Twenty-four studies (n=9,468) were included: 9 experimental studies (n=4134), 6 quasi-experimental studies (n=2,198), 7 descriptive studies (n=2,890), 1 cross-sectional pre-test post-test study (n=217) and 1 qualitative study (n=29).
Access to heath services (16 studies).
The studies showed that CHWs improved access to health services for prenatal care (1 study), mammography screening (5 studies), Pap testing (4 studies), sick-child visits (1 study), pre- and postnatal care (2 studies), sexually transmitted disease testing (2 studies), smoking cessation programmes (1 study) and maternal-child health visits (1 study). The overall attrition rates ranged from 19 to 60%.
Knowledge (7 studies).
Five of the 7 studies showed improvement in knowledge for general health practices among Latino farm-working women (1 study), cervical cancer among Native American Indian women (2 studies), diabetes-related self-care among African American women (1 study), and acquired immune deficiency syndrome among homeless African American and Hispanic/Latino women (1 study).
The two studies that showed no increased knowledge had methodological problems, such as high attrition rates, small sample sizes and a lack of standardised instruments.
Behaviour (6 studies).
Five of the 6 studies showed positive outcomes with CHWs: weight loss in church-based African American women (1 study); longer duration of breast-feeding in African American women (2 studies); reductions in noninjection drug use and the number of sexual partners and increased condom use in homeless ethnic minority women (1 study); and increased physical activity in African American women with type 2 diabetes (1 study).
One study showed no change in the behaviour of Hispanic women receiving diabetic education and case management.