SROCt


< Previous

R is a free software environment for statistical computing and graphics. It compiles and runs on a wide variety of UNIX platforms, Windows and MacOS. To download R, please click on this link and choose your preferred CRAN mirror.

We propose the SROCt package. This is a collection of simple R functions that were used for computing the time-dependant ROC curve for a prognostic marker from aggregated data (survival probabilities in strata of the marker) and from several studies. More precisely, the cut-offs defining the strata of the marker are allowed to be different throughout the studies and need to be specified. The marker is assumed to be normally distributed. The survival function given by the marker level is modeled with a piecewise-constant hazard function. The association of the marker is allowed to be different in each interval of time. Random effects are introduced in the models to take the inter-study variability into account. The regression coefficients are assessed by using the R package nlme. Illustrative data are provided: data from the cohort DIVAT, and data from a published meta-analysis (de Azambuja et al. 2007).

Download the SROCt package and save the resulting file in your computer


Install the SROCt package

  • Shut down R if it is running.
  • Using a command-line window navigate (using the "cd" command) to the directory where you saved the file and issue the command: sudo R CMD INSTALL sroct_0.x.tar.gz.
  • Shut down R if it is running.
  • Restart R from the Start Menu.
  • From the Packages menu, choose the Install package(s) from local zip files.
  • Navigate to and choose the file you just downloaded: sroct_0.x.zip.
If you find any bugs in SROCt package, please e-mail us and we will take your comments into account in the next version.

For non R-users, you can directly download the following files

The aggregated dataset from the KI-67 meta-analysis (xls file). Available aggregated data are exhaustively presented. It concerns the meta-analysis already published by Azambuja et al., aiming at evaluating if KI-67 can be considered as a good prognostic marker for breast cancer survival. The KI-67 is assumed log-normal distributed. This is a table with 406 observations (rows) and with the 10 following variables (columns):
  • "classe" represents the groups of patients defined using KI-67. 1 is the first group which is defined by the lowest KI-67 values.
  • "n" represents the number of recipients at the baseline in each group (date of KI-67 collection).
  • "year" represents the survival time (in years).
  • "surv" represents the survival probabilities at each year.
  • "nrisk" represents the number of subjects at-risk of the event at the corresponding year.
  • "proba" represents the proportion of patients for a given paper who belong to the corresponding group (obtained by using the Kaplan and Meier estimator from the published papers).
  • "log.marker.min" represents the logarithm of the minimum value of the KI-67 interval.
  • "log.marker.max" represents the logarithm of the maximum value of the KI-67 interval.
  • "study.num" This numeric vector identifies the studies.
  • "author" vector identifies the first author of the paper.
  • "year.paper" identifies the year of publication.


The aggregated dataset from the DIVAT cohort(xls file).We have considered a subpopulation of 4195 adult patients who have received a first kidney graft between January 1996 and Jun 2008 and with a measure of the 1-year creatinine. Five centers have participated. A total of 511 graft failures have been observed (346 returns to dialysis and 165 patient deaths with a functional kidney). From this database, we have constructed an aggregated dataset to perform a meta-analysis on 5 published monocentric studies. This is a table with 106 observations (rows) with the 8 following variables (columns):
  • "classe" represents the groups of recipients defined using the 1-year serum creatinine. 1 is the first group with the lowest values of 1-year serum creatinine.
  • "n" represents the number of recipients at the baseline (date of the transplantation) in each group.
  • "year" represents the post transplant time (in years).
  • "surv" represents the survival probabilities at each year (obtained using the Kaplan and Meier estimator from the individual data).
  • "nrisk" represents the number of subjects at-risk of the event at the corresponding year.
  • "proba" represents the proportion of the patients in a center which belong to the corresponding group.
  • "marker.min" represents the minimum value of the interval of the 1-year serum creatinine (in µmol/l).
  • "marker.max" represents the maximum value of the interval of the 1-year serum creatinine (in µmol/l).
  • "centre.num" represents the centers.


You can find here the corresponding patient and graft survival curves per centers from the DIVAT network and according to the 1-year serum creatinine level (pdf file).