surv.boot

This function calculates the standard error for the estimated potential cumulative incidence function
and treatment effect. Two methods to calculate the standard error are considered: the asymptotic standard error
based on the explicit formula and bootstrapping.

Analysis of treatment effects in clinical trials with time-to-event outcomes is complicated by intercurrent events. This package implements methods for estimating and inferring the cumulative incidence functions for time-to-event (TTE) outcomes with intercurrent events (ICE) under the five strategies outlined in the ICH E9 (R1) addendum, see Deng (2025) <doi:10.1002/sim.70091>. This package can be used for analyzing data from both randomized controlled trials and observational studies. In general, the data involve a primary outcome event and, potentially, an intercurrent event. Two data structures are allowed: competing risks, where only the time to the first event is recorded, and semicompeting risks, where the times to both the primary outcome event and intercurrent event (or censoring) are recorded. For estimation methods, users can choose nonparametric estimation (which does not use covariates) and semiparametrically efficient estimation.

Yi Zhou

tteICE

Treatment Effect Estimation for Time-to-Event Data with
Intercurrent Events

Yuhao Deng

surv.boot function

<dl><dt>fit</dt>
<dd>A fitted object returned by the function <code>tteICE</code>, <code>surv.tteICE</code>, or <code>scr.tteICE</code>.</dd>
<dt>nboot</dt>
<dd>Number of resamplings in the boostrapping method. If <code>nboot</code> is 0 or 1, then
asymptotic standard error based on the explicit form is calculated instead of bootstrapping.</dd>
<dt>seed</dt>
<dd>Seed for bootstrapping.</dd></dl>

Arguments

Calculate standard errors for estimated CIFs and treatment effects — surv.boot

<dl>

<dt>fit</dt>
<dd>A fitted object returned by the function <code>tteICE</code>, <code>surv.tteICE</code>, or <code>scr.tteICE</code>.</dd>


<dt>nboot</dt>
<dd>Number of resamplings in the boostrapping method. If <code>nboot</code> is 0 or 1, then
asymptotic standard error based on the explicit form is calculated instead of bootstrapping.</dd>


<dt>seed</dt>
<dd>Seed for bootstrapping.</dd>

</dl>

surv.boot: Calculate standard errors for estimated CIFs and treatment effects

Description

Usage

Value

Arguments

See Also