Group Sequential Methods with Applications to Clinical Trials

Jennison, Christopher; Turnbull, Bruce W.

doi:10.1201/9780367805326

Cited by 484 publications

(423 citation statements)

References 0 publications

Supporting

Mentioning

394

Contrasting

Order By: Relevance

“…Decision rules for stopping consist in setting boundaries for a predetermined test statistic so that some error probability requirements are satisfied. There is a rich literature on this topic [ 36 , 37 , 38 , 39 ]. Adaptive early termination has been applied, for instance, in the recent 2020 World Health Organization Solidarity Trial for COVID-19 [ 40 ].…”

Section: Adaptive Designs: the Frequentist Approachmentioning

confidence: 99%

The Bayesian Design of Adaptive Clinical Trials

Giovagnoli

2021

IJERPH

View full text Add to dashboard Cite

This paper presents a brief overview of the recent literature on adaptive design of clinical trials from a Bayesian perspective for statistically not so sophisticated readers. Adaptive designs are attracting a keen interest in several disciplines, from a theoretical viewpoint and also—potentially—from a practical one, and Bayesian adaptive designs, in particular, have raised high expectations in clinical trials. The main conceptual tools are highlighted here, with a mention of several trial designs proposed in the literature that use these methods, including some of the registered Bayesian adaptive trials to this date. This review aims at complementing the existing ones on this topic, pointing at further interesting reading material.

show abstract

Section: Adaptive Designs: the Frequentist Approachmentioning

confidence: 99%

The Bayesian Design of Adaptive Clinical Trials

Giovagnoli

2021

IJERPH

View full text Add to dashboard Cite

show abstract

“…The design and proofs of these two theorems can also be extended to other data types using standard arguments given, for instance, by Jennison and Turnbull. 7 In the situation of normal data with unknown variance, Student's t statistics are used. The null distributions of all needed statistics are well defined and can be approximated by simulation (after the rescaling of case (b), one could use the lower original degrees of freedom to be mildly conservative).…”

Section: Characterization Of the Proceduresmentioning

confidence: 99%

“…The roles of preliminary clinical and pre-clinical studies, pharmacokinetics and pharmacodynamics, and expert judgment in their choice and their use deserve further study. Existing methods for point and interval estimation such as those based on bias-adjusted estimators, 10 test inversion, and median-unbiased point estimators are applicable, 7,[11][12][13] but the specific implementations of these methods and comparisons of their performances are worth consideration.…”

Section: Stage 2 Randomize 2n 2 Additional Patients Equally To D Imentioning

confidence: 99%

Adding dose modifications into Phase II and Phase II/III seamless trials

Spivack

Cheng

Levin

2019

Stat Methods Med Res

View full text Add to dashboard Cite

We present a technique for adding dose modifications into seamless Phase II and Phase II/III trials featuring dose selection at an interim analysis. The method is convenient to apply and can be used either in a fully prespecified, structured way or as a response to new considerations that emerge at interim. Strong control of the familywise error rate regarding false declarations of efficacy versus control is maintained. Two examples are given. One illustrates how the method could potentially “save” a trial performed in a Phase II context. The other is a seamless Phase II/III trial that uses an adaptive exploration strategy for an assumed nonmonotonic dose-response curve. It can result in greatly improved efficiency over a standard “promote the winner” rule.

show abstract

“…Such sequential designs have a long history, with Wald 1 being the earliest proponent for quality control applications, and Armitage 2 extending it to the biomedical field. Group sequential designs are also driven by ethical, financial and administrative requirements as noted by Enas et al, 3 Jennison and Turnbull, 4 and Ellenberg et al 5 The statistical literature on GSDs is vast with many books published on this topic such as the books by Whitehead, 6 Jennison and Turnbull, 7 Proschan et al, 8 Dmitrienko et al, 9 and Wassner and Brannath. 10 Additionally, overview articles on this topic can be found in Whitehead, 11 Todd, 12 and Mazumdar and Bang.…”

Section: Introductionmentioning

confidence: 99%

“…Alternatively, upper and/or lower boundaries can also be drawn based on p-values from a normal distribution. Some authors such as Jennison and Turnbull 7 have noted that the RST approach used for continuous and binary outcomes can be readily extended to time-to-event outcomes. They show such examples using the log-rank and stratified log-rank test, and separately, using the assumption of proportional hazards (PH).…”

Section: Introductionmentioning

confidence: 99%

Group sequential design for time-to-event data using the concept of proportional time

Phadnis¹,

Mayo

2019

Stat Methods Med Res

View full text Add to dashboard Cite

Sequential monitoring of efficacy and safety is an important part of clinical trials. A Group Sequential design allows researchers to perform interim monitoring after groups of patients have completed the study. Statistical literature is well developed for continuous and binary outcomes and relies on asymptotic normality of the test statistic. However, in the case of time-to-event data, existing methods of sample size calculation are done either assuming proportional hazards or assuming exponentially distributed lifetimes. In scenarios where these assumptions are not true, as evidenced from historical data, these traditional methods are restrictive and cannot always be used. As interim monitoring is driven by ethical, financial, and administrative considerations, it is imperative that sample size calculations be done in an efficient manner keeping in mind the specific needs of a clinical trial with a time-to-event outcome. To address these issues, a novel group sequential design is proposed using the concept of Proportional Time. This method utilizes the generalized gamma ratio distribution to calculate the efficacy and safety boundaries and can be used for all distributions that are members of the generalized gamma family using an error spending approach. The design incorporates features specific to survival data such as loss to follow-up, administrative censoring, varying accrual times and patterns, binding or non-binding futility rules with or without skips, and flexible alpha and beta spending mechanisms. Three practical examples are discussed, followed by discussion of the important aspects of the proposed design.

show abstract

Group Sequential Methods with Applications to Clinical Trials

Cited by 484 publications

References 0 publications

The Bayesian Design of Adaptive Clinical Trials

The Bayesian Design of Adaptive Clinical Trials

Adding dose modifications into Phase II and Phase II/III seamless trials

Group sequential design for time-to-event data using the concept of proportional time

Contact Info

Product

Resources

About