My SciELO
Services on Demand
Journal
Article
Indicators
- Cited by SciELO
- Access statistics
Related links
- Cited by Google
- Similars in SciELO
- Similars in Google
Share
Archivos de Prevención de Riesgos Laborales
On-line version ISSN 1578-2549
Abstract
SANCHEZ-NIUBO, Albert; G. FORERO, Carlos and G. BENAVIDES, Fernando. The application of causal diagrams to conceptualize mechanisms in occupational epidemiology. Arch Prev Riesgos Labor [online]. 2016, vol.19, n.2, pp.103-106. ISSN 1578-2549. https://dx.doi.org/10.12961/aprl.2016.19.02.4.
Although a goal of epidemiological research is to identify causal relationships between a risk factor and a health problem, the methodology employed often sacrifices internal validity to gain capacity to detect associations. There are new graphical and statistical methods that can help unravel the possible causal mechanisms and better understand this "black box". This paper presents causal diagrams, one of the most useful tools for mapping out, prior to analysis, whether a possible association is causal or just due to bias. To demonstrate its usefulness, we use occupational health examples, showing how associations may arise through non-causal pathways as a result of bias. In conclusion, we recommend the routine practice of using causal diagrams in epidemiological research.
Keywords : Epidemiology; causality; directed acyclic graphs; bias; confounding; selection bias.