Objectives:

To conduct a systematic review of the evidence on characteristics of community health workers (CHWs) and CHW interventions, outcomes of such interventions, costs and cost-effectiveness of CHW interventions, and characteristics of CHW training.

Data sources:

We searched MEDLINE^®, Cochrane Collaboration resources, and the Cumulative Index to Nursing and Allied Health Literature for studies published in English from 1980 through November 2008.

Review methods:

We used standard Evidence-based Practice Center methods of dual review of abstracts, full-text articles, abstractions, quality ratings, and strength of evidence grades. We resolved disagreements by consensus.

Results:

We included 53 studies on characteristics and outcomes of CHW interventions, 6 on cost-effectiveness, and 9 on training. CHWs interacted with participants in a broad array of locations, using a spectrum of materials at varying levels of intensity. We classified 8 studies as low intensity, 18 as moderate intensity, and 27 as high intensity, based on the type and duration of interaction.

Regarding outcomes, limited evidence (five studies) suggests that CHW interventions can improve participant knowledge when compared with alternative approaches such as no intervention, media, mail, or usual care plus pamphlets. We found mixed evidence for CHW effectiveness on participant behavior change (22 studies) and health outcomes (27 studies): some studies suggested that CHW interventions can result in greater improvements in participant behavior and health outcomes when compared with various alternatives, but other studies suggested that CHW interventions provide no statistically different benefits than alternatives. Low or moderate strength of evidence suggests that CHWs can increase appropriate health care utilization for some interventions (30 studies). The literature showed mixed results of effectiveness when analyzed by clinical context: CHW interventions had the greatest effectiveness relative to alternatives for some disease prevention, asthma management, cervical cancer screening, and mammography screening outcomes. CHW interventions were not significantly different from alternatives for clinical breast examination, breast self-examination, colorectal cancer screening, chronic disease management, or most maternal and child health interventions.

Six studies with economic and cost information yielded insufficient data to evaluate the cost-effectiveness of CHW interventions relative to other community health interventions.

Limited evidence described characteristics of CHW training; no studies examined the impact of CHW training on health outcomes.

Conclusions:

CHWs can serve as a means of improving outcomes for underserved populations for some health conditions. The effectiveness of CHWs in numerous areas requires further research that addresses the methodological limitations of prior studies and that contributes to translating research into practice.