A novel statistical test and lower bound estimation procedure to quantify the strength of unobserved confounding in observational studies, leveraging data from randomized trials.
