Abstract
Abstract Unmeasured confounding is an important threat to the validity of observational studies. A common way to deal with unmeasured confounding is to compute bounds for the causal effect of interest, that is, a range of values that is guaranteed to include the true effect, given the observed data. Recently, bounds have been proposed that are based on sensitivity parameters, which quantify the degree of unmeasured confounding on the risk ratio scale. These bounds can be used to compute an E-value, that is, the degree of confounding required to explain away an observed association, on the risk ratio scale. We complement and extend this previous work by deriving analogous bounds, based on sensitivity parameters on the risk difference scale. We show that our bounds can also be used to compute an E-value, on the risk difference scale. We compare our novel bounds with previous bounds through a real data example and a simulation study.
Highlights
The estimation of causal effects in observational studies is often hampered by unmeasured confounding
Novel bounds for causal effects based on sensitivity parameters on the risk difference scale 191 the literature, but these typically require rather special conditions, such as a single binary confounder [e.g., refs 8–10] or no exposure–confounder interaction [e.g., refs 11–13]; we refer to ref. [7] for a thorough review
Some subject matter experts may find it more intuitive to speculate about the degree of unmeasured confounding on the risk difference scale, regardless of the chosen scale for the target causal effect
Summary
The estimation of causal effects in observational (non-randomized) studies is often hampered by unmeasured confounding. A common way to deal with unmeasured confounding is to compute bounds for the causal effect of interest, that is, a range of values that is guaranteed to include the true effect, given the observed data. DV derived bounds for the causal exposure effect, as functions of the sensitivity parameters and the observed data distribution. Novel bounds for causal effects based on sensitivity parameters on the risk difference scale 191 the literature, but these typically require rather special conditions, such as a single binary confounder [e.g., refs 8–10] or no exposure–confounder interaction [e.g., refs 11–13]; we refer to ref. The bounds that we derive are functions of sensitivity parameters that quantify the degree of unmeasured confounding on the risk difference scale. Where the first equality follows from the law of total probability, the second from conditional exchangeability (3) and the third from consistency (1)
Published Version (Free)
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have