et al., 2014). The ‘Care Assess Respond Empower’ program successfully incorporated computer-assisted suicide assessment interview and a resilience-based coping and support program delivery by a mental health professional to reduce suicide risk factors and increase protective factors (Hooven et al., 2012).

Screening or Assessment

Five programs involved a screening or assessment component in their design (Hetrick et al., 2014; Hooven et al., 2012; Landgrave & Gomez-Maqueo, 2011; Schilling et al., 2014; Wasserman et al., 2015), in order to identify at-risk students and ensure they received additional support if required. In each case, these were administered only once, before the program began. While the screening process itself did not reduce suicide attempts or suicidal ideation (Wasserman et al., 2015), the ability to identify and refer those in need was identified as beneficial.

There were several issues identified with screening students. A single pre-program assessment has potential for generating false positive (Katz et al., 2013), drawing resources and attention to students who may not need it at the expense of those who may require support as the program unfolds, and other iatrogenic effects (Gould, Greenberg, Velting, & Shaffer, 2003). The process also poses considerable legal and ethical concerns (Miller, 2014), and Jacob (2009) states that schools are responsible for determining whether screening results are ‘valid, fair, and useful for identification of students at risk for suicidal behaviors, and whether the potential benefits of such screenings outweigh possible harm’ (p. 241).

Recommendations

Based on the combination of the issues identified through previous analysis and existing best practices, key issues and considerations were utilized to generate ten recommendations for designers to consider when considering a school-based adolescent suicide prevention program.

R1: Employ longer-term strategies

It is well established that unless the learner has an opportunity to reflect on the material being presented and to make it applicable to their own experiences it is unlikely to have an impact (Lonka & Ahola, 1995). For this reason, most mental health education and skills training programs last a minimum of four sessions. However, as noted from the review, targeted gatekeeper programs are often delivered in one or two sessions. Research revealed only one such program that lasted longer than two sessions (Wyman et al., 2010) and which, accordingly, demonstrated effective outcomes. Taken in conjunction with a recent review by Fountoulakis, Gonda, and Rihmer (2011), there is clear evidence that suicide programs with very short duration are not effective in reducing levels of suicide.

R2: Be aware of contextual factors

As reported in reviews of the Question, Persuade, Refer program, the context and manner in which suicide prevention programs are delivered directly impact on how participants share and use the training (Cross et al., 2011; White et al., 2012). One-off courses delivered by a non-specialist to the class as a whole will therefore be received differently to smaller, more interactive groups or discussions facilitated by a specialist in this area. These contextual factors need to be considered when deciding upon the aims and focus of programs.

R3: Clearly define learning outcomes

Although best pedagogic practices state that ‘increased knowledge’ is not a viable learning outcome (Anderson et al., 2001), analysis of existing programs revealed this to be one of the most commonly cited goals of suicide prevention programs (Johnson & Parsons, 2012; Schilling et al., 2014; Strunk et al., 2014; Tompkins et al., 2010). Suicide programs need to have clearly specified learning outcomes that state exactly what will change and/or be evident in the learner following the intervention. This clarity can only be achieved by adherence to effective pedagogic techniques such as the use of established taxonomies (Bloom, Engelhart, Furst, Hill, & Krathwohl, 1956) and constructive alignment (Biggs & Tang, 2011).

Clearly defined and observable outcomes also enable the effective evaluation of a program, a continuing challenge in suicide awareness training. Given the complexity of the issue, existing studies with poorly defined outcomes have subsequently struggled to definitely establish the impact of their programs (White et al., 2012). However, the evaluation component can be simplified by defining a small number of concise and succinct outcomes at the outset of the program design and determining if these have or have not been achieved.

R4: A preparatory phase is essential

As reported by Wasserman et al. (2012), a preparatory phase is an essential, but often overlooked component. This provides an opportunity for a site visit by the program facilitator prior to delivery to raise awareness, identify and troubleshoot potential difficulties, ensure all stakeholders are aware of the agreed-upon protocols, and to establish context (Wasserman et al., 2012). This also enables the school principal and administrators to discuss and endorse the program and to clarify its aims (Stein et al., 2010) and to affirm the importance attributed to the training. The preparatory phase also provides an opportunity to invite student feedback and input in order to directly identify and address their specified needs.

