Plescia 2008.

Methods	Study design: controlled before‐after study Sampling frame: residents of 14 neighborhoods in Charlotte, NC Sampling method: random samples of intervention and control populations Collection method: behavioral risk factor survey Description of the community coalition: The Charlotte REACH coalition grew out of a national funding program (REACH 2010) to address health disparities in minority communities.  The Charlotte REACH coalition included community members, community grassroots organizations, community health center, neighborhood association, county health department, a community‐based substance abuse program, and other community service providers. Health disparities were identified, and cardiovascular disease and diabetes were identified as priorities. Decisions were made by consensus, with an external consultant providing mediation
Participants	Communities: 14 neighborhoods in Charlotte, NC Country: USA Ages included in assessment: adults (n = 3738) Reasons provided for selection of intervention community: predominantly urban, low‐income African American communities with high risk for cardiovascular disease and diabetes Intervention community (population size): 19,670 Comparison community (population size): not reported
Interventions	Name of intervention: Charlotte REACH Theory: Social‐Ecologic Model Aim: to reduce behavioral risk factors for cardiovascular disease and diabetes Description of costs and resources: not described Components of the intervention: (1) increasing community resources to remove barriers to healthy behavior; (2) improving quality of care at health center; (3) initiating campaign to change social norms; and (4) engaging in political advocacy for evidence‐based policy interventions Start date: 2001 Duration: 60 months
Outcomes	Outcomes and measures Reduction in physical inactivity Increase in consumption of ≥ 5 vegetables/fruits daily Decrease in current smoking — prevalence and 95% CIs from survey results   Time points: 2001 (baseline) and 2005 (follow‐up)
Notes	Funding source: government
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	High risk	No randomization
Allocation concealment (selection bias)	High risk	No allocation concealment
Baseline outcome measurement similar	High risk	Intervention group less likely to be physically active and more likely to be smokers at baseline
Baseline characteristics similar	Unclear risk	Women and older age groups better represented in intervention group
Blinding of outcome assessment (detection bias) All outcomes	Low risk	Population‐based survey
Incomplete outcome data (attrition bias) All outcomes	Low risk	Independent samples imply no attrition; response rates stable over time (63% to 69%)
Blinding of participants and personnel (performance bias) All outcomes	High risk	No blinding
Protection against contamination	Unclear risk	Contamination not addressed; intervention and control within same state
Selective reporting (reporting bias)	Unclear risk	Unclear how 3 outcomes were chosen from 60 survey questions