A note on a sensitivity analysis for unmeasured confounding, and the related E-value