《学术论文集锦[开错]》紫微客

13、013

　　4.3.3 Adaptive dose-ranging trials

　　In addition to the above experimental design for small sample data, based on the actual context of this study (the tumor chemosensitive assay in vitro of HB), we also can conduct adaptive dose ranging studies. Using this method can greatly improve the efficiency and effectiveness of dose selection in the tumor chemosensitive assay in vitro, that is, to adjust the choice of drug concentration applied to the new patients according to the interim analysis and to control the number of patients used as samples in order to reduce the time and cost required for the study operation[14]. In this study, I chose a Design-focused adaptive approach: general adaptive dose allocation approach (GADA). The core of the GADA method is still the Bayesian model, which uses the data obtained from the previous experimental subjects (patients) as prior information, and then the appropriate dose is determined for each new patient by the posterior probability[12]. The key steps of this process are shown in Figure 4.

　　Figure4：GADA approach based on Bayesian theory.
　　Key steps:
　　1.New patients are randomly assigned to a control group or an effective dose experimental group according to the principle of optimal allocation.
　　2.Translating dose allocation decisions into appropriate dosing instructions.
　　3.Send the interim and final results to the central system.
　　4.Use the final results to update the estimated dose-response model.
　　5.According to the posterior probability of response at the target dose, decide whether to stop as nullify, superior effect, or continue dose finding research.
　　6.Select the optimal dose by simulating the effect of the next experimental subject receiving each dose.