Spencer 2011.

Methods	Study design: randomized controlled trial Sampling frame: medical facility records of physician‐diagnosed type 2 diabetes, self identified as African American or Latino, living in target zip codes Sampling method: medical record review Collection method: laboratory data, survey Description of the community coalition: REACH Detroit Steering Committee. Composed of community health leaders, clinical providers, researchers, and REACH Detroit staff
Participants	Communities: southwest Detroit, eastside Detroit Country: USA Ages included in assessment: > 18 years old (n = 164) Reasons provided for selection of intervention community: communities with predominantly Latino or African American populations with low average income who are historically underserved Intervention community (population size): not reported Comparison community (population size): not reported
Interventions	Name of intervention: REACH (Racial and Ethnic Approaches to Community Health) Detroit Initiative Theory: Empowerment Theory Aim: to determine whether a culturally tailored community health worker (CHW) intervention for diabetes self management improves HbA1c levels, blood pressure, lipid levels, diabetes knowledge, diabetes self management behavior, and diabetes‐related distress amongst low‐income, inner city African Americans and Latinos with diabetes Description of costs and resources: cost information not provided Components of the intervention: trained community health workers (“family health" advocates) conducted diabetes education classes and tailored home visits and a clinic visit with patient and care provider; both intervention and control groups had access to REACH community activities (healthy eating, physical fitness, farmers' markets) and received health care at facilities with trained REACH healthcare providers Start date: September 2004 Duration: 22 months
Outcomes	Outcomes and measures: physiological measures (HbA1c, LDL, BP), self management knowledge, diabetes self efficacy, physical activity, dietary practices Time points: baseline and 6 months (f/u intervention: n = 58; control: n = 69)
Notes	Funder: US Centers for Disease Control and Prevention Funding source: government
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Random sequence generation not reported
Allocation concealment (selection bias)	High risk	Allocation concealment not described
Baseline outcome measurement similar	Low risk	Baseline HbA1c similar
Baseline characteristics similar	Low risk	Intervention group was significantly younger than control; this was controlled for in the data analysis
Blinding of outcome assessment (detection bias) All outcomes	High risk	Not blinded
Incomplete outcome data (attrition bias) All outcomes	High risk	All measured outcomes appear to be reported. Laboratory data were not available for a significant percentage of those initially enrolled. However, ITT analysis was performed
Blinding of participants and personnel (performance bias) All outcomes	High risk	No blinding reported
Protection against contamination	High risk	Not stated. Because interventions occurred in the same communities, likelihood of contamination is high
Selective reporting (reporting bias)	Low risk	All measured outcomes appear to be reported
Other bias	Unclear risk	Modest sample size may limit power; laboratory test results obtained through medical chart reviews