Package accrual. R topics documented: October 20, Type Package Title Bayesian Accrual Prediction Version 1.3 Date

Transcription

1 ype Package itle Bayesia Accrual Predictio Versio 1.3 Date Package accrual October 20, 2017 Author Juhao Liu, Yu Jiag, Ce Wu, Steve Sio, Matthew S. Mayo, Raa Raghava, Byro J. Gajewski Maitaier Juhao Liu Depeds R(>= 3.1.3), tcltk2 Iports fgui, SMPracticals Subject recruitet for edical research is challegig. Slow patiet accrual leads to delay i research. Accrual oitorig durig the process of recruitet is critical. Researchers eed reliable tools to aage the accrual rate. We developed a Bayesia ethod that itegrates researcher's experiece o previous trials ad data fro the curret study, providig reliable predictio o accrual rate for cliical studies. I this R package, we preset fuctios for Bayesia accrual predictio which ca be easily used by statisticias ad cliical researchers. Licese GPL-2 LazyLoad yes NeedsCopilatio o Repository CRAN Date/Publicatio :34:53 UC R topics docueted: accrual-package accrual.data accrual.gui accrual.ulti accrual..hedgig accrual..ifor accrual..plot

2 2 accrual-package accrual.plot.ulticeter accrual.plots accrual..hedgig accrual..ifor accrual..plot Idex 11 accrual-package Bayesia Accrual Predictio Details : Subject recruitet for edical research is challegig. Slow patiet accrual leads to delay i research. Accrual oitorig durig the process of recruitet is critical. Researchers eed reliable tools to aage the accrual rate. We developed a Bayesia ethod that itegrates researcher s experiece o previous trials ad data fro the curret study,providig reliable predictio o accrual rate for cliical studies. I this R package, we preset fuctios for Bayesia accrual predictio which ca be easily used by statisticiaa ad cliical researchers. Package: accrual ype: Package Versio: 1.2 Date: Licese: GPL-2 here are ajor eight futios i the package. he accrual.gui fuctio provides the gui versio. Maitaier:Juhao Liu <jliu4@kuc.edu> Refereces [1] Gajewski BJ, Sio SD, Carlso SE (2008). Predictig accrual i cliical trials with Bayesia posterior predictive distributios. Stat Med. 27(13): [2] Jiag, Y., Sio, S., Mayo, M. S., & Gajewski, B. J. (2015). Modelig ad validatig Bayesia accrual odels o cliical data ad siulatios usig adaptive priors. Statistics i edicie, 34(4), accrual..ifor(=300, =36, P=0.5, =100, t=10, p=36)

3 accrual.data 3 accrual..plot(=300, =36, P=0.5, =100, t=10, p=36, Method="Iforative Prior") accrual..plot(=300, =36, P=0.5, =100, t=10, p=300, Method="Iforative Prior") accrual.gui() accrual.data Exaple Accrual Data A exaple dataset for subject accrual. accrual.data str(accrual.data) plot(accrual.data) accrual.plots(accrual.data) accrual.gui GUI Versio of the Bayesia Accrual Predicito he R GUI iterface oly eeds the researchers to iput the origial desig iforatio that are required iforatio for IRBs (total tie proposed ad total subjects proposed) ad the updated accrual data (tie sice start ad subjects accrual). It uses Bayesia predictio odel i the backgroud of calculatio. accrual.gui() accrual.gui()

4 4 accrual.ulti. accrual.ulti. Predictio of Multiceter Accrual with Iforative Prior i Fixed ie Frae Produce a output for predictio of the uber of subjects ca be recruited i a fixed tie frae with Iforative Prior for a ulticeter trial. accrual.ulti.(,,p,j,,sj,,pred,all) arget copletio tie P he prior certaity, rage 0-1 J sj pred all he uber of sites he start date for each site Saple observed to date for each site he specific tie that wat to predict the recruitet Usig all the sites (rue/false) accrual.ulti.(=300,=36,p=0.5,j=10,=10,sj=c(0,0,0,0,0,0,0,0,0,0), =c(9,10,10,10,11,11,11,12,12,12),pred=36,all=rue)[[1]]

5 accrual..hedgig 5 accrual..hedgig Predictio of Accrual with Hedgig Prior i Fixed ie Frae Produce a output for predictio of the uber of subjects ca be recruited i a fixed tie frae with Hedgig Prior. accrual..hedgig(,,, t, p) t p arget copletio tie Saple observed to date he specific tie that wat to predict the recruitet accrual..hedgig(=300, =36, =100, t=10, p=36)[[1]] accrual..ifor Predictio of Accrual with Iforative Prior i Fixed ie Frae Produce a output for predictio of the uber of subjects ca be recruited i a fixed tie frae with Iforative Prior. accrual..ifor(,, P,, t, p)

