### Mediator model Call: lm(formula = bili ~ trt, data = data) Residuals: Min 1Q Median 3Q Max -3.4000 -2.5000 -1.7000 0.4434 24.3000 Coefficients: Estimate Std. Error t value Pr(>|t|) (Intercept) 2.2132 0.8784 2.520 0.0123 * trt 0.7434 0.5532 1.344 0.1801 --- Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1 Residual standard error: 4.594 on 274 degrees of freedom Multiple R-squared: 0.006548, Adjusted R-squared: 0.002923 F-statistic: 1.806 on 1 and 274 DF, p-value: 0.1801 ### Outcome model Call: glm(formula = platelet ~ trt + bili, family = poisson(link = "log"), data = data) Coefficients: Estimate Std. Error z value Pr(>|z|) (Intercept) 5.5390793 0.0120167 460.949 < 2e-16 *** trt 0.0323306 0.0074641 4.331 1.48e-05 *** bili -0.0062492 0.0008434 -7.409 1.27e-13 *** --- Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1 (Dispersion parameter for poisson family taken to be 1) Null deviance: 9286.7 on 275 degrees of freedom Residual deviance: 9216.2 on 273 degrees of freedom AIC: 11247 Number of Fisher Scoring iterations: 4 ### Mediation analysis est se Z p lower upper cde 0.03879670 0.008956943 4.331467 1.481194e-05 0.02124141 0.056351983 pnde 0.03879670 0.008956943 4.331467 1.481194e-05 0.02124141 0.056351983 tnie -0.00557464 0.004215782 -1.322326 1.860595e-01 -0.01383742 0.002688142 tnde 0.03879670 0.008956943 4.331467 1.481194e-05 0.02124141 0.056351983 pnie -0.00557464 0.004215782 -1.322326 1.860595e-01 -0.01383742 0.002688142 te 0.03322206 0.009849044 3.373125 7.432012e-04 0.01391829 0.052525829 pm -0.17107851 0.157177384 -1.088442 2.763999e-01 -0.47914053 0.136983498 exp(est) exp(lower) exp(upper) cde 1.0395591 1.0214686 1.057970 pnde 1.0395591 1.0214686 1.057970 tnie 0.9944409 0.9862579 1.002692 tnde 1.0395591 1.0214686 1.057970 pnie 0.9944409 0.9862579 1.002692 te 1.0337801 1.0140156 1.053930 pm NA NA NA Evaluated at: avar: trt a1 (intervened value of avar) = 2.3 a0 (reference value of avar) = 1.1 mvar: bili m_cde (intervend value of mvar for cde) = 1.4 cvar: c_cond (covariate vector value) = Note that effect estimates do not vary over m_cde and c_cond values when interaction = FALSE.