Bayesian sensitivity analyses.

(A) Bayesian sensitivity analysis for survival in patients with a normal karyotype. Hypothetical values for the mean GO-IA TE are plotted on the horizontal axis. The solid vertical lines correspond to the mean TEs assumed from the FAIG and FAIG plus ATRA trials (± 0.396 and ± 0.328, respectively). The plotted curves are the posterior probabilities that GO with or without IL-11 is harmful relative to IA (vertical axis) as a function of the hypothesized mean TE. For GO with or without IL-11 to be superior to IA (posterior probability < 0.50), TEs at least 2-fold greater than those observed in the FAIG and FAIG plus ATRA trials must be postulated (dotted lines). The effects must also be postulated to be in the correct direction. (B) Bayesian sensitivity analysis for survival in patients with −5/−7. TEs at least 5-fold greater than those observed in the FAIG and FAIG plus ATRA trials—and TEs in the correct direction—must be postulated for GO with or without IL-11 to be superior to IA (dotted lines). (C) Bayesian sensitivity analysis for survival in patients with other abnormal cytogenic findings. TEs at least 4-fold greater than those observed in the FAIG and FAIG plus ATRA trials—and TEs in the correct direction—must be postulated for GO with or without IL-11 to be superior to IA (dotted lines).