Bayesian Methods for Benefit/Risk Assessment

Slides:



Advertisements
Similar presentations
Bayes rule, priors and maximum a posteriori
Advertisements

1 Bayesian CTS FDA/Industry Workshop September 18, 2003Copyright Pharsight Case Study in the Use of Bayesian Hierarchical Modeling and Simulation.
OPC Koustenis, Breiter. General Comments Surrogate for Control Group Benchmark for Minimally Acceptable Values Not a Control Group Driven by Historical.
Continued Psy 524 Ainsworth
INTRODUCTION TO MACHINE LEARNING Bayesian Estimation.
Hypothesis testing and confidence intervals by resampling by J. Kárász.
Bayesian inference of normal distribution
Uncertainty and confidence intervals Statistical estimation methods, Finse Friday , 12.45–14.05 Andreas Lindén.
Bayesian Estimation in MARK
MPS Research UnitCHEBS Workshop - April Anne Whitehead Medical and Pharmaceutical Statistics Research Unit The University of Reading Sample size.
Sample size optimization in BA and BE trials using a Bayesian decision theoretic framework Paul Meyvisch – An Vandebosch BAYES London 13 June 2014.
What is Statistical Modeling
Predictive Automatic Relevance Determination by Expectation Propagation Yuan (Alan) Qi Thomas P. Minka Rosalind W. Picard Zoubin Ghahramani.
Basics of Statistical Estimation. Learning Probabilities: Classical Approach Simplest case: Flipping a thumbtack tails heads True probability  is unknown.
1 Equivalence and Bioequivalence: Frequentist and Bayesian views on sample size Mike Campbell ScHARR CHEBS FOCUS fortnight 1/04/03.
Results 2 (cont’d) c) Long term observational data on the duration of effective response Observational data on n=50 has EVSI = £867 d) Collect data on.
Dynamic Treatment Regimes, STAR*D & Voting D. Lizotte, E. Laber & S. Murphy Psychiatric Biostatistics Symposium May 2009.
Using ranking and DCE data to value health states on the QALY scale using conventional and Bayesian methods Theresa Cain.
Computer vision: models, learning and inference
July 3, A36 Theory of Statistics Course within the Master’s program in Statistics and Data mining Fall semester 2011.
Bayesian Statistics in Clinical Trials Case Studies: Agenda
Modeling Menstrual Cycle Length in Pre- and Peri-Menopausal Women Michael Elliott Xiaobi Huang Sioban Harlow University of Michigan School of Public Health.
Overview G. Jogesh Babu. Probability theory Probability is all about flip of a coin Conditional probability & Bayes theorem (Bayesian analysis) Expectation,
Additional Slides on Bayesian Statistics for STA 101 Prof. Jerry Reiter Fall 2008.
Prof. Dr. S. K. Bhattacharjee Department of Statistics University of Rajshahi.
Bayesian Extension to the Language Model for Ad Hoc Information Retrieval Hugo Zaragoza, Djoerd Hiemstra, Michael Tipping Presented by Chen Yi-Ting.
HSRP 734: Advanced Statistical Methods June 19, 2008.
Maximum Likelihood Estimator of Proportion Let {s 1,s 2,…,s n } be a set of independent outcomes from a Bernoulli experiment with unknown probability.
Mixture Models, Monte Carlo, Bayesian Updating and Dynamic Models Mike West Computing Science and Statistics, Vol. 24, pp , 1993.
Lecture 4: Statistics Review II Date: 9/5/02  Hypothesis tests: power  Estimation: likelihood, moment estimation, least square  Statistical properties.
Bayesian Approach For Clinical Trials Mark Chang, Ph.D. Executive Director Biostatistics and Data management AMAG Pharmaceuticals Inc.
Confidence Interval & Unbiased Estimator Review and Foreword.
