1Carl-Fredrik Burman, 11 Nov 2008 RSS / MRC / NIHR HTA Futility Meeting Futility stopping Carl-Fredrik Burman, PhD Statistical Science Director AstraZeneca.

Slides:



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

Inferential Statistics and t - tests
Bayesian Health Technology Assessment: An Industry Statistician's Perspective John Stevens AstraZeneca R&D Charnwood Bayesian Statistics Focus Team Leader.
NIHR Research Design Service London Enabling Better Research Forming a research team Victoria Cornelius, PhD Senior Lecturer in Medical Statistics Deputy.
Stopping Trials for Futility Ranjit Lall (May 2009)
Bayesian posterior predictive probability - what do interim analyses mean for decision making? Oscar Della Pasqua & Gijs Santen Clinical Pharmacology Modelling.
Hypothesis Testing A hypothesis is a claim or statement about a property of a population (in our case, about the mean or a proportion of the population)
Introduction Clinical trials Why clinical trials? The Clinical Trial Process Informed consent Patients‘ interests Rights and protection Trials Registers.
MPS Research UnitCHEBS Workshop - April Anne Whitehead Medical and Pharmaceutical Statistics Research Unit The University of Reading Sample size.
Optimal Drug Development Programs and Efficient Licensing and Reimbursement Regimens Neil Hawkins Karl Claxton CENTRE FOR HEALTH ECONOMICS.
Elements of Decision Problems
Turning Questions into Trials: Innovation in Surgical Oncology Jennifer E. Rosen MD FACS Assistant Professor of Surgery and Molecular Medicine Boston University.
Obtaining Informed Consent: 1. Elements Of Informed Consent 2. Essential Information For Prospective Participants 3. Obligation for investigators.
How does the process work? Submissions in 2007 (n=13,043) Perspectives.
Focusing on the key challenges Decision-making & drug development Peter Hertzman Paul Miller.
Stopping Trials for Futility RSS/NIHR HTA/MRC 1 day workshop 11 Nov 2008.
Using ranking and DCE data to value health states on the QALY scale using conventional and Bayesian methods Theresa Cain.
Large Phase 1 Studies with Expansion Cohorts: Clinical, Ethical, Regulatory and Patient Perspectives Accelerating Anticancer Agent Development and Validation.
Adaptive Designs for Clinical Trials
© 2002 Prentice-Hall, Inc.Chap 7-1 Statistics for Managers using Excel 3 rd Edition Chapter 7 Fundamentals of Hypothesis Testing: One-Sample Tests.
Inference in practice BPS chapter 16 © 2006 W.H. Freeman and Company.
Clinical Trials. What is a clinical trial? Clinical trials are research studies involving people Used to find better ways to prevent, detect, and treat.
Intervention Studies Principles of Epidemiology Lecture 10 Dona Schneider, PhD, MPH, FACE.
EVIDENCE BASED MEDICINE Health economics Ross Lawrenson.
Background to Adaptive Design Nigel Stallard Professor of Medical Statistics Director of Health Sciences Research Institute Warwick Medical School
© 2003 Prentice-Hall, Inc.Chap 7-1 Business Statistics: A First Course (3 rd Edition) Chapter 7 Fundamentals of Hypothesis Testing: One-Sample Tests.
Introduction: Why statistics? Petter Mostad
A National unit for Bayesian Health Decision Science.
Clinical Trial Designs An Overview. Identify: condition(s) of interest, intended population, planned treatment protocols Recruitment of volunteers: volunteers.
Economic evaluation of drugs for rare diseases CENTRE FOR HEALTH ECONOMICS K Claxton, C McCabe, A Tsuchiya Centre for Health Economics and Department of.
Evidence-Based Public Health Nancy Allee, MLS, MPH University of Michigan November 6, 2004.
Value of information Marko Tainio Decision analysis and Risk Management course in Kuopio
Introduction to inference Use and abuse of tests; power and decision IPS chapters 6.3 and 6.4 © 2006 W.H. Freeman and Company.
1 Ethical issues in clinical trials Bernard Lo, M.D. February 10, 2010.
Good Research, Bad Choices? Mary Coombs. What Makes Something Research Rather Than Treatment?
Validation / citations. Validation u Expert review of model structure u Expert review of basic code implementation u Reproduce original inputs u Correctly.
An agency of the European Union Presented by: Paolo Tomasi Data Safety Monitoring Boards / Data Monitoring Committees in paediatric studies Paolo Tomasi,
How to find a paper Looking for a known paper: –Field search: title, author, journal, institution, textwords, year (each has field tags) Find a paper to.
Developing medicines for the future and why it is challenging Angela Milne.
What is a non-inferiority trial, and what particular challenges do such trials present? Andrew Nunn MRC Clinical Trials Unit 20th February 2012.
Cancer Trials. Reading instructions 6.1: Introduction 6.2: General Considerations - read 6.3: Single stage phase I designs - read 6.4: Two stage phase.
Biostatistics in Practice Peter D. Christenson Biostatistician LABioMed.org /Biostat Session 4: Study Size and Power.
Biostatistics in Practice Peter D. Christenson Biostatistician Session 4: Study Size and Power.
Efficient Designs for Phase II and Phase III Trials Jim Paul CRUK Clinical Trials Unit Glasgow.
Department Author Bayesian Sample Size Determination in the Real World John Stevens AstraZeneca R&D Charnwood Tony O’Hagan University of Sheffield.
Biostatistics in Practice Peter D. Christenson Biostatistician Session 4: Study Size for Precision or Power.
Bayesian Approach For Clinical Trials Mark Chang, Ph.D. Executive Director Biostatistics and Data management AMAG Pharmaceuticals Inc.
: An alternative representation of level of significance. - normal distribution applies. - α level of significance (e.g. 5% in two tails) determines the.
Research Issues Can be both sociological and ethical Excesses in the past have made rules and standards necessary Rules are participant oriented and protective.
Copyright, 1996 © Dale Carnegie & Associates, Inc. CLINICAL TRIALS AND HUMAN SUBJECT PROTECTION: A PLAINTIFF’S PERSPECTIVE ALAN MILSTEIN SHERMAN SILVERSTEIN.
© 2004 Prentice-Hall, Inc.Chap 9-1 Basic Business Statistics (9 th Edition) Chapter 9 Fundamentals of Hypothesis Testing: One-Sample Tests.
SUMMARY Hypothesis testing. Self-engagement assesment.
Pilot and Feasibility Studies NIHR Research Design Service Sam Norton, Liz Steed, Lauren Bell.
NIHR Themed Call Prevention and treatment of obesity Writing a good application and the role of the RDS 19 th January 2016.
European Patients’ Academy on Therapeutic Innovation Ethical and practical challenges of organising clinical trials in small populations.
Course: Research in Biomedicine and Health III Seminar 5: Critical assessment of evidence.
EVIDENCE-BASED MEDICINE AND PHARMACY 1. Evidence-based medicine 2. Evidence-based pharmacy.
Introduction to Biostatistics, Harvard Extension School, Fall, 2005 © Scott Evans, Ph.D.1 Sample Size and Power Considerations.
Critical Appraisal II Prepared by Dr. Hoda Abd El Azim.
Is a Clinical Trial Right for Me?
Cost effectiveness Analysis: Valuing Health; Valuing Research!
Patient Focused Drug Development An FDA Perspective
Within Trial Decisions: Unblinding and Termination
Statistics for Managers using Excel 3rd Edition
Is a Clinical Trial Right for Me?
Aiying Chen, Scott Patterson, Fabrice Bailleux and Ehab Bassily
Issues in Hypothesis Testing in the Context of Extrapolation
The DMC’s role in monitoring futility
Data Monitoring committees and adaptive decision-making
How to apply successfully to the NIHR HTA Board?
Presentation transcript:

1Carl-Fredrik Burman, 11 Nov 2008 RSS / MRC / NIHR HTA Futility Meeting Futility stopping Carl-Fredrik Burman, PhD Statistical Science Director AstraZeneca R&D

Carl-Fredrik Burman, 11 Nov 2008 RSS / MRC / NIHR HTA Futility Meeting 2 Stakeholder perspectives  The patient  A pharmaceutical company  The public (MRC, NIHR)

Carl-Fredrik Burman, 11 Nov 2008 RSS / MRC / NIHR HTA Futility Meeting 3 The fundamental design requirement: Ethics ”My old mother – principle” The trial is ethical if (and only if) I would recommend my mother to take part in the trial, given that she would be eligible.

Carl-Fredrik Burman, 11 Nov 2008 RSS / MRC / NIHR HTA Futility Meeting 4 Interim stopping  Stop the trial as soon as I would not include my mother, e.g. if  One (publicly available) treatment is clearly better  A “new” treatment fails to show sufficient effect, when it has known safety disadvantages  No ethical obligation to stop  If two treatments with similar safety have no clear difference in effect

Carl-Fredrik Burman, 11 Nov 2008 RSS / MRC / NIHR HTA Futility Meeting 5 (Genuine) informed consent  The patient should get  Full information regarding the trial treatments (and procedures), including previous data, potential risks, etc.  Help to understand the information and  Apply it to his/her specific situation (health status, preferences)  When would a fully informed, fully competent patient give consent?  If and only if it is better (not worse) for him/her to take part in the trial, as compared to receiving standard therapy.  Cf. “my old mother” principle

Carl-Fredrik Burman, 11 Nov 2008 RSS / MRC / NIHR HTA Futility Meeting 6 Easy-going clinical equipose is not enough  Clinical equipose  If there is uncertainty about which treatment is better  (Alternatively, compelling evidence of one treatment being better)  (Alternatively, medical experts disagree)  It’s far too easy to say that we are uncertain  I expect my doctor to say what he believes is best

Carl-Fredrik Burman, 11 Nov 2008 RSS / MRC / NIHR HTA Futility Meeting 7 Our old Mother  Scientific equipose  Not every expert agree on CO 2 -induced global warming  Do you suggest a randomised N-of-1 trial?  Of course not — choose the treatment we believe is best Earth

Carl-Fredrik Burman, 11 Nov 2008 RSS / MRC / NIHR HTA Futility Meeting 8 What is ”best” for the patient? May depend on e.g.  Effect (best guess + uncertainty)  Safety  Better care in the trial?  Economic compensation (but beware of exploitation)  Altruism Likely effect will differ between individuals (covariates) Preferences are different Decision theory may help decide (at least in theory …)

Carl-Fredrik Burman, 11 Nov 2008 RSS / MRC / NIHR HTA Futility Meeting 9 Decision analysis (DA) Patient perspective  Utility function U(effect, safety, QoL, cost, …)  Model for effect, safety, etc., based on best information (data, expert knowledge, …). Often Bayesian prior.  Choose decision (volunteer to participate in trial, or not) to maximise expected utility The DA approach can also be used by a trial sponsor

Carl-Fredrik Burman, 11 Nov 2008 RSS / MRC / NIHR HTA Futility Meeting 10 A pharmaceutical company perspective (simplified)  A new drug will be licensed if and only if the (next) phase III trial has a statistically significant effect (p<5%)  If licensed, the company will make a profit of V (unit: £)  The trial cost is k·N, where N is the sample size  The assumed (believed) treatment effect is .  Maximise V · Power(N) – k · N Of course, this model is wrong (as all models are). Should e.g. have V=V(T)=V(T(N)), where T is time.

Carl-Fredrik Burman, 11 Nov 2008 RSS / MRC / NIHR HTA Futility Meeting 11 Gain Net gain = Gain – Cost Cost Optimal sample size N opt = 1010

Carl-Fredrik Burman, 11 Nov 2008 RSS / MRC / NIHR HTA Futility Meeting 12 The interim decision (continue vs. stop for futility)  Value V if significant  Conditional power CP if trial is continued  C additional trial cost if continued (compared to if stopped)  Continue iffV · CP > C, that is, iff CP > C / V

Carl-Fredrik Burman, 11 Nov 2008 RSS / MRC / NIHR HTA Futility Meeting 13 DA vs. ”least clinically relevant” effect  DA approach:  Maximise expected utility based on ”best guess” effect (or prior)  Traditional approach:  90% power at ”least clinically relevant” effect  What is the least clinically relevant effect?  If no adverse effects, no cost  And the outcome is death  One single saved life is clinically relevant … at least to the one saved  What is a relevant effect depends on safety, cost etc.

