ipd: IPD

In raredd/kmdata: A Database of Reconstructed Individual Patient Level Data from Oncology Clinical Trials

IPD

Description

Generate individual patient data (IPD) from digitized survival curve data using the method described by Guyot (2012).

Usage

ipd(
  time,
  prob,
  t.atrisk,
  n.atrisk,
  total = NA,
  arm = "arm",
  tau = max(time, t.atrisk),
  interpolate = FALSE
)

Arguments

time, prob	vectors of survival time and probabilities
t.atrisk, n.atrisk	vectors of times and number at-risk
total	(optional) total number of events
arm	(optional) arm label
tau	maximum follow-up time
interpolate	logical or integer; if TRUE or an integer, constant interpolation will be performed for the time-probability vectors; this may be helpful in situations with a small number of digitized points or if the censoring distribution is difficult to estimate

References

Guyot, Patricia, et al. Enhanced secondary analysis of survival data: reconstructing the data from published Kaplan-Meier survival curves. BMC Medical Research Methodology 2012, 12:9.

Examples

## Not run: 
xy <- system.file(
  'etc', 'init', 'Checkmate_067_S3A_Nivolumab.csv',
  package = 'kmdata'
)
ar <- system.file(
  'etc', 'init', 'Checkmate_067_At_Risk.csv',
  package = 'kmdata'
)
dd <- read.csv(xy)
ar <- read.csv(ar)

## generate the patient-level censored data
ipd1 <- ipd(dd$T, dd$S, ar$trisk, ar$nrisk, arm = 'Nivo')
with(ipd1, tapply(time, event, summary))

## convenience function to plot
kmplot(ipd1, xaxis.at = ar$trisk)

## overlay raw coordinates from digitization
lines(dd$T, dd$S, col = 2)

## End(Not run)

raredd/kmdata documentation built on June 15, 2025, 9:33 a.m.