survConcordance: Compute a concordance measure.

Description

This function computes the concordance between a right-censored survival time and a single continuous covariate

Usage

survConcordance(formula, data, weights, subset, na.action)
survConcordance.fit(y, x, strata, weight)

Arguments

formula

a formula with a survival time on the left and a single covariate on the right, along with an optional strata() term. The left hand term can also be a numeric vector.

data

a data frame

weights,subset,na.action

as for coxph

x, y, strata, weight

predictor, response, strata, and weight vectors for the direct call

Value

an object containing the concordance, followed by the number of pairs that agree, disagree, are tied, and are not comparable.

Details

The survConcordance.fit function computes the result but does no data checking whatsoever. It is intended as a hook for other packages that wish to compute concordance, and the data has already been assembled and verified.

Concordance is defined as Pr(agreement) for any two randomly chosen observations, where in this case agreement means that the observation with the shorter survival time of the two also has the larger risk score. The predictor (or risk score) will often be the result of a Cox model or other regression.

For continuous covariates concordance is equivalent to Kendall's tau, and for logistic regression is is equivalent to the area under the ROC curve. A value of 1 signifies perfect agreement, .6-.7 is a common result for survival data, .5 is an agreement that is no better than chance, and .3-.4 is the performance of some stock market analysts.

The computation involves all n(n-1)/2 pairs of data points in the sample. For survival data, however, some of the pairs are incomparable. For instance a pair of times (5+, 8), the first being a censored value. We do not know whether the first survival time is greater than or less than the second. Among observations that are comparable, pairs may also be tied on survival time (but only if both are uncensored) or on the predictor. The final concordance is (agree + tied/2)/(agree + disagree + tied).

There is, unfortunately, one aspect of the formula above that is unclear. Should the count of ties include observations that are tied on survival time y, tied on the predictor x, or both? By default the concordance only counts ties in x, treating tied survival times as incomparable; this agrees with the AUC calculation used in logistic regression. The Goodman-Kruskal Gamma statistic is (agree-disagree)/(agree + disagree), ignoring ties. It ranges from -1 to +1 similar to a correlation coefficient. Kendall's tau uses ties of both types. All of the components are returned in the result, however, so people can compute other combinations if interested. (If two observations have the same survival and the same x, they are counted in the tied survival time category).

The algorithm is based on a balanced binary tree, which allows the computation to be done in O(n log n) time.

Examples

Run this code

# NOT RUN {
survConcordance(Surv(time, status) ~age, data=lung)

options(na.action=na.exclude)
fit <- coxph(Surv(time, status) ~ ph.ecog + age + sex, lung)
survConcordance(Surv(time, status) ~predict(fit), lung)
# }
# NOT RUN {
  n=227 (1 observations deleted due to missing values)
Concordance= 0.6371102 , Gamma= 0.2759638 
concordant discordant  tied risk  tied time 
     12544       7117        126         28 
# }