1 METHODS FOR DETERMINING SIMILARITY OF EXPOSURE-RESPONSE BETWEEN PEDIATRIC AND ADULT POPULATIONS Stella G. Machado, Ph.D. Quantitative Methods and Research.
Lecture 2: Statistical learning primer for biologists
Simulation Study for Longitudinal Data with Nonignorable Missing Data Rong Liu, PhD Candidate Dr. Ramakrishnan, Advisor Department of Biostatistics Virginia.
Bayesian Travel Time Reliability
Chapter 20 Classification and Estimation Classification – Feature selection Good feature have four characteristics: –Discrimination. Features.
INTRODUCTION TO CLINICAL RESEARCH Introduction to Statistical Inference Karen Bandeen-Roche, Ph.D. July 12, 2010.
Timothy Aman, FCAS MAAA Managing Director, Guy Carpenter Miami Statistical Limitations of Catastrophe Models CAS Limited Attendance Seminar New York, NY.
A shared random effects transition model for longitudinal count data with informative missingness Jinhui Li Joint work with Yingnian Wu, Xiaowei Yang.
 An exposure-response (E-R) analysis in oncology aims at describing the relationship between drug exposure and survival and in addition aims at comparing.
How does Biostatistics at Roche typically analyze longitudinal data
The Unscented Particle Filter 2000/09/29 이 시은. Introduction Filtering –estimate the states(parameters or hidden variable) as a set of observations becomes.
Matrix Models for Population Management & Conservation March 2014 Lecture 10 Uncertainty, Process Variance, and Retrospective Perturbation Analysis.
Statistical NLP: Lecture 4 Mathematical Foundations I: Probability Theory (Ch2)
- 1 - Preliminaries Multivariate normal model (section 3.6, Gelman) –For a multi-parameter vector y, multivariate normal distribution is where  is covariance.
Density Estimation in R Ha Le and Nikolaos Sarafianos COSC 7362 – Advanced Machine Learning Professor: Dr. Christoph F. Eick 1.
Outline Historical note about Bayes’ rule Bayesian updating for probability density functions –Salary offer estimate Coin trials example Reading material:
Institute of Statistics and Decision Sciences In Defense of a Dissertation Submitted for the Degree of Doctor of Philosophy 26 July 2005 Regression Model.
Canadian Bioinformatics Workshops
Markov Chain Monte Carlo in R
Data Analysis Patrice Koehl Department of Biological Sciences
Bayesian Semi-Parametric Multiple Shrinkage
Multiple Imputation using SOLAS for Missing Data Analysis
Bayesian data analysis
Model Inference and Averaging
Ch3: Model Building through Regression
Predictive distributions
Mark Rothmann U.S. Food and Drug Administration September 14, 2018
Monitoring rare events during an ongoing clinical trial
Comparisons among methods to analyze clustered multivariate biomarker predictors of a single binary outcome Xiaoying Yu, PhD Department of Preventive Medicine.
Aiying Chen, Scott Patterson, Fabrice Bailleux and Ehab Bassily
Statistical NLP: Lecture 4
Ch13 Empirical Methods.
Parametric Methods Berlin Chen, 2005 References:
CS639: Data Management for Data Science
Bayesian Approaches for Benefit-Risk Assessment with Examples Ram Tiwari and Chul Ahn Division of Biostatistics Center for Device and Radiological Health.
Yu Du, PhD Research Scientist Eli Lilly and Company
A Bayesian Design with Conditional Borrowing of Historical Data in a Rare Disease Setting Peng Sun* July 30, 2019 *Joint work with Ming-Hui Chen, Yiwei.
Assessing similarity of curves: An application in assessing similarity between pediatric and adult exposure-response curves July 31, 2019 Yodit Seifu,
Presentation transcript:

