Table 4.

Assessment of the robustness (using E-values ¹) of observed associations for all-cause mortality with steroid exposure presented in Figure 2.

Timing of Steroid Exposure Prior to ICI Initiation	0 to ≤3 Months Post ICI Initiation Hazard Ratios	3 to ≤6 Months Post ICI Initiation Hazard Ratios	≥6 Months Post ICI initiation Hazard Ratios 1
Steroids ≤ 1 month prior to ICI	3.27	3.44	1.58
Steroids 1 to ≤ 3 months prior to ICI	2.82	1.77	1.49
Steroids 3 to 12 months prior to ICI	1.89	1.69	1.40

¹ E-value measures the minimum strength of association that an unmeasured confounder must have with both the exposure and outcome, and considers the covariates to nullify the observed exposure–outcome association. The further an E-value shown here is from 1, the less likely that unmeasured confounding would account for its corresponding HR in Figure 2.