blaTum: Bladder Tumors Cancer Recurrences
In jun-yan/spef: Semiparametric Estimating Functions

Description Usage Format Note Source See Also Examples

A data frame contains data on recurrences of bladder cancer, used by many people to demonstrate methodology for recurrent event modelling. This data set is adapted from http://onlinelibrary.wiley.com/journal/10.1111/(ISSN)1467-9868/homepage/62_2.htm, containing records of 85 patients from two treatment arms: 48 are from the placebo arm, and the rest 37 are from the thiotepa arm.

blaTum

id:	patient id
treatment:	placebo = 0, thiotepa = 1
size:	size (cm) of largest initial tumour
num:	initial number of tumours (8=8 or more)
time:	observation time
count:	number of new tumors since last observation time

To our surprise, the two-treatment (placebo and thiotepa) subset of the full version bladTumor do not match the two-treatment version blaTum.

Byar, D. P. (1980). The Veterans administration study of chemoprophylaxis for recurrent stage I bladder tumors: comparisons of placebo, pyridoxine, and topical thiotepa. Bladder Tumors and Other Topics in Urological Oncology, pp. 363–370. New York: Plenum.

Sun, J. and Wei, L. J. (2000) Regression analysis of panel count data with covariate dependent observation and censoring times. Journal of the Royal Statistical Society, Series B: Statistical Methodology, 62 293–302.

Huang, C. Y., Wang, M. C. and Zhang, Y. (2006). Analysing panel count data with informative observation times. Biometrika, 93(4): 763–776.

bladTumor

## Load data set
data(blaTum)

## Plot bladder tumor data
library(ggplot2)
ggplot(blaTum, aes(time, id)) + geom_tile(aes(fill=count)) +
   facet_grid(treatment ~ ., scales="free_y", )

## Fit a semiparametric regression model
formula <- PanelSurv(id, time, count) ~ num + size + treatment
fit <- panelReg(formula, data=blaTum, method="AEE", se="Sandwich")
fit