6 6 accrual..plot arget copletio tie P he prior certaity, rage 0-1 Saple observed to date t p he specific tie that wat to predict the recruitet accrual..ifor(=300, =36, P=0.5, =100, t=10, p=36)[[1]] accrual..plot Plot for Predictio of Accrual i Fixed ie Frae Produce a plot ad output for predictio of the uber of subjects ca be recruited i a fixed tie frae. accrual..plot(,, P,, t, p, Method) P t p Method arget copletio tie he prior certaity, rage 0-1; For Accelerated Prior, P = 1-/ Saple observed to date he specific tie that wat to predict the recruitet Iforative Prior, Accelerated Prior, Hedgig Prior accrual..plot(=300, =36, P=0.5, =100, t=10, p=36, Method="Iforative Prior") accrual..plot(=300, =36, =100, t=10, p=36, Method="Accelerated Prior") accrual..plot(=300, =36, =100, t=10, p=36, Method="Hedgig Prior")

7 accrual.plot.ulticeter 7 accrual.plot.ulticeter Plot for Predictio of Multiceter Accrual i Fixed ie Frae Produce a plot ad output for predictio of the uber of subjects for a ulticeter trial ca be recruited i a fixed tie frae. accrual.plot.ulticeter(,,p,j,,sj,,all) arget copletio tie P he prior certaity, rage 0-1 J sj all he uber of sites he start date for each site Saple observed to date for each site Usig all the sites (rue/false) accrual.plot.ulticeter(=300,=36,p=0.5,j=10,=10,sj=c(0,0,0,0,0,0,0,0,0,0), =c(9,10,10,10,11,11,11,12,12,12),all=rue) accrual.plots Digostic Plots he diagostic pael shows four figures that help to uderstad the data distributio. he figure o the top left is the expoetial quatile plot, which checks whether the distributio of waitig ties is expoetial. he top right figure shows the histogra of the waitig ties, with the red lie is the theoretical expoetial distributio. he figure of waitig tie verse cuulative accrual tie is show o the botto left. he figure of total accrual verse cuulative accrual tie is show o the botto right.

8 8 accrual..hedgig accrual.plots(w) w he accrual dataset accrual.plots(accrual.data) accrual..hedgig Predictio of ie with Hedgig Prior Predictio of tie frae with Hedgig Prior for a certai uber of subjects. accrual..hedgig(,,, t, p) t p arget copletio tie Saple observed to date he specific uber of subjects wat to be predicted accrual..hedgig(=300, =36, =100, t=10, p=300)[[1]]

9 accrual..ifor 9 accrual..ifor Predictio of ie with Iforative Prior Predictio of tie frae with Iforative Prior for a certai uber of subjects. accrual..ifor(,, P,, t, p) arget copletio tie P he prior certaity, rage 0-1 Saple observed to date t p he specific uber of subjects wat to be predicted accrual..ifor(=300, =36, P=0.5, =100, t=10, p=300)[[1]] accrual..plot Plot for Predictio of ie Produce a plot ad output for predictio of tie frae for a certai uber of subjects. accrual..plot(,, P,, t, p, Method)

10 10 accrual..plot P t p Method arget copletio tie he prior certaity, rage 0-1; For Accelerated Prior, P = 1-/ Saple observed to date he specific uber of subjects wat to be predicted Iforative Prior, Accelerated Prior, Hedgig Prior accrual..plot(=300, =36, P=0.5, =100, t=10, p=300, Method="Iforative Prior") accrual..plot(=300, =36, =100, t=10, p=300, Method="Accelerated Prior") accrual..plot(=300, =36, =100, t=10, p=300, Method="Hedgig Prior")

11 Idex opic Bayesia accrual-package, 2 accrual.gui, 3 accrual.ulti., 4 accrual..hedgig, 5 accrual..ifor, 5 accrual..plot, 6 accrual.plot.ulticeter, 7 accrual..hedgig, 8 accrual..ifor, 9 accrual..plot, 9 opic Diagostic accrual.plots, 7 opic accrual accrual-package, 2 accrual.ulti., 4 accrual..hedgig, 5 accrual..ifor, 5 accrual..hedgig, 8 accrual..ifor, 9 opic datasets accrual.data, 3 opic expoetial accrual.plots, 7 opic gui accrual.gui, 3 opic plot accrual..plot, 6 accrual.plot.ulticeter, 7 accrual..plot, 9 accrual.plots, 7 accrual..hedgig, 8 accrual..ifor, 9 accrual..plot, 9 accrual (accrual-package), 2 accrual-package, 2 accrual.data, 3 accrual.gui, 3 accrual.ulti., 4 accrual..hedgig, 5 accrual..ifor, 5 accrual..plot, 6 accrual.plot.ulticeter, 7 11