A school-based health promotion programme to increase help-seeking for substance use and mental health problems: study protocol for a randomised controlled trial

Adolescence is a high-risk time for the development of mental health and substance use problems. Indeed, half of all lifelong mental disorders (including substance use disorders) commence by the age of 14 years, with three quarters beginning before the age of 25 [1]. In Australia, over a quarter of 16–24 year-olds meet criteria for a mental disorder in the previous 12 months, with anxiety (15.4 %), depression (6.3 %) and substance use disorders (12.7 %) being the most commonly experienced conditions [2]. Despite cross-sectional data showing an increase in non-drinking among Australian adolescents [3], harmful use of alcohol is still the most common and concerning substance use issue among this cohort. The most recent Australian Secondary Students Alcohol and Drug Survey identified that just over half of all 12–17 year-olds had consumed alcohol in the past year, and 50.7 % of this group were drinking at harmful levels by age 17 [4]. Equally concerning, in 2013, surveys showed that around one in six people aged between 12 to 18 had consumed 11 or more standard drinks on a single drinking occasion in the past 12 months [5]. Untreated mental health issues and early onset substance use often co-occur and can lead to a range of short-term harms. Ultimately, substance use can adversely impact relationships, educational and developmental milestones, as well as later mental and physical health [6].

Seeking help early is widely recognised as a generic protective factor, and promoting early and prompt treatment is critical in order to reduce the adverse impacts of mental health and substance use problems [6]. However, fewer than one in four 16–24 year-olds with a current disorder access health services, with those experiencing a substance use disorder being the least likely to seek professional help [2]. Rather than seeking professional help, research indicates that young people are keeping their problems to themselves or turning to their peers or key adults in their lives for help [7]. This is despite evidence that many parents and peers have poor mental health literacy [8], as indicated by their limited ability to recognise specific disorders, poor knowledge of how to seek mental health information, and poor knowledge of risk factors and causes, self-treatments and professional help available, as well as attitudes that do not promote recognition and appropriate help-seeking [9]. In addition, adolescents have knowledge, attitudes and beliefs about help-seeking and substance use that act as barriers to seeking professional help, and these are likely to have been established before the age of 13 [10]. Barriers identified include stigma, fears about lack of confidentiality, limited trust, lack of problem recognition, reliance on oneself, and concerns about helper characteristics. These help-seeking beliefs and preferences highlight the importance of building the mental health literacy of adolescents, including ensuring that they know when and how to assist their peers to access support.

Schools are an ideal and opportunistic setting in which to reach out to young people [11], particularly in terms of facilitating future help-seeking for mental health and substance use issues. Given the low help-seeking intentions of adolescents, teachers and peers are ideally placed to play a gatekeeping role (i.e. identifying issues and intervening), by supporting and helping young people to access appropriate professional support. To be effective, gatekeepers require the skills to identify mental health issues, engage the young person and help them overcome the barriers to accessing and engaging with professional help. Additionally, there is evidence from the mental health first-aid literature that teaching people how to help their peers seek help not only improves gatekeeping skills but is an innovative approach to improving their own mental health and help-seeking attitudes [12].

Although some school drug and mental health education programmes have been produced that focus on youth participation and peers as educators [13, 14], to our knowledge these have generally not focussed on exploring the barriers to helping a friend, taught students the skills necessary to overcome these barriers, nor facilitated professional help-seeking. In short, they do not focus on teaching practical steps for peers to become effective gatekeepers for their friends, which in turn, would signify that they have become proficient in terms of their own help-seeking skills. To this end, we previously reported findings from a pilot help-seeking intervention study (MAKINGtheLINK), with a focus on cannabis use and mental health, delivered over two 48-minute periods to 10 year-10 classes (182 students) at a Melbourne high school, as part of their standard curriculum [15]. The delivery of the programme was found to be both acceptable and feasible within a school setting, with students reporting increased confidence and awareness of how to seek help for themselves or a friend. A second pilot, which focussed primarily on help-seeking for alcohol and mental health, was conducted over three 50-minute periods, to 16 year-8 and year-9 classes (370 students) from three Melbourne high schools, as part of their standard curriculum [16]. Students reported that the programme led to increased knowledge about alcohol, awareness of help-seeking options and confidence to seek help for an alcohol problem. While these findings are encouraging, the pilot studies included few schools, without a control arm, and did not measure any change in actual help-seeking. Before the programme can be finalised and embedded within a national school framework, the efficacy of the programme needs to be established utilising a rigorous methodological and longitudinal design.

This cluster randomised trial seeks to demonstrate the efficacy of a universal, school-based intervention that focusses on reducing barriers and improving help-seeking and peer support for students who are experiencing poor mental health and/or misusing alcohol or other drugs.