Bayesian Methods for Benefit/Risk Assessment Ram C. Tiwari Associate Director Office of Biostatistics, CDER, FDA Ram.Tiwari@fda.hhs.gov

Disclaimer This presentation reflects the views of the author and should not be construed to represent FDA’s views or policies. Benefit-risk Assessment

Benefit-risk Assessment

Outline Introduction Methodology Illustration and simulation study Commonly-used Benefit-risk (BR) measures Methodology BR measures based on Global benefit-risk (GBR) scores and a new measure Bayesian approaches Power prior Illustration and simulation study Future work Benefit-risk Assessment

Introduction The benefit-risk assessment is the basis of regulatory decisions in the pre-market and post market review process. The evaluation of benefit and risk faces several challenges. Benefit-risk Assessment

Commonly used B-R measures Various measures have been proposed to assess benefit and risk simultaneously: Q-TWiST by Gelbert et al. (1989) Ratio of benefit and risk by Payne (1975) The Number Needed to Treat and the Number Needed to Harm by Holden et al. (2003) Global Benefit Risk (GBR) scores by Chuang-Stein et al. (1991) Benefit-risk Assessment

BR categories A five-category multinomial random variable to capture the benefit and risk of a drug product on each individual simultaneously: Table 1: Possible outcomes of a clinical trial with binary response data   Benefit No benefit No AE Category 1 Category 3 AE Category 2 Category 4 withdrawal Category 5 Benefit-risk Assessment

Example: Hydromorphone Data was provided by Jonathan Norton. Benefit-risk Assessment

GBR scores Benefit-risk Assessment

Methodology: BR measures BR measures based on the global scores proposed by Chuang-Stein et al. BR measures based on the global scores are for each arm (treatment and comparator) separately. BR_Linear can take a continuous value on a scale of -4 to 4 (inclusive). Benefit-risk Assessment

Methodology: New BR measure A new indicator based measure is proposed: BR_Indicator compares two arms simultaneously. It takes a integer value between -6 to 6 (inclusive). Benefit-risk Assessment

Methodology: Dirichlet prior Dirichlet distribution is used as the conjugate prior for multinomial distribution, and the posterior distribution of the five-category random variable is derived at each visit using sequentially updated posterior as a prior. Benefit-risk Assessment

Methodology: Sequential Updating Sequential updating of the posteriors are given by: The posterior mean (i.e., Bayes estimate) and 95% credible interval for each of the four measures are obtained using a Markov chain Monte Carlo (MCMC) technique. Benefit-risk Assessment

Methodology: Decision Rules For a BR measure, If the credible interval include the value zero, the benefit does not outweigh the risk; If the lower bound of the credible interval is greater than zero, the benefit outweighs the risk; If the upper bound of the credible interval is less than zero, the risk outweighs the benefit. Benefit-risk Assessment

Methodology: Power Prior Power prior (Ibrahim and Chen, 2000) is used through the likelihood function to discount the information from previous visits, and the posterior distribution of the five-category random variable is obtained using the Dirichlet prior for p and a Beta (1, 1) as a power prior for . Benefit-risk Assessment

Methodology: Model Fit The model fit of the two models (with and without power prior) is assessed through the conditional predictive ordinate (CPO) and the logarithm of the pseudo-marginal likelihood (LPML). The larger the value of LPML, the better fit the model is. Here, n(i) is the data with ni removed. Benefit-risk Assessment

Back to our example: Hydromorphone Benefit-risk Assessment

Illustration: Posterior Means and 95% Credible Intervals for BR_Linear Measure without power prior with power prior Benefit-risk Assessment

Illustration: Posterior Means and 95% Credible Intervals for BR_Indicator Measure without power prior with power prior Benefit-risk Assessment

Illustration: Results a. The model without power prior b. The model with power prior Benefit-risk Assessment

Illustration: Posterior Means and 95% Credible Intervals for Power Prior Parameter Benefit-risk Assessment

Illustration: Model Fit   LPML values Treatment Control Model without power prior -14.230 -14.209 Model with power prior -6.432 -6.190 Benefit-risk Assessment

Simulation study Correlated longitudinal multinomial data are simulated using the R package SimCorMultRes.R, which uses an underlying regression model to draw correlated ordinal response. Two scenarios are simulated: The treatment arm is similar to the control arm in terms of benefit-risk; The treatment arm is better than control arm in the sense that the benefit outweighs risk. Benefit-risk Assessment

Simulation study: Scenarios Scenario 1: Treatment benefit does not outweigh risk compared to control Scenario 2: Treatment benefit outweighs risk compared to control Benefit-risk Assessment

Simulation study: Scenario 1 Treatment benefit does not outweigh risk compared to control a. The model without power prior b. The model with power prior Benefit-risk Assessment

