Systematic dose-response of environmental epidemiologic studies: Dose and response pre-analysis

Abstract

Meta-analysis approaches can be used to assess the human risks due to exposure to environmental chemicals when there are numerous high-quality epidemiologic studies of priority outcomes in a database. However, methodological issues related to how different studies report effect measures and incorporate exposure into their analyses arise that complicate the pooled analysis of multiple studies. As such, there are "pre-analysis" steps that are often necessary to prepare summary data reported in epidemiologic studies for dose-response analysis. This paper uses epidemiologic studies of arsenic-induced health effects as a case example and addresses the issues surrounding the estimation of mean doses from censored dose- or exposure-intervals reported in the literature (e.g., estimation of mean doses from high exposures that are only reported as an open-ended interval), calculation of a common dose metric for use in a dose-response meta-analysis (one that takes into consideration inter-individual variability), and calculation of response "effective counts" that inherently account for confounders. The methods herein may be generalizable to 1) the analysis of other environmental contaminants with a suitable database of epidemiologic studies, and 2) any meta-analytic approach used to pool information across studies. A second companion paper detailing the use of "pre-analyzed" data in a hierarchical Bayesian dose-response model and techniques for extrapolating risks to target populations follows.

Keywords: Dose conversion; Dose-response; Inter-individual variability; Meta-analysis; Uncertainty.

Figure 1:. Analysis Flow Chart.

The analyses covered in this paper are highlighted in blue whereas the analyses highlighted in orange and green are covered in the second paper of this series (Allen et al., 2020)

Figure 2:. Dose Pre-Analysis and Uncertainty Flow Chart in Relation to “Best,” Low-end, and High-end Dose Sets

¹ High group means minimized or maximized subject to constraint that −2 × (LL – MLL) ≤ 2.706 (a 95% bound on the high-group mean). LL is the log-likelihood for the lognormal distribution for the candidate parameter vector; MLL is the maximum log-likelihood. When a published study reports the mean or median values for each group, those values are used directly as the group-specific dose values, with no lognormal fitting. ² The terminology “low-end,” “high-end,” and “best” estimates are used to avoid confusing the values with credible (or confidence) interval bounds having a specific numerical value (e.g., 95%). Combining the log-likelihood bounds for group-specific means, with percentiles from the Monte Carlo analysis does not allow determination that the bounding estimates have any identifiable associated “confidence level.” They do, however produce reasonable semi-quantitative limits on how uncertain the resulting estimates are.

Figure 3:. Effect of Group Mean Estimations (MLE, High, Low) on Final Distribution of Daily Intake Dose Estimates for the Lowest and Highest Exposure Group in Chen et al. (2010)

Plot represents the distribution of the percentiles of the dose-estimation MCMC analysis for the low- and high-exposure groups. Frequencies were calculated in Excel using the “Frequency” function and user-defined bin values.