The primary hypothesis to be tested is that:

We will also test the following secondary hypotheses:

The study (see Fig. 1) is a randomised controlled trial (RCT) with schools as clusters and individual students as participants, and will follow the Standard Protocol Items: Recommendations for Interventional Trials (SPIRIT) Statement for reporting trial protocols (see Additional file 1 for checklist) [17]. All consenting participants will be assessed at four time points: baseline, 6 weeks post baseline, 6 months post baseline and 12 months post baseline. Participants from the control schools will be assessed on equivalent dates at the four time points.

Twenty-two Victorian secondary schools will be recruited for the study. Schools will be randomly allocated to either receive the intervention immediately or to form a waitlist control group and receive the intervention after completion of the fourth survey. Randomisation will be stratified by the school’s Index of Community Socio-Educational Advantage (ICSEA) score, with two strata defined as <1000 (‘disadvantaged’) and 1000+ (‘advantaged’). The random allocation list will be generated using Stata statistical software, Release 12 (StataCorp, College Station, TX, USA: 2011) with random block sizes of 2 or 4 within each strata.

Eligible participants are year-9 students (aged 14–15 years) who consent to participate in the study. Eligible schools are state, Catholic or independent schools in Victoria with between 40 and 250 year-9 students. Schools will be sent an email explaining the study and inviting participation. To be included in the sample, schools will have to be willing to support their 2013, 2014 or 2015 year-9 cohorts to participate in the health education programme offered as part of this research, as they will not know in which year they will receive the programme until randomisation and consent form collection has occurred.

Sample size calculations accounting for cluster randomisation have been estimated using Stata 12 software (StataCorp, College Station, TX, USA: 2011). Useful background information is available from a controlled pre-post pilot study conducted in three regional high schools that were matched on socioeconomic characteristics [18]. The trial group included 171 year-11 students from mainstream classes who received a school-based intervention (Building Bridges to General Practice) that addressed adolescents’ psychological help-seeking barriers in presentations delivered by trained general practitioners (GPs) [18]. At baseline, only 1 (0.6 %) of the 171 students reported visiting a GP. At 12-week follow-up, 12 students (7.0 %) indicated receiving a consultation. Assuming the pre-test consultation rate represents the control group post measure for the proposed trial, we estimated the effect size in schools (where the intervention is implemented by trained staff as the expected difference in proportions between the control arm and the intervention arm) to be 6.4 % (pooled SD 18.8 %). Assuming 80 % successful consent and randomisation, followed by a 25 % attrition rate between baseline and follow-up, the number of students in the analysis would be 60 % of a school’s year-9 level. We anticipate that this will lead to an approximate average cluster size for analysis of 51 students per school.

It is important to note that the assumed effect size may be a conservative estimate considering the intervention’s focus on help-seeking from both formal and informal sources, rather than just GPs, who young people are reluctant to access. In other words, both the prevalence at baseline and the difference in proportions at 12-month follow-up, of students seeking help from formal and informal sources are likely to be greater than the pilot study indicates for seeking help from GPs.

For the proposed project, the intra-class correlation (ICC) is unknown, although a comparable school-based cluster randomised trial conducted by Jorm and colleagues reported an ICC of 0.05 [19]. Setting ICC (ρ) at 0.05, power (1 − β) at 80 %, and a two-sided level of significance at 5 % (α = 0.05), the required number of individuals per treatment arm is 476 students corresponding to a need for 10 schools per arm. The desired sample size was 1020 and this was increased to 1360 to allow for 25 % dropout by the 12-month follow-up assessment.

The help-seeking intervention draws upon two models of behaviour change – the Information-Motivation-Behavioural Skills Model (IMB) and the Theory of Planned Behaviour (TpB). The IMB model is a well-validated, comprehensive health behaviour change framework that has been used in schools, particularly for HIV education and prevention [20]. The TpB is a health behaviour change framework that has been used extensively to guide experimental health intervention trials [21]. In this study, the help-seeking activities to be trialled provide students with information about how to seek help and from whom (Information), investigate participants’ psychological barriers to help-seeking (Beliefs, Intentions), investigate risky behaviours associated with alcohol use and mental health problems (Symptom levels), and provide opportunities and videos for skill rehearsal (Behavioural skills) which, according to our composite model, will lead to increased intention to seek help and actual help-seeking (Behavioural outcome).