Simulation study: Scenario 2 Treatment benefit outweighs risk compared to control a. The model without power prior b. The model with power prior Benefit-risk Assessment

Simulation study: Results Scenario 1: Treatment benefit does not outweigh risk compared to control Scenario 2: Treatment benefit outweighs risk compared to control Benefit-risk Assessment

Simulation study: Model Fit   LPML values Treatment Control Scenario 1: Model without power prior -23.536 -23.354 Model with power prior -8.472 -7.667 Scenario 2: -27.099 -21.840 -8.532 -8.393 Benefit-risk Assessment

Future Work in BR Assessment Benefit-risk Assessment

Future work in BR assessment Frequentist approaches: Bootstrap approach General linear mixed model (GLMM) approach Other Bayesian approaches: Normal priors Dirichlet process Benefit-risk Assessment

Bootstrap Approach Approximate underlying distribution using the empirical distribution of the observed data; Resample from the original dataset; Calculate the estimates and confidence intervals (CIs) of the BR measures based on the bootstrap samples; Percentile bootstrap CIs; Basic bootstrap CIs; Studentized bootstrap CIs; Bias-Corrected and Accelerated CIs. Apply the decision rules. Benefit-risk Assessment

Bootstrap Approach-Results Benefit-risk Assessment

General linear mixed model (GLMM) approach Within each arm (T or C), the ith subject falls in the jth category (vs. the first category) at kth visit can be modeled as, where, α0 is the baseline effect assumed common across all categories, βj is the category effect, and γk is the longitudinal effect at kth visit, with and, . Benefit-risk Assessment

GLMM approach Note that different variance-covariance structures can be used for (γ1,γ2,…γ8), to model the longitudinal trend. Compound-symmetry Power covariance structure Unstructured covariance structure The estimates of the confidence intervals of the global measures can be derived from Monte Carlo samples, and the decision rules can be determined based on the confidence intervals. Benefit-risk Assessment

General linear mixed model approach-Results Benefit-risk Assessment

Bayesian approaches with GLMM (α0 , βj ; j=1,…,5)~ independent Normal with means 0 and large variances; Variance parameters~ IG Dirichlet Process Approach: Let α0 to depend on subjects, that is, assume that α0i |G ~ iid G, with G~ DP(M, G0), M>0 concentration parameter and G0 a baseline distribution such as a normal with mean 0 and large variance. βj ; j=1,…,5 are independent normal with means 0, and large variances. The posterior distributions of the probability and the global measures can be derived, and the decision rules can be determined based on the credible intervals. Benefit-risk Assessment

Discussion Quantitative measure of benefit and risk is an important aspect in the drug evaluation process. The Bayesian method is a natural method for longitudinal data by sequentially updating the prior; Power prior can be used to discount information from previous visits. Frequentist approaches such as bootstrapping method and general linear mixed model can be applied for benefit risk assessment. Continuous research in longitudinal assessment of drug benefit-risk is warranted. Benefit-risk Assessment

Benefit-risk Assessment

Selected References Gelber RD, Gelman RS, Goldhirsch A. A quality-of-life oriented endpoint for comparing treatments. Biometrics. 1989;45:781-795 Payne JT, Loken MK. A survey of the benefits and risks in the practices of radiology. CRC Crit Rev Clin Radiol Nucl Med. 1975; 6:425-475 Holden WL, Juhaeri J, Dai W. “Benefit-Risk Analysis: A Proposal Using Quantitative Methods,” Pharmacoepidemiology and Drug Safety. 2003; 12, 611–616. 154 Chuang-Stein C, Mohberg NR, Sinkula MS. Three measures for simultaneously evaluating benefits and risks using categorical data from clinical trials. Statistics in Medicine. 1991; 10:1349-1359. Norton, JD. A Longitudinal Model and Graphic for Benefit-risk Analysis, with Case Study. Drug Information Journal. 2011; 45: 741-747. Ibrahim, JG, Chen, MH. Power Prior Distributions for Regression Models. Statistical Science. 2000; 15: 46-60. Benefit-risk Assessment

Q & A Benefit-risk Assessment