Atnaujinkite slapukų nuostatas

El. knyga: Hybrid Frequentist/Bayesian Power and Bayesian Power in Planning Clinical Trials [Taylor & Francis e-book]

  • Formatas: 188 pages, 24 Tables, black and white; 36 Line drawings, color; 36 Illustrations, color
  • Serija: Chapman & Hall/CRC Biostatistics Series
  • Išleidimo metai: 20-Jun-2022
  • Leidėjas: Chapman & Hall/CRC
  • ISBN-13: 9781003218531
Kitos knygos pagal šią temą:
  • Taylor & Francis e-book
  • Kaina: 170,80 €*
  • * this price gives unlimited concurrent access for unlimited time
  • Standartinė kaina: 244,00 €
  • Sutaupote 30%
Hybrid Frequentist/Bayesian Power and Bayesian Power in Planning Clinical TrialsDidesnis paveikslėlis
  • Formatas: 188 pages, 24 Tables, black and white; 36 Line drawings, color; 36 Illustrations, color
  • Serija: Chapman & Hall/CRC Biostatistics Series
  • Išleidimo metai: 20-Jun-2022
  • Leidėjas: Chapman & Hall/CRC
  • ISBN-13: 9781003218531
Kitos knygos pagal šią temą:
Taylor & Francis e-booksMore info about Taylor & Francis e-books
Partnerių interneto svetainė: https://www.taylorfrancis.com/books/9781003218531
Hybrid Frequentist/Bayesian Power and Bayesian Power in Planning Clinical Trials provides a practical introduction to unconditional approaches to planning randomised clinical trials, particularly aimed at drug development in the pharmaceutical industry. This book is aimed at providing guidance to practitioners in using average power, assurance and related concepts. This book brings together recent research and sets them in a consistent framework and provides a fresh insight into how such methods can be used.

Features:





