Melnyk 2013.

Methods	Study name: COPE (Creating Opportunities for Personal Empowerment) Study design: cluster RCT at the level of the school Intervention arm: COPE Comparator arm: Healthy Teens attention control programme Sample size calculation performed: yes Subgroups prespecified: N/A Subgroup analyses: none Start date: January 2010 Duration of follow‐up: 6 months Number of follow‐ups: 2 Follow‐up time points: post‐test and 6 months following end of intervention Intracluster correlation coefficient: not reported
Participants	Number of schools randomised: 11 Number of participants randomised: 807 total, 374 intervention, 433 control Age (range or mean (SD)) or grade at the start: mean age 14.75 years intervention; 14.74 years control Gender: 54.5% female in intervention, 49.20% female in control Ethnicity: American Native: 2.8% intervention, 4% control; Asian: 2% intervention, 5.7% control; black: 8.4% intervention, 11.2% control; white: 8.7% intervention, 18.8% control; Hispanic: 76.8% intervention, 59.6% control; other: 1.4% intervention, 0.7% control Socioeconomic status: N/S Inclusion criteria: adolescents aged 14 to 16 years who were enrolled in a health education course. Teens of any gender, ethnicity, or SES and those who could read and speak English Exclusion criteria: medical condition that would prevent participation in the physical exercise component
Interventions	Timing of randomisation: before baseline survey Duration of the intervention: 15 weeks Description of the intervention arm(s): a 15‐session educational and cognitive‐behavioural skills building programme that incorporated 15 to 20 minutes of physical activity in each session. Areas covered included healthy lifestyles, self‐esteem, goal‐setting, and problem‐solving; stress and coping; emotional regulation; effective communication; overcoming barriers to goal progression; food and nutrition information (e.g. portion sizes, nutrients, food groups, snacks); and physical activity. Homework activities were conducted to reinforce the content of the programme, and 4 parent newsletters were sent home for review. Brief description of the theoretical model: cognitive‐behavioural theory Description of the comparator arm(s): attention control "Healthy Teens", which provided the same number and length of sessions
Outcomes	Primary outcomes: healthy lifestyle behaviours, BMI Secondary outcomes: mental health, alcohol and drug use, social skills, academic performance
Setting	Country: USA; Place: South West Setting: secondary school Focus: school ‐ universal
Process measures	Process data reported: yes Method: quantitative Description: 78% of COPE teens reported on the post‐evaluation questionnaire that the programme was helpful. The most helpful programme elements were reported to be content on stress and coping, nutrition, and exercise. 92% of parents indicated it was helpful for their teens; 94% said they would recommend it to family or friends. 82% of parents agreed that information shared through newsletters was useful. Study authors also stated that there was less than adequate intervention fidelity by some of the teachers in the study.
Statistics	Sample size: 807 participants Unit of randomisation: school Unit of analyses: individual accounting for clustering Methods to promote equivalence between groups: N/S Statistical models: repeated logistical regression models using generalised estimating equations were used to analyse binary outcomes. Baseline differences adjustment: yes Repeated measures methods in analysis: yes
Notes	Equity: N/A Funding: NIH/National Institute of Nursing Research (1R01NR012171) Randomisation method: unclear Clustering accounted for in sample size calculation: yes Cluster randomisation methods to account for clustering in analysis: generalised estimating equations
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Lack of information
Allocation concealment (selection bias)	Unclear risk	Lack of information
Blinding of participants and personnel (performance bias) All outcomes	Unclear risk	Study authors stated that this was a blinded cluster RCT that compared the intervention vs an attention control.
Blinding of outcome assessment (detection bias) All outcomes	High risk	Self‐report measures
Incomplete outcome data (attrition bias) All outcomes	Unclear risk	Insufficient information
Selective reporting (reporting bias)	High risk	Reporting of certain outcomes not consistent with protocol
Other bias	Low risk	No other sources of bias