Lessons learned from the application of BEES-C: Systematic assessment of study quality of epidemiologic research on BPA, neurodevelopment, and respiratory health

- BEES-C is the seminal evaluation tool for exposure assessments.
- BEES-C is designed for research on biomonitoring of short-lived chemicals.
- BEES-C is tested using BPA, neurodevelopment and respiratory effects literature.
- Recommendations for refining BEES-C are provided.

Epidemiologic studies evaluating associations between biomarkers of exposure to short-lived chemicals and health endpoints in humans face special challenges. Perhaps the most critical challenges are the need to determine the type and optimal number of samples, and the proper timing of specimen collection. Further, as many short-lived chemicals are ubiquitous in the environment, utmost care is required to avoid sample contamination. A separate set of challenges is associated with appropriate interpretation and reporting of results from multiple simultaneous analyses, which are becoming increasingly feasible. The Biomonitoring, Environmental Epidemiology, and Short-Lived Chemicals (BEES-C) instrument is specifically designed to evaluate the quality of epidemiologic studies that measure biomarkers of chemicals with short physiologic half-lives. The instrument provides systematic guidance for evaluating 14 different aspects of study quality divided into three broad categories: 1) biomarker selection and measurement, 2) strategy and execution of exposure assessment, and 3) general considerations of study design and reporting. We evaluated the utility of the BEES-C instrument using epidemiologic studies of exposure to bisphenol A and its association with neurodevelopmental and respiratory health indicators. Each BEES-C element was assessed with respect to needed modifications and concordance among reviewers using professional, scientific judgment. Based on this first use of the BEES-C instrument, we found that most of its elements were effective in comparing the quality of available studies, with reviews generally concordant and justifications consistent. However, we note that certain elements would be improved with slight adjustments and that one of the elements appeared redundant and should be removed.