We provide sensitivity analyses for unmeasured confounding in estimates of effect heterogeneity and causal interaction.
We describe a pregnancy cohort recruited and followed during the COVID-19 pandemic, as well as some challenges encountered during the process.
We show that it is possible to bound the total composite bias due to unmeasured confounding, selection, and differential misclassification, and to use that bound to assess the sensitivity of a risk ratio to any combination of these biases.
Commenting on an article by Rudolph et al., we describe goals of mediation analysis for informing policy.
We extend a sensitivity analysis approach for selection bias to account for poor control selection in case-control studies.
We introduce the concept of mediation and provide examples that solidify understanding of mediation for valid discovery and interpretation in the field of reproductive medicine.
Understanding the way in which a study sample relates to the target population is critical for avoiding and addressing bias. Communication about selection bias is aided by the use of causal graphs.
We show how E-values can be used and easily interpreted for sensitivity analysis in mediation.
We provide bounds for selection bias under a variety of situations.