Restricted mean survival time over a fixed horizon

RMST answers what a hazard ratio does not: how much longer, on average, over a stated window, and it needs no proportional-hazards assumption. The horizon is not optional in substance, because past the last observed time the curve is a plateau the data cannot pin down, so t_star is required here rather than defaulted.

Usage

Survrls(fit, event, t_star)

Arguments

fit: Observed event or censoring times.
event: Event indicator, 1 = event, 0 = censored.
t_star: Restriction horizon, positive.

Value

List with estimate, tau, n_events, method.

Details

A thin alias for Rmst; the estimator already exists and is not duplicated here.

Formula: RMST(t*) = integral_0^{t*} S(u) du, the survival curve being a step function integrated as summed rectangles.

References

Royston, P. and Parmar, M. K. B. (2013). Restricted mean survival time. BMC Medical Research Methodology 13:152. doi:10.1186/1471-2288-13-152 Klein, J. P. and Moeschberger, M. L. (2003). Survival Analysis, 2nd ed. Springer, section 4.5.

Examples

Survrls(c(1, 2, 3), c(1, 1, 1), 3)
#> $estimate
#> [1] 2
#> 
#> $tau
#> [1] 3
#> 
#> $n_events
#> [1] 3
#> 
#> $method
#> [1] "RMST(t*) = area under the KM curve [Royston & Parmar 2013]"
#>