Table 3 Summary table of included studies

Audrey et al., 2008 [34]

UK

Tobacco

A Stop Smoking In Schools Trial (ASSIST)

cRCT.

questionnaire interviews

30 ASSIST schools & 29 control

Process evaluation to examine the context, implementation and receipt of the intervention

Framework method of data management. (reading, coding & identifying themes, & sorting material according to key issues)

Teachers welcomed external training—it interested pupils, prevented difficulties of discussing smoking with teachers and relieved staff burden. Implementation appeared compatible with the school ethos and timetable. Smoking was perceived as a difficult issue and staff welcomed a new initiative. Disruption to the timetable was inevitable, and the importance of communication between ASSIST staff and teachers was important

CASP:

Moderate

Barr et al., 2002 [33]

USA

Tobacco

Tobacco Use Prevention Education (TUPE)

Telephone survey

296 middle school teachers & 282 high school teachers

Relations between TUPE teachers’ receptivity or amenability to implement TUPE programs and features of implementation settings

Cluster analyses for amenability to implementation. A one-way ANOVA for associations between amenability and implementation. A hierarchal multiple-regression for staff effectiveness perceptions

Indicators of staff amenability were variable. The most amenable staff reported consistently covering each activity with few barriers. For staff perceptions of effectiveness to prevent smoking initiation: Tobacco related norms accounted for 9.9% of variance, staff training & TUPE support or barriers—4.2%, and class activities—4.0%. For staff perceptions of TUPE for cessation: Tobacco norms—6.6% of variance, staff training & TUPE support—6.3%, class activities—5.5%

EPHPP:

Moderate

Basen-Engquist et al., 1994 [31]

USA

Tobacco.

Minnesota Smoking Prevention (MSPP)

Questionnaire

39 districts in live training & 33 in video training. Mean number of pupils was 41, 2.8 teachers

Assessing how the type of teacher training affects implementation via a live workshop or video training

Fisher exact test & Mann-Whitney U for differences in teacher implementation

Two group t-tests tested differences between students in the live and video districts

The relationship between type of training and use of the curriculum was significant. Districts who were assigned to the video training condition were less likely to teach the curriculum. However, implementing teachers from both groups reported high levels of implementation. Students in live workshops were more likely to recall discussions and activities

EPHPP:

Weak

Garrahan 1995 [40]

USA

Substance Use

Systems approach model

800 students

Not stated

Baseline substance use data was collected via a survey & analysed

Involving school personnel in a building-wide manner and monitoring efforts and outcomes was important. All implemented intervention aspects were linked to existing components of the school, and this gave the impression that what was implemented was based on common sense or self-evident reasoning

EPHPP:

Weak

Jarrett et al., 2009 [35]

USA

Tobacco

Not-On-Tobacco (NOT)

Survey

769 pupils who reported regular smoking.

Perceptions of facilitator characteristics & the relationship between perceptions & outcomes

Descriptive analyses used to determine overall ranking of facilitator characteristics. Chi-square test to determine if facilitator ratings differed by race or sex

88.7% of pupils rated facilitators as favourable. No nagging or preaching, nonjudgmental, trustworthy, caring, & confidentiality were scored highly. There were few differences in ratings by race. Favorability scores were associated with changes in smoking (quit or reduce). Pupils who perceived facilitators favourably showed significant smoking reduction and cessation rates, regardless of sex or race

EPHPP:

Weak

MacDonald and Green 2001 [44]

Canada

Substance Misuse

Interviews and observations with Project Workers (PWs)

100 interviews in 6 sites with school admins, teachers, pupils, parents, & agency staff

Participants were probed around the level of understanding and support for prevention, implementation experiences, implementation barriers & facilitators, support for PWs and the school’s problem with drug and alcohol issues

Constant comparative method of grounded theory

Field notes were recorded and used to support analysis

PWs needed to establish legitimacy and familiarity within schools, by overcoming staff opposition. They had to address conflicting expectations, resulting from poor preparation. Schools had to be ready and willing to implement, and PWs faced issues selling the model, and facilitating participation. Training sought to teach PWs to understand the model, but this did not occur and PWs realised they did not understand it enough to implement to others and few achieved it as intended. Some tried, but were discouraged by school barriers. Some retained key features, but omitted elements due to admin pressure or context demands

CASP:

Strong

McBride et al., 2002 [43]

Australia

Alcohol

The School Health & Alcohol Harm Reduction Project

(SHAHRP)

Longitudinal study

41 classes.28 teachers 6 schools

Series of methods to optimise and assess implementation fidelity including training, critical assessment and self report

Spearman’s rank measured fidelityTheme matrices described qualitative responses

SHAHRP was taught 80.7% as intended, with fidelity ranging from 78.9 to 83.4%. Implementation was optimised by: training, staff and pupil motivation and timing. Teachers found too much work in some lessons, interruptions reduced classroom time and implementation effectiveness was pupil dependent. Expectations needed to be lowered for difficult pupils and some activities were not implemented as intended

CASP:

Weak

EPHPP:

Strong/Moderate

McCormick et al., 1995 [32]

USA

Tobacco

RCT

21 districts, 50 schools, and 3000 pupils.

Districts were assigned to control or intervention.

