In a randomized parallel group multi-site trial looking at achievement of standard recommended treatment goals at baseline and one year later by 2135 patients with diabetes, hypertension, and congestive heart failure, we tested the hypothesis that adding a lay person with brief training to usual care would improve clinical outcomes. These "care guides" were culturally matched to patients served, similar to community health workers, but were located in 6 diverse primary care health clinics where they could meet patients face-to-face. They were asked to assist communication in both directions between providers and patients, taking advantage of their status as non-authority figures. They were given two weeks' training about these diseases and behavior change theory. This intervention was designed to be low cost, easy to implement, and to integrate into rather than change clinic work flow.
