Example: 5-FU Trial (continued) The 5-FU trial was designed according to the optimal Bayesian feasible dose escalation method known as EWOC (Babb et al., 1998). For this trial the feasibility bound was set equal to a = 0.25, this value being a compromise between the therapeutic aim of the trial and the need to avoid treatment attributable toxicity. Consequently, escalation of 5-FU between successive patients was to the dose level determined to have posterior probability equal to 0.25 of being an overdose (i.e., greater than the MTD). The ﬁrst patient accrued to the trial received the preselected dose 140 mg/m 2 . Based on the EWOC algorithm, as implemented according to Rogatko and Babb (1998), the doses administered the next four patients were selected according to the schedule given in Figure 3. In contrast to the Bayesian feasible methods, the prediction approaches of Robinson (1978) and Shih (1989) provide sequential search procedures which control the probability that a patient will exhibit DLT. Their formulation is non-Bayesian, being based on the coverage distri-
This is the end of the preview.
access the rest of the document.
Maximum likelihood, Likelihood function, Bayesian inference, Bayesian statistics, dLt