McCormick 1995.

Study characteristics
Methods	Trial name: The North Carolina School Health and Tobacco Education Project (SHTEP)/Skills Management and Resistance Training (SMART) Study design: cluster‐RCT Intervention duration: 4 years Length of follow‐up from baseline: 4 years Differences in baseline characteristics: not reported Unit of allocation: school district Unit of analysis: school district and individual teachers
Participants	School type: junior high or middle schools Region: North Carolina, USA Demographic/socioeconomic characteristics: not reported Inclusion/exclusion criteria: to be eligible for participation, a school district had to have ≥ 2 junior high or middle schools (51/140 school districts in North Carolina met this criterion). All teachers who were eligible to teach health. Number of schools allocated ‐ 21 districts (11 intervention, 10 control) ‐ Initial implementation = 69 teachers (51 intervention, 18 control) ‐ Maintained implementation = 136 teachers (86 intervention, 50 control) ‐ 42 schools Numbers by trial group n (controls baseline) = 10 n (controls follow‐up) = 10 n (interventions baseline) = 11 n (interventions follow‐up) = 11 Recruitment School districts:28 districts were randomly selected and a 5‐stage strategy was used to recruit districts to the study. After a school district agreed to participate, it was randomly assigned to either the experimental or control condition, resulting in 11 experimental and 10 control districts. Schools:in both experimental and control school districts, ≥ 2 schools and ≥ 2 classrooms per school participated in the study Teachers: all teachers in study districts identified by their schools as "eligible to teach health" were included in the sample. Students: not reported Recruitment rate School districts:21 school districts of 28 identified, therefore, 75% Schools:approximately 50 schools Teachers at baseline 69/115 = 60% Teachers at follow‐up 136/175 = 78% Classrooms: approximately 100 classrooms Students:approximately 3000 students exposed to 1 of the tobacco prevention curricula
Interventions	Number of experimental conditions: 2 (1 intervention, 1 control) Policies, practices or programmes targeted by the intervention School health/tobacco prevention curricula (schools districts offered choice of 3: Growing Healthy, Teenage Health Teaching Modules, Project SMART). Implementation strategies EPOC: educational meetings ‐ Implementation intervention: in‐depth training for teachers and administrators on the use of the specific curriculum that had been adopted. EPOC: educational materials ‐ Schools were provided with curricular materials (schools chose the most suited of 3 options). EPOC: local consensus processes ‐ Signing of a Memorandum of Understanding: a Memorandum of Understanding, specifying which curriculum was adopted, how many lessons were to be taught, and how many classes would receive instruction, was signed by each district. Adoption intervention: process consultation for adoption. A consultation workshop was conducted with each experimental district to inform school personnel about the 3 health curricula that were being disseminated. Theoretical underpinning: diffusion of innovation Description of control: districts in the control conditions were mailed curricula materials and provided TA upon request, but did not receive the training sessions (EPOC: educational material).
Outcomes	Outcome relating to the implementation of school policies, practices or programmes ‐ % later implementation of curriculum for school district ‐ Extent later implementation mean for school district (% of total curriculum activities taught) Data collection method: each teacher who was eligible to teach health received an implementation check sheet and was asked to indicate which activities were taught. Data were also aggregated at school district level. Assessed at year 4 follow‐up "later implementation". A dichotomous measure of implementation (i.e. yes/no) necessary but not sufficient to assess the implementation of a school health curriculum. Therefore, Implementation check sheets were also used to assess implementation as the percentage of total curriculum activities that were taught (extent of implementation). Validity of measures used: not reported/self‐report methods Data collection method: 13‐item measure of awareness and concern among teachers and administrators about tobacco use among students Validity of measures used: validity was not reported, although the authors reported the instrument was described elsewhere. Outcome relating to cost: not reported Outcome relating to adverse consequences: not reported Outcome relating to child diet, PA or weight status: not reported
Notes	Research funding: National Cancer Institute (#5 R01 CA 459907‐02). Conflicts of interest: not reported
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Random sequence generation procedure not described.
