Estimate the hazard ratio and compute a confidence interval for it via a special application of the MantelHaenszel procedure. A separate 2 x 2 table is constructed for each event time. The underlying assumption is that the hazard ratio is constant over the followup period.
1  mantelCoxHR(time, event, group, conf.level = 0.95)

time 
Event times, censored or observed. 
event 
Censoring indicator, 1 for event, 0 for censored. 
group 
Factor with two levels, e.g. treatment group. 
conf.level 
Significance level for confidence interval for hazard ratio. 
mantelCox.hr 
Hazard ratio estimate. 
ci.hr 
Wald confidence interval at the level specified by 
p.val.logrank 
pvalue of logrank test for a comparison of the survival curves between the two groups. 
coxph.hr 
Hazard ratio estimated via Coxregression. 
Note that in general the MantelCox estimate and the hazard ratio estimate received via Coxregression do not coincide.
Kaspar Rufibach
kaspar.rufibach@gmail.com
Kirkwood, B.R. and Sterne, J.A.C. (2003). Essential Medical Statistics. Blackwell Science. See p. 283 ff.
1 2 3  ## use Acute Myelogenous Leukemia survival data contained in package 'survival'
time < leukemia[, 1]; status < leukemia[, 2]; x < as.factor(leukemia[, 3])
mantelCoxHR(time, status, x)

Questions? Problems? Suggestions? Tweet to @rdrrHQ or email at ian@mutexlabs.com.
Please suggest features or report bugs with the GitHub issue tracker.
All documentation is copyright its authors; we didn't write any of that.