The intervention will consist of five interactive classroom activities run over two school periods (average period is 75 minutes), plus a booster session 1 month later (to reiterate key messages and help students gain practical experience by applying the help-seeking skills they have learnt), delivered by an experienced external facilitator with the assistance of the regular classroom teacher. The rationale for the addition of a booster session is based on feedback and data from our pilot investigations, as well as research indicating the importance of booster sessions in terms of enhancing and maintaining treatment effects [22]. Activities will cover (1) recognising when a friend needs help (vignettes about poor mental health and risky drinking), (2) what types of helpers are available, (3) myths and facts about substance use and mental health, (4) identifying and overcoming barriers to professional help-seeking, (5) assisting a friend to access help, and (6) accessing reliable sources of help (see Table 1).

Recruitment for the trial began in August 2013. Emails providing information about the programme opportunity and inviting expressions of interest from schools were sent to all Catholic, independent, and state schools within 50 km of Melbourne’s CBD with a year-9 cohort of between 40 and 250 students. Schools were required to obtain consent forms signed by the students’ parent/guardian for them to participate in the study, returning a minimum of 60 % positive consents of year-9 students and guardians to be eligible to participate in the trial. In order to maximise return of consent forms, each school principal assigned a staff member to the MAKINGtheLINK project in order to take responsibility for their collection. In addition, the research team ran two presentations for year-9 students to explain why participation in the survey was important, provided clear instructions about where and when to return consent forms, and provided extra copies of forms where required. After these initial steps were taken, contacts at each school emailed or telephoned parents to follow up with them about the consent forms.

In spite of these efforts, we were unable to obtain the required sample size at four of the seven schools that had agreed to participate by December 2013. In order to increase participation rates, and supported by the Department of Education and Early Childhood Development, a passive/opt-out consent process was adopted from 2014 onwards where parents were no longer required to opt in to the research, and students could make an informed choice as to whether they wanted to participate. Under this process, students are assured that participation is optional and that they will not be penalised for choosing not to participate, and are required to check the ‘consent’ box before completing the online survey. Parents are informed of the research at least 1 month in advance and can opt out of the research trial by contacting the school if they do not want their children involved.

All consenting students subsequently undergo baseline testing during class prior to intervention delivery. Intervention schools then receive the programme in their regular class time. Participants will be re-tested the week following the intervention or at an equivalent time for the control schools (6 weeks post baseline). Follow-up assessments will also be completed online during class at 6 and 12 months post baseline. All students will be provided with information on how to seek help for substance use, depression and other problems from a range of resources and services at the completion of the surveys. Figure 2 shows the SPIRIT diagram for the trial procedure.

All information in the survey will be de-identified and confidential, and the records will be kept securely for 5 years at the research centre (Turning Point) after which time they will be destroyed. However, if students report high scores on a screen for depression the researchers will notify the student welfare coordinator (SWC) who will contact the child to check that they are safe and refer for assistance as appropriate. Any student who becomes upset while completing the surveys can stop at any time. The researcher will be available to provide debriefing, and the student will be encouraged to talk to their parents or one of their teachers. They will also be provided with contact details for services, if necessary.

In addition to demographic information (age, gender, postcode, language spoken at home, living arrangements, parental occupation, country of birth), the following measures will be administered.

Schools will be unaware of each school’s allocation until after they have provided consent. Participating individuals and their parents will be unaware of whether they have been allocated to the intervention or control group until after the first data collection point.

The research staff will co-ordinate all appropriate data management and cleaning prior to analysis. Data on screening, refusals and dropout will be coded and reported as per Consolidated Standards of Reporting Trials (CONSORT) guidelines for participant flow through the trial (see Fig. 3). A description of the baseline characteristics of schools and of individuals who participate in the two intervention arms will be compiled using descriptive statistics such as mean and standard deviation, median and interquartile range, and percentages.

In order to evaluate the main aim of the study, which is to examine the association between participation in the help-seeking programme (control versus intervention arm) and actual help-seeking behaviour at 12 months post intervention, logistic regression analyses will be undertaken testing for group differences between help-seeking behaviour as measured by the AHSQ. In a secondary analysis, gender, depression and anxiety scores, reported alcohol problems, and any characteristics of schools or participants that are imbalanced to a large extent at baseline, will be included in an extended logistic regression model for help-seeking behaviour in addition to the control/intervention arm indicator variable.

For secondary aims, the outcomes are measured by continuous variables, and as such linear regression analysis will be used to compare the two study arms with the primary analysis comprising models with treatment indicator and the baseline value of the outcome score, i.e. analysis of covariance. Secondary analyses will adjust for imbalanced characteristics in the same way as for the primary analysis. All analyses will utilise robust standard errors to adjust for clustering by school.

Each analysis described above will be conducted using participants with all necessary data for that analysis, according to the intention-to-treat principle. Characteristics of participants who were lost to follow-up and who completed follow-up will be compared. Analyses will be repeated using all participants on the basis of multiple imputation involving responses at all four time points to impute for missing values.