Fig. 4

Representative Plots of Interaction between ARB dose and Effect of Angiotensin-II on Main Study Outcomes. A–D Displays marginal effect sizes from multivariable models for key study outcomes in angiotensin-II and placebo arms at clinically relevant ARB doses in losartan equivalents. The models excluded the n = 29 ACEi-exposed patients (total sample n = 292). Losartan equivalents were modeled on a natural log scale to accommodate observed distributions. Unexposed patients were considered to have an ARB dose = 0 Log(mg). To facilitate interpretation, the estimates were plotted for the corresponding untransformed values at clinically relevant ARB doses along the X-axes. Results are plotted at representative time points for MAP (hr1), NED (hr6), study drug dose (hr3), and change in renin (hr3). Circles/squares show point estimates for placebo/angiotensin-II arms, respectively. Error bars 95% CIs. Key effect estimates from the multivariable models are displayed on the graph as follows: Ang-II = the treatment effect (i.e., in ARB-unexposed patients), Log(mg) = the main-effect of losartan equivalents (i.e., the effect in the placebo group), and Interaction = the interaction-effect of losartan equivalents with treatment (i.e., change in treatment effect per log(mg) increase in ARB dose). Brackets show the 95% confidence intervals for the effects. Full model output, including main and interaction effect sizes, and their confidence intervals can be found in Additional file 1: Tables S4–S8. Ang-II angiotensin-II, ACEi angiotensin converting enzyme inhibitor, ARB angiotensin receptor blocker, MAP mean arterial pressure, NED norepinephrine equivalent dose