Comparing three regularization methods to avoid extreme allocation probability in response-adaptive randomization

Du, Yining; Cook, John D.; Lee, John

doi:10.1080/10543406.2017.1293077

Cited by 9 publications

(12 citation statements)

References 27 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…It is a common practice in bandit approaches to apply a “burn-in phase” when using a Bayesian decision rule (Du et al, 2018; Kano et al, 2017; McInerney, Roberts, & Rezek, 2010). A burn-in phase helps to minimize the effect of outliers due to sampling error at the beginning of the experiment.…”

Section: Overview Of Decision Rules For Mab Approachesmentioning

confidence: 99%

“…Apart from a burn-in phase, there are also other regulatory methods in the respective literature that can be used to avoid extreme allocation probabilities, such as the “power transformation method” and the “clip method” (cf. Du et al, 2018). All these methods can be used to adjust the level at which exploration and exploitation are blended.…”

Section: Limitations and Practical Constraintsmentioning

confidence: 99%

See 1 more Smart Citation

Rethinking the Gold Standard With Multi-armed Bandits: Machine Learning Allocation Algorithms for Experiments

Kaibel

Biemann

2019

Organizational Research Methods

View full text Add to dashboard Cite

In experiments, researchers commonly allocate subjects randomly and equally to the different treatment conditions before the experiment starts. While this approach is intuitive, it means that new information gathered during the experiment is not utilized until after the experiment has ended. Based on methodological approaches from other scientific disciplines such as computer science and medicine, we suggest machine learning algorithms for subject allocation in experiments. Specifically, we discuss a Bayesian multi-armed bandit algorithm for randomized controlled trials and use Monte Carlo simulations to compare its efficiency with randomized controlled trials that have a fixed and balanced subject allocation. Our findings indicate that a randomized allocation based on Bayesian multi-armed bandits is more efficient and ethical in most settings. We develop recommendations for researchers and discuss the limitations of our approach.

show abstract

Section: Overview Of Decision Rules For Mab Approachesmentioning

confidence: 99%

Section: Limitations and Practical Constraintsmentioning

confidence: 99%

Rethinking the Gold Standard With Multi-armed Bandits: Machine Learning Allocation Algorithms for Experiments

Kaibel

Biemann

2019

Organizational Research Methods

View full text Add to dashboard Cite

show abstract

“…Measures can be taken, such as starting with an equal randomization phase, then switching to adaptive randomization, and restricting the randomization probability to 0.1-0.9, to avoid extreme allocation and undesirable properties. [72][73][74] There is still a need for new trial methods. As cancer therapy continues to change, some treatments are considered useful if they turn a deadly cancer into a chronic disease.…”

Section: Resultsmentioning

confidence: 99%

Bayesian clinical trials at The University of Texas MD Anderson Cancer Center: An update

