Identifying direction of bias in environmental epidemiology studies
On this page:
Effect estimates from observational epidemiology may be subject to different sources of bias. Bias may lead to under-estimation or over-estimation of the true toxicant effect and can vary in direction across. Many approaches exist for evaluating the risk of bias due to exposure misclassification in environmental epidemiology studies, but these instruments lack instructions for determining the direction or magnitude of any identified bias and seldom address the co-occurrence of multiple sources or types of bias.
We used a systematic evidence mapping approach to identify and synthesize research findings related to determining the direction of bias in observational epidemiology studies. A list of bias-related search terms was developed via expert input and used to query databases such as PubMed and Scopus. A sample set of positive seed articles were used to rank the results, and the top 107 articles underwent full-text screening for relevance using DistillerSR software. Articles were relevant if they included information useful for determining bias directionality and/or magnitude. Study details (e.g., bias type, bias source) were extracted, along with a description of the relevant information.
Out of the 107 screened studies, 64 met relevancy criteria. The majority of studies (n=59) were related to the direction of information biases, 11 of which pertained specifically to misclassification due to measurement error. Preliminary findings indicate… Compilation and synthesis of the concepts on bias directionality as observed in these studies may enhance the implementation of RoB evaluation tools and impact our understanding of bias when using epidemiology data for dose-response analyses in human hazard assessments of chemicals.