R5: Design and delivery should be flexible

Stein et al. (2010) report that programs should be designed to be flexible and to accommodate issues as they arise within the specified structure. Inbuilt flexibility permits adoption of alternative strategies to reflect unique circumstances and to tailor the program to more accurately address the needs of the audience. Accordingly, this flexibility should be incorporated into the design and delivery of the program. While there is a need to address clearly defined aims, the content should be responsive to issues that arise during delivery.

One key concern is the resistance or tension associated with discussions relating to suicide. Although this undoubtedly requires sensitivity and care (Wasserman et al., 2012), exploring this can provide ‘fertile ground or a more critically engaged pedagogy: one that invites students to consider the multiple meanings that might be available for thinking about suicide, self-other relations, [and] moral responsibility’ (White et al., 2012, p. 353).

R6: Use external, expert facilitators instead of staff

Since students are more reluctant to accept and to engage in teacher-driven interventions (Petrova et al., 2015; Wyman et al., 2010), it is recommended that, where possible, intervention programs should be delivered by external specialists or facilitators (Wasserman et al., 2015). There are several reasons for this. Firstly, related to the issue of context, it assures the students that this is being treated as a serious issue which the school wishes to take seriously. Secondly, the expertise of the specialist will ensure that any difficult questions can be answered and revelations managed in an appropriate manner. The use of a specialist facilitator may also address the issues surrounding student screening, by providing continuous observational evaluation of students’ participation and responses, and ensuring suitable interventions in conjunction with the school when deemed necessary. Finally, from an ethical perspective the specialist is in a better position to identify and support students who are affected by any issues raised during the session.

The detachment afforded by an external facilitator also provides a buffer for both the student (who can discuss issues more openly) and the teacher (who is removed from discussions and personal revelations). If staff members are involved in design and delivery, there is a need for suitable training and consultation (Hetrick et al., 2014; Landgrave & Gomez-Maqueo, 2011), regular review of outcomes and materials (White et al., 2012), and supervision (Jegannathan et al., 2014) to avoid issues such as burnout, compassion fatigue, and vicarious traumatization (Erbacher, Singer, & Poland, 2014).

R7: Don’t be restrictive

Given the complexity and interaction of factors that may lead to suicidal ideation, prevention programs should move beyond prioritizing and addressing single issues. Several studies reviewing skills-based training advocated targeting a broader range of factors to develop skills and awareness among adolescents. Suggestions include the need to promote awareness of the interactive nature of factors such as the psychodemographics associated with mental health and suicide, common myths and misconceptions, and information about national and local supports and resources (Miller, 2014; Wasserman et al., 2012). Other issues in school programs may include recognizing emotions, relationships, examining the link between thinking, feeling and acting, assertiveness training, self-talk and positive thinking, brain development (psychoeducation), issues around social media, and (un)healthy coping strategies.

R8: Don’t over-emphasize risk factors

The review of existing studies demonstrated the preponderance of risk factors associated with increased suicidal ideation, including mental health difficulties (Davidson & Linnoila, 2013), bullying (Klomek et al., 2011), sexual orientation (Mustanski & Liu, 2013), body image (Brausch & Gutierrez, 2009), stress (Wilbum & Smith, 2005), loss or bereavement (Harrison & Harrington, 2001), alcohol and substance abuse, victimization, and school problems (Borowsky, Ireland, & Resnick, 2001). Over-emphasizing specific risk factors, however, may result in overlooking others, or in under-identifying those who are at risk of making impulsive suicidal attempts (Spokas, Wenzel, Brown, & Beck, 2012).

Furthermore, while knowledge of risk factors is a vital component of prevention programs, age, gender, or sexual orientation will not change by participating in suicide prevention training. For this reason, identification of risk factors should not be the main focus of any program. It is widely accepted that within the school setting there should be a focus on building resilience in young people to enable them to cope with the various challenges they encounter during adolescence (Seligman, Ernst, Gillham, Reivich, & Linkins, 2009; White & Waters, 2015).

A number of the reviewed programs reported success in reducing risk factors and increasing protective factors (Hooven et al., 2010, 2012; Jegannathan et al., 2014; Landgrave & Gomez-Maqueo, 2011; Schmidt et al., 2015; Strunk et al., 2014; Wasserman et al., 2015; Wyman et al., 2010). Future program design, then, should integrate this evidence-based, resilience-building approach alongside multi-level and recovery-focused training. Multi-component prevention and promotion programs that focus simultaneously on different levels, such as changing the school environment, improving students’ individual skills, and involving parents, are more effective than those that intervene on only one level (WHO, 2014).