Use of ‘Level of use’ tool and implementation check-sheets

Population means, median, frequencies & correlations used for summary. Non-parametric tests tested for differences between control and intervention

Overall implementation completeness was low, with the mean % implemented being 70% and 23% implemented ≥90%. Larger districts were more likely to implement than small ones. Districts with favourable climates were more likely to implement and reported higher usage. Trained teachers were more likely to implement curricula and more likely to implement higher proportion

EPHPP:

Moderate

Pettigrew et al., 2013 [42]

USA

Substance use

keepin’ it Real (kiR)

Ethnography

39 schools; 14 Control, 14 Rural: Mean number of pupils per school = 99, with a range from 27 to 226

An assessment of teacher implementation using the indicators; delivery methods, consistency of delivery, teaching standards

Coding provided; quantitative implementation ratings—quality adherence, adaptation, delivery and engagement, whilst qualitative codes identified adaptation and engagement

Analysis identified teacher control as passive, coordinated, or strict, and pupil participation as disconnected, attentive, or participatory; serving as a classroom typology for kiR implementation. Passive teachers were linked with passive pupils, strict teachers had attentive pupils, whilst classes with participatory pupils were taught by coordinated teachers. Teachers who taught kiR frequently tended to display similar control and pupils participated consistently

CASP:

Moderate

Rohrbach et al., 2007 [37]

USA

Substance Use

Project Towards No Drug Abuse (TND)

RCT

18 schools—6 in each different condition. Pupils ranged from 13 to 19 years of age

Study compared teachers with Program Specialists (PSs). Questionnaire assessed implementation fidelity of TND via adherence, classroom process and perceived pupil acceptance

Inter-rater reliability was calculated for each item. To test the effect of implementer on fidelity and outcomes, a mixed-linear model was used

Of the 4 indexes of fidelity, only delivery quality differed between PSs and teachers. Both teachers and PSs achieved effects on 3 of the 5 immediate outcome measures, including program knowledge, addiction concern, and self-control. Pupils’ posttest ratings of the program and the quality of delivery showed no difference between teacher and specialist-led classrooms

EPHPP:

Moderate

Skara et al., 2005 [36]

USA

Substance Use

Project Towards No Drug Abuse (TND).

Questionnaire

18 schools—6 in each different condition. 2735 students completed pretest questionnaires: 85% completed post-program

Questionnaire assessed implementation fidelity of TND via questions open and closed questions

Data was analysed using a generalised mixed-linear model using SAS

The curriculum was implemented as intended, received favourable ratings, and significantly improved knowledge. Providers reported high adherence to lesson plans and lessons were not difficult to teach. Adherence and delivery quality did not differ by curriculum or school. Individual ratings of delivery quality were favourable, including providers’ perceptions of pupil participation, pupil interest, provider’s maintenance of class control & providers’ perceptions of effectiveness

EPHPP:

Moderate

Sloboda et al., 2009 [39]

USA

Substance Use

Take Charge of Your Life (TCYL)

Observation and surveys

TCYL was delivered by 140 Drug Abuse Resistance Education (DARE) officer instructors

Implementation fidelity measured using instructional strategy (IS)

Descriptive statistics & analyses between content coverage and IS & scores from targeted lessons were conducted using hierarchical linear modelling to gain 2-level random intercept models

Higher content was correlated with IS. There were no correlation between age, sex, race, education, content coverage or use of IS. Pupils with higher coverage scored higher on the consequences measure. Results indicated pupils with a higher proportion of the content had greater perceptions of negative consequences. Greater exposure and greater content coverage was related to negative alcohol expectancies

EPHPP:

Weak

Stead et al., 2007 [38]

UK

Substance Use

Blueprint

Observations and interviews

30 schools in 4 Local Authority areas: 24 intervention & 6 control.

Year 7 (11–12 years) & Year 8 (12–13 years)

Implementation fidelity measured via adherence, exposure, participant responsiveness, quality of delivery and program differentiation

Observation schedule used to generate descriptive statistics.

The mean content fidelity was 72%. As teachers got familiar with lessons, they were likely to modify or omit elements. Fidelity was highest in teacher-pupil lessons & lowest for pupil-pupil. Resource use was variable and teachers found timing and completing content difficult. Teachers were unsure of interactive sessions due to disruption & unpredictable outcomes. Some teachers expressed concern about answering questions about drugs, but there was no difference in delivery quality of teachers with experience & those without

CASP:

Moderate

Sussman et al., 1993 [30]

USA

Tobacco

Project Towards No Tobacco Use

Questionnaire

4852 7th grade pupils. 9 Health Educators. 76 observers collected teacher data

Key implementation measures were around program completion and delivery (fidelity- adherence, exposure, reinvention)

Pupils & educators gave ratings of implementation. Post hoc comparisons were used between pairs of means and one-way ANOVAs predicted response means

Adherence did not vary by condition and high levels of implementation were observed in all conditions. Pupils preferred physical consequences and enthusiasm was rated the lowest. Health educators’ enthusiasm, effort and class enthusiasm differed by condition. Teachers did not differ in their ratings of class control or understandability

EPHPP:

Weak

Thaker et al. 2008 [41]

USA

Substance Use

Reconnecting Youth (RY) program

Organisational diffusion study

At risk of drop out students from grades 9–12. 5 schools from each district took part

Three diffusion of innovation indicators used: perceived advantage, complexity and compatibility. Capacity, school turbulence and leadership/admin support were also explored to assess how they could affect implementation

Survey data was analysed using SPSS whereas interview data was transcribed and analysed using qualitative content analysis

Teachers reported learning RY difficult, as they were not prepared & needed to plan. RY was rigid, complex and difficult to implement the timelines & content. School capacity (skills and resources) varied & affected implementation. Other issues were budget shortfalls, funding cuts, difficulties finding rooms and school turbulence (transient pupil populations, school reorganisation, schedule changes, & staff turnover). RY lacked leadership and admin support. Only 50% of staff reported principles being supportive. Whilst only 1/3 of district admins considered RY important

EPHPP:

Weak