Allocation concealment (selection bias)	Unclear risk	No information provided about allocation concealment and, therefore, it was unclear if allocation was concealed.
Blinding of participants and personnel (performance bias) Implementation outcome	High risk	Outcome group: all Given the nature of the intervention, participants and study personnel were likely to have been aware of study allocation and, therefore, high risk of performance bias.
Blinding of outcome assessment (detection bias) Implementation outcome	High risk	Outcome group: all. All outcomes subjective, self‐report. There was no mention of blinding of participants and personnel. Teachers completed check sheets that assessed their delivery of curricula activities and researchers conducted site visits at implementing schools. High risk of performance bias as the implementation outcomes likely to be influenced by knowledge of group allocation.
Incomplete outcome data (attrition bias) Implementation outcome	High risk	Outcome group: organisational climate. The implementation measures represent cross‐sectional assessments. There was low attrition as only 1 experimental and 1 control district dropped out of the study (low risk of bias). Outcome group: organisational size Organisational size represented the total student enrolment for each school district and was obtained from North Carolina Department of Public Instruction records. The smallest district had 4134 students and the largest district had 29,532 students. Intervention = 9 districts (82%) Control = 7 districts (70%) Imbalance across intervention and control district response (high risk of bias) Outcome group: time of adoption Time of adoption range 9.57–19.86 weeks Intervention = 8 districts (73%) Control = 7 districts (70%) (high risk of bias) Outcome group: Awareness – concern – interest The response rate for teacher's awareness and interest was 69% (n = 432), and for concern was 52% (n = 324). Insufficient reporting of attrition/exclusions to permit judgement of 'Low risk' or 'High risk' (e.g. no reasons for missing data provided) (unclear risk of bias). Outcome group: Level of Use – District District level grouping of Level of Use instruments sent to teachers (252/570 returned, response rate 44%). Mean percent of implementation for each district reported. Intervention = 9 districts (82%) Control = 7 districts (70%) Imbalance across intervention and control district response (high risk of bias). Outcome group: Level of Use – Teacher Of the 570 Level of Use instruments sent to teachers, 252 were returned and completed, for a response rate of 44%. Only the Level of Use instruments for School Health and Tobacco Education Project curricula were analysed (n = 71). Intervention = 52 teachers Control = 20 teachers Insufficient reporting of attrition/exclusions to permit judgement of 'Low risk' or 'High risk' (e.g. no reasons for missing data provided) (unclear risk of bias). Outcome group: Implementation Checklist – District Initial implementation Intervention = 8 districts (73%) Control = 6 districts (60%) Later implementation Intervention = 8 districts (73%) Control = 6 districts (60%) Imbalance across intervention and control district response (high risk of bias). Outcome group: Implementation Checklist – Teachers During the third year, 115 Implementation Checksheets were sent to teachers and 69 were returned (60%). Later implementation was measured in year 4 of the project. During this year, 136/175 (78%) Implementation Checksheets were returned. Insufficient reporting of attrition/exclusions to permit judgement of 'Low risk' or 'High risk' (no reasons for missing data provided) (unclear risk of bias).
Selective reporting (reporting bias)	Unclear risk	No study protocol, therefore, it was unclear if there was selective outcome reporting.
Other bias	Low risk	Did not appear to be at risk of contamination as districts were the unit of randomisation.
Recruitment to cluster	Unclear risk	District selected which 2 middle or junior high schools participated.
Baseline imbalance	Low risk	Districts randomised to condition.
Loss of cluster	Low risk	Small loss of clusters (1 experimental, 1 control) and equal dropout across conditions.
Incorrect analysis	High risk	No adjustment for clustering reported in analysis.
Compatibility with individually randomised RCTs	Unclear risk	Unable to determine if a herd effect existed.
Overall risk of bias assessment	Unclear risk	Most domains were at unclear risk of bias.