Learning for Dose Allocation in Adaptive Clinical Trials with Safety Constraints

Authors: Cong Shen, Zhiyang Wang, Sofia Villar, Mihaela Van Der Schaar

ICML 2020 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental Through numerical experiments using both synthetic and real-world datasets, we show that SEEDA outperforms state-of-the-art clinical trial designs by finding the optimal dose with higher success rate and fewer patients. and 5. Experiments To investigate the operational characteristics and evaluate the performance of the proposed adaptive designs, we present an experimental study with K = 6 dose levels and n = 300 trial cohorts, with each cohort consists of 3 patients.
Researcher Affiliation Academia Cong Shen 1 University of Virginia, USA 2 Zhiyang Wang University of Pennsylvania, USA 3 Sofía S. Villar University of Cambridge, United Kingdom 4 Mihaela van der Schaar University of California, Los Angeles, USA.
Pseudocode Yes Algorithm 1 The Safe Efficacy Exploration Dose Allocation (SEEDA) Algorithm and Algorithm 2 The SEEDA-Plateau Algorithm
Open Source Code No The paper does not provide explicit statements or links indicating that source code for the described methodology is publicly available.
Open Datasets Yes We evaluate the SEEDA algorithms in two real-world datasets neurodeg and IBSCovars based on (Biesheuvel & Hothorn, 2002).
Dataset Splits No The paper describes using a total number of patients 'n' for trial cohorts but does not specify distinct training, validation, or test dataset splits in the conventional sense. The process involves sequential patient recruitment and adaptive dose allocation.
Hardware Specification No The paper does not provide specific details about the hardware specifications (e.g., GPU/CPU models, memory) used for running its experiments.
Software Dependencies No The paper mentions 'an R package implementation provided by (Yoshida, 2019)' but does not specify version numbers for this package or other software dependencies.
Experiment Setup Yes To investigate the operational characteristics and evaluate the performance of the proposed adaptive designs, we present an experimental study with K = 6 dose levels and n = 300 trial cohorts, with each cohort consists of 3 patients. The estimation is updated after observing all individual outcomes from a cohort. All experiment results are obtained with 1000 trial repetitions. The MTD threshold is set as θ = 0.35.