A focus on normal theory linking average power, expected power, predictive power, assurance, conditional Bayesian power and Bayesian power. Extensions of the concepts to binomial, and time-to-event outcomes and non-inferiority trials An investigation into the upper bound on average power, assurance and Bayesian power based on the prior probability of a positive treatment effect Application of assurance to a series of trials in a development program and an introduction of the assurance of an individual trial conditional on the positive outcome of an earlier trial in the program, or to the successful outcome of an interim analysis Prior distribution of power and sample size Extension of the basic approach to proof-of-concept trials with dual success criteria Investigation of the connection between conditional and predictive power at an interim analysis and power and assurance Introduction of the idea of surety in sample sizing of clinical trials based on the width of the confidence intervals for the treatment effect, and an unconditional version.
List of Figures
xi
List of Tables
xiii
Preface xv
Acknowledgements xix
Author xxi
List of Acronyms
xxiii
1 Introduction
1(8)
2 AH Power Is Conditional Unless It's Absolute
9(24)
2.1 Introduction
9(1)
2.2 Expected, Average and Predicted Power
10(8)
2.2.1 Averaging Conditional Power with Respect to the Prior -- Analytic Calculation
11(3)
2.2.2 Calculating the Probability of Achieving "Significance" -- Predictive Power
14(2)
2.2.3 Averaging Conditional Power with Respect to the Prior -- Numerical Integration
16(1)
2.2.4 Averaging Conditional Power with Respect to the Prior -- Simulation
17(1)
2.3 Bounds on Average Power
18(3)
2.4 Average Power for a Robust Prior
21(3)
2.5 Decomposition of Average Power
24(5)
2.6 Average Power -- Variance Estimated
29(4)
2.6.1 Bound on Average Power when the Variance Is Estimated
31(2)
3 Assurance
33(26)
3.1 Introduction
33(1)
3.2 Basic Considerations
33(1)
3.3 Sample Size for a Given Average Power/Assurance
34(3)
3.4 Sample Size for a Given Normalised Assurance
37(1)
3.5 Applying Assurance to a Series of Studies
38(6)
3.6 A Single Interim Analysis in A Clinical Trial
44(5)
3.7 Non-Inferiority Trials
49(10)
3.7.1 Fixed Margin
52(2)
3.7.2 Synthesis Method
54(3)
3.7.3 Bayesian Methods
57(2)
4 Average Power in Non-Normal Settings
59(16)
4.1 Average Power Using a Truncated-Normal Prior
59(1)
4.2 Average Power When the Variance Is Unknown: (a) Conditional on a Fixed Treatment Effect
60(1)
4.3 Average Power When the Variance Is Unknown: (b) Joint Prior on Treatment Effect and Variance
61(3)
4.4 Average Power When the Response Is Binary
64(4)
4.5 Illustrating the Average Power Bound for a Binary Endpoint
68(1)
4.6 Average Power in a Survival Context
69(6)
4.6.1 An Asymptotic Approach to Determining the AP
69(2)
4.6.2 The Average Power for the Comparison of One Parameter Exponential Distributions
71(1)
4.6.3 A Generalised Approach to Simulation of Assurance for Survival Models
72(1)
Note
73(2)
5 Bayesian Power
75(12)
5.1 Introduction
75(1)
5.2 Bayesian Power
75(1)
5.3 Sample Size for a Given Bayesian Power
76(1)
5.4 Bound on Bayesian Power
77(2)
5.5 Sample Size for a Given Normalised Bayesian Power
79(1)
5.6 Bayesian Power When the Response Is Binary
80(1)
5.7 Posterior Conditional Success Distributions
81(6)
5.7.1 Posterior Conditional Success Distributions -- Success Defined By Significance
82(2)
5.7.2 Posterior Conditional Success Distributions -- Success Defined By a Bayesian Posterior Probability
84(1)
5.7.3 Use of Simulation to Generate Samples from the Posterior Conditional Success and Failure Distributions
85(1)
5.7.4 Use of the Posterior Conditional Success and Failure Distributions to Investigate Selection Bias
86(1)
6 Prior Distributions of Power and Sample Size
87(14)
6.1 Introduction
87(1)
6.2 Prior Distribution of Study Power -- Known Variance
88(4)
6.3 Prior Distribution of Study Power -- Treatment Effect Fixed, Uncertain Variance
92(2)
6.4 Prior Distribution of Study Sample Size -- Variance Known
94(2)
6.5 Prior Distribution of Sample Size -- Treatment Effect Fixed, Uncertain Variance
96(2)
6.6 Prior Distribution of Study Power and Sample Size -- Uncertain Treatment Effect and Variance
98(1)
6.7 Loss Functions and Summaries of Prior Distributions
99(2)
7 Interim Predictions
101(12)
7.1 Introduction
101(2)
7.2 Conditional and Predictive Power
103(6)
7.3 Stopping for Futility Based on Predictive Probability
109(2)
7.4 "Proper Bayesian" Predictive Power
111(2)
8 Case Studies in Simulation
113(14)
8.1 Introduction
113(1)
8.2 Case Study 1 -- Proportional Odds Primary Endpoint
114(6)
8.2.1 Background
114(1)
8.2.2 The Wilcoxon Test for Ordered Categorical Data
115(2)
8.2.3 Applying Conditional Power to the Proportional Odds Wilcoxon Test
117(1)
8.2.4 Statistical Approach to Control Type I Error
118(1)
8.2.5 Simulation Set-Up
119(1)
8.2.6 Simulation Results
120(1)
8.3 Case Study 2 -- Unplanned Interim Analysis
120(7)
8.3.1 Background
121(1)
8.3.2 Interim Data
121(1)
8.3.3 Model for Prediction
121(6)
9 Decision Criteria in Proof-of-Concept Trials
127(22)
9.1 Introduction
127(1)
9.2 General Decision Criteria for Early Phase Studies
127(1)
9.3 Known Variance Case
128(6)
9.4 Known Variance Case -- Generalised Assurance
134(1)
9.5 Bounds on Unconditional Decision Probabilities for Multiple Decision Criteria
135(1)
9.6 Bayesian Approach to Multiple Decision Criteria
136(4)
9.7 Posterior Conditional Distributions with Multiple Decision Criteria
140(3)
9.8 Estimated Variance Case
143(4)
9.9 Estimated Variance Case -- Generalised Assurance
147(1)
9.10 Discussion
148(1)
10 Surety and Assurance in Estimation
149(22)
10.1 Introduction
149(2)
10.2 An Alternative to Power in Sample Size Determination
151(2)
10.3 Should the Confidence Interval Width Be the Sole Determinant of Sample Size?
153(3)
10.4 Unconditional Sample Sizing Based on CI Width
156(3)
10.4.1 Modified Cook Algorithm
158(1)
10.4.2 Harris et al. (1948) Algorithm
158(1)
10.5 A Fiducial Interpretation of (10.14)
159(12)
Note
160(1)
References
161(10)
Appendix 1 Evaluation of a Double Normal Integral 171(2)
Appendix 2 Besag's Candidate Formula 173(2)
Index 175
Andrew P. Grieve is a Statistical Research Fellow in the Centre of Excellence in Statistical Innovation at UCB Pharma. He is a former Chair of PSI (Statisticians in the Pharmaceutical Industry) and a past-President of the Royal Statistical Society. He has over 45 years of experience as a biostatistician working in the pharmaceutical industry and academia and has been active in most areas of pharmaceutical R&D in which statistical methods and statisticians are intimately involved, including drug discovery, pre-clinical toxicology, pharmaceutical development, pharmacokinetics and pharmacodynamics, phase IIV of clinical development, manufacturing, health economics and clinical operations.
Pastovi nuoroda: https://www.kriso.lt/db/9781003218531_pe.html
Raktažodžiai: