1 / 16

Outline

Outline. Model fitting in WinBUGS Choosing next dose Pre-trial simulations. Model fitting: NDLM. Observation equation response Y i is neurological score at 13 weeks b i is baseline neurological score subject i at dose Z j.

london
Download Presentation

Outline

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript


  1. Outline • Model fitting in WinBUGS • Choosing next dose • Pre-trial simulations

  2. Model fitting: NDLM • Observation equation • response Yi is neurological score at 13 weeks • biis baseline neurological score • subject i at dose Zj • Yi - bi = j+ ij = 1,…,J; i = 1,…,Ii ~ i.i.d. N(0, 2)

  3. sampling precision samplingdistribution θ is changefrom baseline specify prior vague, half-Normal prior on σ Observation equation: WinBUGS model{ for (i in 1:I){ Y[i] ~ dnorm(mu[i], sigma2inv) mu[i] <- baseline[i] + theta[d[i]] } sigma2inv <- 1 / (sigma * sigma) sigma ~ dnorm(0,0.1)I(0,) }

  4. Locally around z = Zj a straight line with level qj and slope dj Parameters (qj , dj ) change between doses by adding a (small) evolution noise NDLM Evolution Variance = Smoother

  5. Model fitting: NDLM • Evolution (system) equation where ωj and ej ~ i.i.d. N(0, Wj 2)

  6. vague prior on placebo leveland slope θ dependson previous θ and δ randomwalk evolution variance of θ, δ is W * σ2 uniform prior on W, fraction of sampling variance Evolution equation:WinBUGS theta[1] ~ dnorm(mu.theta0, prec.theta0) delta[1] ~ dnorm(mu.delta0, prec.delta0) for(j in 2:J){ theta[j] ~ dnorm(mu.theta[j], prec.theta[j]) mu.theta[j] <- theta[j-1] + delta[j-1] delta[j] ~ dnorm(delta[j-1], prec.delta[j]) prec.theta[j] <- 1 / (W * sigma * sigma) prec.delta[j] <- 1 / (W * sigma * sigma) } W ~ dunif(0.001,1)

  7. Choosing next dose • Select utility function • -V(response at ED95) • -V(ED95) • -det(VCOV(ED95, response at ED95) • Randomisation rule • placebo or optimal dose • probability proportional to utility of each dose • placebo or doses at or ‘near’ optimal utility

  8. Choosing next dose • Estimating utility of each dose • full MCMC estimation of utility posterior predictive distribution • simpler estimation of expected utility • predict an observed response at each dose • calculate ED95 expected value by importance sampling • hence for each dose get utility -V(ED95)

  9. Estimating utility

  10. Pre-trial simulations • During actual trial, efficient computing less important • Critical for pre-trial simulations • underlying dose response curve • settings of longitudinal model • choice of covariates • utility function • randomisation rule • compare to ‘standard’ designs

  11. Pre-trial simulations • call WinBUGS using x command options noxwait xmin; x cd &bugsdir; x winbugs14.exe /PAR &scriptname;

  12. Construct text filesfor analysis, WinBUGS script,WinBUGS data set SAS WinBUGS Run in WinBUGS Import MCMC samples,predicted observation Trial stoppingrule triggered? Stop trial,call report Y Randomiseanotherpatient N Estimate utility Simulating a trial

  13. proportionallocatedeach dose utility dose-responsecurve estimate ED95posterior Pre-trial simulations

  14. Summary • Adaptive design (NDLM) straightforward in WinBUGS • Generic software simplifies implementation and validation • Interaction with SAS permits wide scope of pre-trial simulations • …and ease of integration with in-house reporting systems in industry

  15. Acknowledgements • UK Medical Research Council • Pfizer Global Research & Development:Andy Grieve, Margaret Jones, Mike Smith, Mike Krams • Tessella: Tom Parke • Duke University: Peter Müller, Don Berry

More Related