Wagner 2000a.

Methods	Study design: cluster‐randomized controlled trial (independent samples) Sampling frame: non‐institutionalized adults age 18+ with working telephones residing in intervention or control communities Sampling method: Waksberg method random‐digit dialing Collection method: telephone interview Description of the community coalition: The 11 grantees of the Community Health Promotion Grant Program were expected to establish coalitions that encompassed a broad spectrum of community agencies and organizations. Sponsoring agency for the coalition at this site (“Community G”) was a university; no further site‐specific information on coalition composition or structure is reported
Participants	Communities: urban communities in Western USA (no further location reported) Country: USA Ages included in assessment: 18+ Reasons provided for selection of intervention community: not reported Intervention community (population size): “Community G”: urban community with large Hispanic population in Western USA (80,953) Comparison community (population size): 4 urban communities (population size not reported)
Interventions	Name of intervention: Henry J. Kaiser Family Foundation’s Community Health Promotion Grants Program Theory: not reported Aim: to address nutrition problems and increased risk of cancer and cardiovascular disease in the local Hispanic community Description of costs and resources: $150,000 per year from Kaiser Family Foundation plus any external funding or in‐kind donations the program could obtain (total funding not reported by community; overall average = $237,000 per year per site total) Components of the intervention: community health screenings, school‐based nutrition education, grocery store interventions, community nutrition classes  Start date: 1987 Duration: 5 years
Outcomes	Outcomes and measures: nutrition (self reported): (1) percent calories from fat; (2) days eating red meat as main meal; (3) percent who drink low‐fat milk; and (4) fruit and vegetable intake Time points: baseline (1988) and follow‐up (1992)
Notes	Because the 11 grantee communities had different populations, interventions, and evaluation study designs, and because the evaluation was stratified by site, the 2 communities with a minority target group and complete evaluation results are presented as separate studies (see Wagner 2000b)  Source of funding: Kaiser Foundation
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Applicants for funding were stratified by urbanicity, then were randomly divided into intervention and control groups by the “finite selection model” to ensure the desired mix of health problems, minority target populations, and geographic locations
Allocation concealment (selection bias)	Unclear risk	Not reported
Baseline outcome measurement similar	Low risk	Similar rates of CVD/cancer at baseline; similar with respect to diet outcomes, except intervention community more likely to drink low‐fat milk
Baseline characteristics similar	High risk	3 of 4 control communities had significantly different racial composition; 2 of 4 control communities had significantly lower poverty levels
Blinding of outcome assessment (detection bias) All outcomes	Unclear risk	Cross‐sectional surveys unlikely to be the subject of detection bias
Incomplete outcome data (attrition bias) All outcomes	Unclear risk	Independent cross‐sectional samples imply no attrition; overall response rate of 50% did not differ between intervention and control
Blinding of participants and personnel (performance bias) All outcomes	High risk	No blinding
Protection against contamination	Unclear risk	Intervention and control site proximity not reported
Selective reporting (reporting bias)	Unclear risk	Presented outcomes were selected on the basis of hypothesized intervention effects