R9: Delivery should be varied, interactive, and engaging

Suicide prevention programs should avoid the pitfall of ‘death by PowerPoint’ (Kerr, 2001). Delivery methods shown to be effective include interactive workshops, discussions, group activities/exercises, booklets, posters, cards, home assignments, and video vignettes, while the development and dissemination and accessible takeaway resources have also been suggested as a means of promoting conversations between students and their parents (Freedenthal, 2010; Schmidt et al., 2015; Wasserman et al., 2012, 2015).

While the Internet has been successfully used to address other mental health issues (Calear & Christensen, 2010), its use in suicide prevention programs has been limited. When used, it has been demonstrated as effective in managing suicidal ideation and detecting and challenging problematic thinking (Hetrick et al., 2014), suggesting the potential for development as an accessible and familiar resource.

Role-play is an effective technique in assisting with suicide prevention (Cross et al., 2011; Petrova et al., 2015; Strunk et al., 2014; Wasserman et al., 2015), as it allows for the practicing of help-seeking behavior and the revision of procedural knowledge (Ornelas, 2012) in a nonjudgemental space. In addition to the interactive nature, it also has the potential to promote empathy, instigate discussion, and build confidence by experimenting with ways and words to ask for help and refer those who need help (Petrova et al., 2015).

R10: Re-evaluate program outcomes regularly

The conclusion of a training program does not signify a conclusion of learning on suicide-related issues. While practical time and resource constraints may prevent continuous, year-long, or back-to-back programs, there is benefit in regularly revisiting and re-evaluating the strategies, skills, and outcomes of previous programs. Cross et al. (2011) suggest several strategies for teachers involved in a gatekeeper training program to support maintenance of skills over time including reminders via video applications for phones, Web-based interactive practice opportunities, and providing feedback as part of a debriefing process.

Discussion and Conclusions

This scoping review sought to clarify existing research in the implementation of school-based suicide prevention programs in order to develop recommendations that would inform the development of effective school-based programs for students aged 12–18 years old. This was the first study to employ a scoping methodology to explore suicide prevention in schools. Studies were reviewed based on the five operational definitions of program types laid out by Katz et al. (2013), education or awareness, gatekeeper, peer leadership, skills training, and screening or assessment programs. Issues and considerations relating to the implementation of programs and gaps in the existing evidence provided the basis for ten recommendations for the design and delivery of a school-based, adolescent suicide prevention program.

This review should be considered as a stepping stone for alternative forms of enquiry into suicide prevention in schools. Future research should explore the implementation of other school-based mental health programs and the relationship between these and suicide prevention programs. Since most programs reviewed are from high income countries, there is a research gap on the outcome of school-based interventions among young people in low- and middle-income countries/areas. Given the global impact of suicide and its prevalence in all societies, there is an urgent need to evaluate the effects of suicide prevention programs in the context of different cultures and countries. Better understanding of factors that predict and protect against suicidal behaviors among racial/ethnic groups of adolescents is needed to identify modifiable factors and develop culturally responsive prevention and intervention strategies (Borowsky et al., 2001).

Strengths, Limitations, and Opportunities

The key strength of this study was the use of an under-used but very effective methodology to summarize a large volume of information. The comprehensive, international review of school-based programs for 11–18-year-old students identified research gaps and examples of best practices that enabled the generation of ten evidence-based recommendations for more effective suicide prevention programs for schools.

There are several limitations with the current study. Firstly, due to the nature of scoping reviews, programs and studies were not graded in terms of their effectiveness and there was a lack of consensus in specific goals which made direct comparisons challenging. Similarly, identifying gaps in the literature is impeded by the absence of a quality marker in relation to program design or to the research itself, since quality assessment does not form part of the scoping review remit. Secondly, while programs aimed at promoting general mental health may hold promise for suicide prevention, they were not included in this review. Finally, the exclusive focus on the literature published within the previous 5 years may have excluded older but beneficial studies. Future research in this area could be extended to ascertain the extent of such omissions.

This review highlights the need for future programs to have an inbuilt flexibility that accommodates issues arising throughout delivery. Future research should reflect this flexibility accordingly by expanding the range of methodologies currently pursued and in the shift from the traditional focus on predicting risk to strengthening resilience and protective factors.