Our review revealed a number of methodological issues regarding the seven included studies. First, interventions were quite clearly described, and most details were reported except for training, supervision (content and duration) and CHW characteristics (e.g., age, gender, educational level, and payment). Therefore, we could not determine whether the type of training, supervision and type of CHW influenced the effectiveness of CHWs. A second issue was that several studies did not mention which components of the CHW interventions were effective and produced the reported effects. Third, we identified only a small number of studies. Therefore, robust conclusions on effectiveness could not be drawn. In addition, all studies have been
conducted in the US which makes generalisation ofthe results of these studies to other Western countries questionable. Finally, none of the included studies reported outcomes regarding knowledge. We would expect that an improvement in behaviour through CHW counselling and education could also have improved knowledge. This hypothesis was confirmed
in a previous review of studies of ethnic minority women of all ages in the US, in which an
increase in knowledge was shown as a result of CHW education