get_oc_QuasiBOIN {UnifiedDoseFinding}

Generate operating characteristics for finding the maximum tolerated dose (MTD) defined by Equivalent Score (ET) using Quasi-CRM design using gBOIN

Description

Obtain the operating characteristics of Quasi-CRM design (Yuan et al. 2007) and Robust-Quasi-CRM design (Pan et al. 2014) for finding the maximum tolerated dose (MTD) using Equivalent Score (ET) derived from toxicity grade information using the gBOIN design (Mu et al. 2017)

Usage

get_oc_QuasiBOIN(target, p.true, score, ncohort, cohortsize, n.earlystop = 100,
                 startdose = 1, p.saf = 0.6 * target, p.tox = 1.4 * target,
                 cutoff.eli = 0.95, extrasafe = FALSE, offset = 0.05,
                 ntrial = 1000, seed = 100)

Arguments

target

the target DLT rate

p.true

the true toxicity probability at each dose level

score

the default value is score = seq(0, 1.5, by = 0.5) / 1.5

ncohort

the number of cohorts

cohortsize

the cohort size

n.earlystop

the early stopping parameter. The default value is n.earlystop = 100

startdose

the starting dose level. The default value is startdose = 1

p.saf

lower bound. The default value is p.saf = 0.6 * target

p.tox

upper bound. The default value is p.tox = 1.4 * target

cutoff.eli

the cutoff to eliminate an overly toxic dose for safety. The default value is cutoff.eli = 0.95

extrasafe

extrasafe set extrasafe = TRUE to impose a more stringent stopping rule. The default value is extrasafe = FALSE

offset

when extrasafe = TRUE will have effect. The default value is offset = 0.05

ntrial

the number of simulated trials

seed

the seed. The default value is seed = 100

Value

get_oc_QuasiBOIN() returns the operating characteristics of Bayesian optimal interval design as a list object, including: (1) the target DLT rate, (2) the true DLT rate at different scale for each dose level, (3) number of cohort, (4) cohortsize, (5) starting dose level, (6) lower bound, (7) upper bound, (8) selection percentage of each dose level, (9) the average number of patients treated at each dose, (10) the average number of patients responded to toxicity at each dose level

Author(s)

Chia-Wei Hsu, Haitao Pan, Rongji Mu

References

Yuan, Z., R. Chappell, and H. Bailey. "The continual reassessment method for multiple toxicity grades: a Bayesian quasi-likelihood approach." Biometrics 63, no. 1 (2007): 173-179.

Pan, Haitao, Cailin Zhu, Feng Zhang, Ying Yuan, Shemin Zhang, Wenhong Zhang, Chanjuan Li, Ling Wang, and Jielai Xia. "The continual reassessment method for multiple toxicity grades: a Bayesian model selection approach." PloS one 9, no. 5 (2014): e98147.

Mu, Rongji, Ying Yuan, Jin Xu, Sumithra J. Mandrekar, and Jun Yin. "gBOIN: a unified model-assisted phase I trial design accounting for toxicity grades, and binary or continuous end points." Journal of the Royal Statistical Society. Series C: Applied Statistics 68, no. 2 (2019): 289-308.

Examples

target <- 0.47 / 1.5
p.true <- matrix(c(0.83, 0.12, 0.04, 0.01,
                   0.75, 0.15, 0.07, 0.03,
                   0.62, 0.18, 0.11, 0.09,
                   0.51, 0.19, 0.14, 0.16,
                   0.34, 0.16, 0.15, 0.35,
                   0.19, 0.11, 0.11, 0.59), ncol = 4, byrow = TRUE)
score <- seq(0, 1.5,by = 0.5) / 1.5
ncohort <- 10
cohortsize <- 3
ntrial <- 4000
get_oc_QuasiBOIN(target = target, p.true = p.true, score = score, ncohort = ncohort,
                 cohortsize = cohortsize, ntrial = ntrial)