Int J Med Sci 2013; 10(11):1459-1461. doi:10.7150/ijms.6455 This issue
Short Research Communication
National Drug Research Institute, Curtin University, Perth, Western Australia, Australia
Epidemiological studies that investigate the relationships between health behaviors and diseases may be affected by both known and unknown confounding factors. Alcohol use is one of these behaviors that have been intensively investigated in epidemiological studies. This manuscript introduced a simple test that can identify confounded epidemiological studies. This approach is sensitive to both known and unknown confounders. It provides a new perspective to develop measures for evidence selection in the future.
Keywords: evidence-based medicine, bias, epidemiology, causality, health behaviors