PCDSpline (version 1.0)

BladTumor: Bladder Tumor Cancer Data

Description

Bladder tumor data were from the bladder cancer study conducted by the Veterans Administration Cooperative Urological Research Group, and used by many people to demonstrate methodology for recurrent event modelling. In this study 118 patients who had superficial bladder tumors were randomized into one of three treatment groups: placebo (48), thiotepa (38), and pyridoxine (32). During the study at each follow-up visit, new tumors since the last visit were counted, measured, and removed transurethrally. For each patient the initial number of tumors and the size of largest initial tumors were also recorded. For more details about this study see Byar et al. (1977).

Usage

data(BladTumor)

Format

A data frame with 116 observations on the following 8 variables.

subject:

patient ID

time:

observation time

count:

cumulative number of tumors

number:

initial number of tumors (8=8 or more)

size:

size(cm) of largest initial tumors

pyridoxine:

dummy variable for pyridoxine treatment

thiotepa:

dummy variable for thiotepa treatment

count1:

number of new tumors since last observation time

Source

Wang, X. and Yan, J. (2011). Fitting semiparametric regressions for panel count survival data with an R package spef. Computer Methods and Programs in Biomedicine 104,2 278-285

Details

This data include 116 subjects who have at least one follow-up observation after the study enrollment.

References

Byar, D.P., Blackard,C., and the VACURG. (1977). Comparisons of placebo, pyridoxine, and topical thiotepa in preventing recurrence of stage I bladder cancer. Urology 10, 556-561.

See Also

BladTumor1

Examples

data(BladTumor)

n<-max(BladTumor$subject)
#record the number of observations for all patients
k<-as.numeric(table(BladTumor$subject))   
K<-max(k)
t<-matrix(,n,K)
z<-matrix(,n,K)

x1<-c();x2<-c();x3<-c();x4<-c();

for (r in 1:n){
rownum<-which(BladTumor$subject==r)
#record all observation times
t[r,][1:k[r]]<-BladTumor[rownum,]$time    
#record all panel counts from non-overlapping intervals
z[r,][1:k[r]]<-BladTumor[rownum,]$count1  
x1[r]<-BladTumor[which(BladTumor$subject==r),]$number[1]
x2[r]<-BladTumor[which(BladTumor$subject==r),]$size[1]
x3[r]<-BladTumor[which(BladTumor$subject==r),]$pyridoxine[1];
x4[r]<-BladTumor[which(BladTumor$subject==r),]$thiotepa[1]
}

x<-cbind(x1,x2,x3,x4)
BladTumor1<-list(t=t,x=x,z=z,k=k,K=K)