Carl-Fredrik Burman, 11 Nov 2008 RSS / MRC / NIHR HTA Futility Meeting 14 Conditional power at interim  Final estimate  is N( , 1/N).  Stage i has sample size N i and estimate  i. Then  = (N 1 ·  1 +N 2 ·  2 )/N  Statistical significance if  > C /  N (where C=1.96 say)  CP = P(  > C /  N ) =  (  ·  N 2 +  1 ·N 1 /  N 2 -C  (N/N 2 ) )  But which  to use when calculating CP?  Original alternative  Alternative ?  Interim estimate  1 ?  Linear combination of  1 and  Alternative ?  Bayesian posterior based on interim data ?

Carl-Fredrik Burman, 11 Nov 2008 RSS / MRC / NIHR HTA Futility Meeting 15 Stop, continue, or something else?  Run a new trial?  Sample size reestimation, based on interim estimates  Flexible design methodology (Bauer & Köhne –94)  Predefined weights for the different stages (generally, weight not proportional to information)  May change the sample size for stage 2 after viewing interim results  Discussion on CP  Somewhat controversial  May be better than design with only futility stopping  Group-sequential designs should often be preferred

Carl-Fredrik Burman, 11 Nov 2008 RSS / MRC / NIHR HTA Futility Meeting 16 Publicly funded trial: Treatments with similar safety  Assume  Whole patient population will receive one of these treatments  Efficacy is the only unknown  Same safety, cost, etc.  The closer the interim effect is to zero, the more value in continuing  Thus, no reason to stop for futility

Carl-Fredrik Burman, 11 Nov 2008 RSS / MRC / NIHR HTA Futility Meeting 17 Example 1: Value of information  Compare 2 treatments with probabilities p A, p B for death.  Assume total future population size is T (10,000 say)  If we knew that , we would choose treatment A  T·  lives would be spared as compared to using B  Similarily, choose B if  <0  Net value T·Abs(  ) or T·Abs(  )/2 if compared to using random treatment

Carl-Fredrik Burman, 11 Nov 2008 RSS / MRC / NIHR HTA Futility Meeting 18 Maximal value of information  Before trial,  p 2  p 1 has approximately normal prior with mean=0, SD=  (say 10%)  What would the value be if we could learn the exact value of  ?  Take the Bayesian expectation of the value T·Abs(  )/2, E prior [T·Abs(  )/2] = T·  /  (2  )  With T=10,000 and  =10%, about 400 lives would be spared Example cont’d

Carl-Fredrik Burman, 11 Nov 2008 RSS / MRC / NIHR HTA Futility Meeting 19

Carl-Fredrik Burman, 11 Nov 2008 RSS / MRC / NIHR HTA Futility Meeting 20 Publicly funded trial: Intervention vs. no treatment (placebo)  Assume  Intervention is associated with some cost, safety risks  Not clear whether intervention has a positive effect  If effect, then the size of the effect will determine the size of the patient population which will get a positive net benefit  First objective: is there any effect?  Reasonable to stop for futility if interim estimate is low  Expected value by continuing study is then small

Carl-Fredrik Burman, 11 Nov 2008 RSS / MRC / NIHR HTA Futility Meeting 21 Information leakage  In regulatory setting, large discussion on who should see interim data  Does the DMC have to be independent from the sponsor  What are the risks of potential information leakage?  Problems may be over-emphasised?  The ethical aspect

Carl-Fredrik Burman, 11 Nov 2008 RSS / MRC / NIHR HTA Futility Meeting 22 Summary  Futility stopping may be an ethical requirement  Industry funded trials: Tradeoff cost and expected value  Publicly funded trials (examples)  Don’t stop for futility if two active treatments differ only in effect  May stop for futility if “active” treatment unlikely to have sufficient effect (tradeoff cost and value)  (If basic science objective …)