et al. 2019

View full text Add to dashboard Cite

Background/aims In our 2009 article, we showed that Bayesian methods had established a foothold in developing therapies in our institutional oncology trials. In this article, we will document what has happened since that time. In addition, we will describe barriers to implementing Bayesian clinical trials, as well as our experience overcoming them. Methods We reviewed MD Anderson Cancer Center clinical trials submitted to the institutional protocol office for scientific and ethical review between January 2009 and December 2013, the same length time period as the previous article. We tabulated Bayesian methods implemented for design or analyses for each trial and then compared these to our previous findings. Results Overall, we identified 1020 trials and found that 283 (28%) had Bayesian components so we designated them as Bayesian trials. Among MD Anderson–only and multicenter trials, 56% and 14%, respectively, were Bayesian, higher rates than our previous study. Bayesian trials were more common in phase I/II trials (34%) than in phase III/IV (6%) trials. Among Bayesian trials, the most commonly used features were for toxicity monitoring (65%), efficacy monitoring (36%), and dose finding (22%). The majority (86%) of Bayesian trials used non-informative priors. A total of 75 (27%) trials applied Bayesian methods for trial design and primary endpoint analysis. Among this latter group, the most commonly used methods were the Bayesian logistic regression model (N = 22), the continual reassessment method (N = 20), and adaptive randomization (N = 16). Median institutional review board approval time from protocol submission was the same 1.4 months for Bayesian and non-Bayesian trials. Since the previous publication, the Biomarker-Integrated Approaches of Targeted Therapy for Lung Cancer Elimination (BATTLE) trial was the first large-scale decision trial combining multiple treatments in a single trial. Since then, two regimens in breast cancer therapy have been identified and published from the cooperative Investigation of Serial Studies to Predict Your Therapeutic Response with Imaging and Molecular Analysis (I-SPY 2), enhancing cooperation among investigators and drug developers across the nation, as well as advancing information needed for personalized medicine. Many software programs and Shiny applications for Bayesian trial design and calculations are available from our website which has had more than 21,000 downloads worldwide since 2004. Conclusion Bayesian trials have the increased flexibility in trial design needed for personalized medicine, resulting in more cooperation among researchers working to fight against cancer. Some disadvantages of Bayesian trials remain, but new methods and software are available to improve their function and incorporation into cancer clinical research.

show abstract

“…Advances in approximation and optimization methods can help alleviate the computational barrier. Also, many authors have provided ad hoc modifications to curtail the extreme allocation property . It is instructive yet complicated to conduct comparisons.…”

Section: Discussionmentioning

confidence: 99%

Frequentist operating characteristics of Bayesian optimal designs via simulation

Zhang

Trippa

Parmigiani

2019

Statistics in Medicine

View full text Add to dashboard Cite

Bayesian adaptive designs have become popular because of the possibility of increasing the number of patients treated with more beneficial treatments, while still providing sufficient evidence for treatment efficacy comparisons. It can be essential, for regulatory and other purposes, to conduct frequentist analyses both before and after a Bayesian adaptive trial, and these remain challenging. In this paper, we propose a general simulation-based approach to compare frequentist designs with Bayesian adaptive designs based on frequentist criteria such as power and to compute valid frequentist p-values. We illustrate our approach by comparing the power of an equal randomization (ER) design with that of an optimal Bayesian adaptive (OBA) design. The Bayesian design considered here is the dynamic programming solution of the optimization of a specific utility function defined by the number of successes in a patient horizon, including patients whose treatment will be affected by the trial's results after the end of the trial.While the power of an ER design depends on treatment efficacy and the sample size, the power of the OBA design also depends on the patient horizon size. Our results quantify the trade-off between power and the optimal assignment of patients to treatments within the trial. We show that, for large patient horizons, the two criteria are in agreement, while for small horizons, differences can be substantial. This has implications for precision medicine, where patient horizons are decreasing as a result of increasing stratification of patients into subpopulations defined by molecular markers. KEYWORDSBayesian adaptive designs, dynamic programming, frequentist analyses, optimal strategy Spiegelhalter et al 5 and Cellamare et al 6 have illustrated the use of Bayesian approaches in clinical trials and discussed their advantages. Ethical concerns 7 associated with adaptive designs are also relevant in Bayesian adaptive settings. On the other hand, frequentist criteria remain the mainstay in regulatory decision-making and in the medical literature. This applies to both trial design requirements, such as the power at a given type I error level, and reported 4026

show abstract

Comparing three regularization methods to avoid extreme allocation probability in response-adaptive randomization

Cited by 9 publications

References 27 publications

Rethinking the Gold Standard With Multi-armed Bandits: Machine Learning Allocation Algorithms for Experiments

Rethinking the Gold Standard With Multi-armed Bandits: Machine Learning Allocation Algorithms for Experiments

Bayesian clinical trials at The University of Texas MD Anderson Cancer Center: An update

Frequentist operating characteristics of Bayesian optimal designs via simulation

Contact Info

Product

Resources

About