twostg3: Compute operating characteristics of two-stage...

View source: R/twostg3.R

twostg3R Documentation

Compute operating characteristics of two-stage trinomial-outcome designs

Description

For a phase II study with three possible ordered outcomes (e.g., response, stable, progression), computes the probability of not rejecting the null hypothesis for a specified two-stage design.

Usage

twostg3(n1, n2, presp, pstab, r1, s1, r2, s2)

Arguments

n1

Number of subjects enrolled in the first stage

n2

Number of additional subjects enrolled in the second stage

presp

The probability of the best (response) category

pstab

The probability of the intermediate (stable only) category

r1

Max number of responses that can occur at the first stage without proceeding to the second stage

s1

Max number cases that can have responses or stable outcomes at first stage without proceeding to the second stage

r2

Max number of responses from stages 1 and 2 combined that can be observed without declaring the treatment to be effective

s2

Max number of cases from stages 1 and 2 combined that can have responses or stable outcomes without declaring the treatment to be effective

Details

Consider a phase II study with an ordered categorical outcome with three categories, which are referred to here as response, stable and progression (based on the best response outcome in oncology). A treatment might improve outcome by either improving the response rate or by improving the disease stabilization rate (where stabilization includes best response of response or stable). In the designs considered here, the treatment is declared to be active (or sufficiently active to be worth studying further) if it shows a sufficient level of activity for either response or disease stabilization.

More formally, let R1 be the number of patients observed to respond and S1 be the number of patients observed to be stable (but not responses) during the first stage of accrual, and similarly let R2 and S2 be the number of additional cases observed to respond or stabilize from the second stage. The study is stopped after the first stage if BOTH R1 <= r1 and R1+S1 <= s1. The study proceeds to the second stage if either R1 > r1 or R1+S1 > s1. The treatment is declared to be inactive if either the study stops after the first stage or if at the end of the second stage BOTH R1+R2 <= r2 and R1+R2+S1+S2 <= s2. The treatment is considered sufficiently active for further investigation if either R1+R2 > r2 or R1+R2+S1+S2 > s2.

Note that presp and pstab are the probabilities of the corresponding disjoint cells in the trinomial model (that is, pstab is the probability that a case will be stable but not a response).

Value

A vector giving the probability of stopping after the first stage (p.stop.1) and the overall probability that the treatment is declared inactive (p.inactive).

See Also

twostg

Examples

twostg3(20, 40, 0.05, 0.10, 1, 3, 5, 22)


raredd/desmon documentation built on May 7, 2024, 3:46 p.m.