Package 'BayesianTrial'

Title: Bayesian Decision-Making for Clinical Trials
Description: Provides tools for Bayesian decision-making in single-arm phase II clinical trials with binary efficacy endpoints. Uses Beta-binomial models to calculate posterior probabilities that the response rate exceeds a prespecified reference value and to estimate predictive probabilities of meeting final-analysis success criteria. Evaluates the frequentist operating characteristics of two-stage designs with posterior-probability-based stopping rules, including type I error rates, statistical power, and probabilities of early stopping for efficacy or futility, using Monte Carlo simulation. For methodological background, see Thall and Simon (1994) <doi:10.2307/2533377> and Lee and Liu (2008) <doi:10.1177/1740774508089279>.
Authors: Jiangyan Zhao [cre, aut], Rongji Mu [aut, cph]
Maintainer: Jiangyan Zhao <[email protected]>
License: GPL (>= 3)
Version: 0.1.0
Built: 2026-11-19 13:38:36 UTC
Source: https://github.com/cran/BayesianTrial

Help Index


Operating Characteristics of a Two-Stage Single-Arm Phase II Trial

Description

Evaluates the operating characteristics of a two-stage single-arm phase II trial with a binary efficacy endpoint and prespecified posterior-probability stopping rules. Monte Carlo simulation is used to estimate the probability of trial success and early stopping probabilities for efficacy and futility.

Usage

oc_singlearm_binary(
  theta_true,
  n_current,
  N_total,
  alpha0 = 0.2,
  beta0 = 0.8,
  theta0_null = 0.2,
  gamma_early = 0.98,
  gamma_futility = 0.5,
  gamma_final = 0.95,
  ntrial = 10000,
  seed = 666
)

Arguments

theta_true

Assumed true treatment response rate for generating trial outcomes; a probability between 0 and 1. Specify the null-scenario rate to assess the type I error rate or a clinically promising alternative rate to assess power.

n_current

Planned number of patients with evaluable outcomes at the single interim analysis; a non-negative integer not exceeding N_total. A conventional two-stage design requires 0 < n_current < N_total.

N_total

Planned maximum number of evaluable patients; a non-negative integer at least as large as n_current.

alpha0, beta0

Positive, finite shape parameters of the Beta prior for the response rate. Their sum must also be finite.

theta0_null

Prespecified unacceptable or historical reference response rate used as the primary efficacy benchmark; a probability between 0 and 1.

gamma_early

Interim posterior probability cutoff for early declaration of efficacy; a probability between 0 and 1. An early positive result is declared if the posterior probability of exceeding theta0_null is strictly greater than this cutoff.

gamma_futility

Interim posterior probability cutoff for stopping due to futility; a probability between 0 and 1 and strictly less than gamma_early. Futility stopping occurs if the posterior probability of exceeding theta0_null is less than or equal to this cutoff. This is not a predictive probability cutoff.

gamma_final

Posterior probability cutoff for declaring a positive result at the final analysis if the trial continues; a probability between 0 and 1.

ntrial

Positive integer number of independently simulated trials.

seed

Optional integer seed for reproducibility.

Value

A named list containing:

power_null_post

Estimated overall probability of declaring a positive trial result, including early efficacy stopping. Interpreted as type I error probability under a null scenario or as power under a specified alternative.

early_stop_rate_futility

Estimated probability of stopping the trial at the interim analysis for futility.

early_stop_rate_efficacy

Estimated probability of stopping the trial at the interim analysis for efficacy.

References

Zhao, J., and Mu, R. (2026). Introduction to Bayesian Decision Method and Its Application in Medical Research. Manuscript in preparation (in Chinese).

Thall, P. F., and Simon, R. M. (1994). Practical Bayesian guidelines for phase IIB clinical trials. Biometrics, 50, 337–349. doi:10.2307/2533377.

Lee, J. J., and Liu, D. D. (2008). A predictive probability design for phase II cancer clinical trials. Clinical Trials, 5, 93–106. doi:10.1177/1740774508089279.

See Also

posterior_predictive_singlearm_binary

Examples

null_oc <- oc_singlearm_binary(
  theta_true = 0.20, n_current = 20, N_total = 40
)
alternative_oc <- oc_singlearm_binary(
  theta_true = 0.40, n_current = 20, N_total = 40
)
null_oc$power_null_post  # Type I error probability at the null rate
alternative_oc$power_null_post  # Power at the alternative rate

Posterior and Predictive Probabilities for a Single-Arm Phase II Trial

Description

Calculates posterior summaries of the treatment response rate and predictive probabilities of meeting prespecified final-analysis efficacy criteria in single-arm phase II trials with binary endpoints.

Usage

posterior_predictive_singlearm_binary(
  x_obs,
  n_current,
  N_total,
  alpha0 = 0.2,
  beta0 = 0.8,
  theta0_null = 0.2,
  theta0_clin = 0.4,
  gamma_final = 0.95,
  gamma_clin_final = 0.9,
  S = 10000,
  type = c("posterior", "predictive", "both"),
  seed = 666
)

Arguments

x_obs

Number of observed responses among n_current patients; a non-negative integer no greater than n_current.

n_current

Number of patients with evaluable outcomes at the current analysis; a non-negative integer no greater than N_total.

N_total

Planned maximum number of patients with evaluable outcomes; a non-negative integer at least as large as n_current.

alpha0, beta0

Positive shape parameters of the Beta prior for the response rate.

theta0_null

Prespecified historical reference or unacceptable response rate; a probability between 0 and 1.

theta0_clin

Clinically meaningful target response rate for a supplementary predictive assessment; a probability between 0 and 1.

gamma_final

Final posterior probability threshold for trial success relative to theta0_null; a probability between 0 and 1.

gamma_clin_final

Final posterior probability threshold for the supplementary criterion relative to theta0_clin; a probability between 0 and 1.

S

Positive integer number of Monte Carlo replications used to estimate predictive probabilities.

type

Results to return: "posterior" (default), "predictive", or "both".

seed

Optional integer seed for reproducibility.

Value

A named list containing the components specified by type. The default returns posterior summaries; "predictive" returns predictive probabilities; "both" returns all components:

posterior

Shape parameters alpha and beta of the current Beta posterior.

posterior_mean

Posterior mean of the response rate.

posterior_sd

Posterior standard deviation of the response rate.

p_null

Posterior probability that the response rate exceeds theta0_null.

ppos_null

Predictive probability of meeting the final efficacy criterion defined by theta0_null and gamma_final.

ppos_clin

Predictive probability of meeting the supplementary criterion defined by theta0_clin and gamma_clin_final.

Predictive probabilities are estimated by Monte Carlo simulation, conditional on the observed data and completion of the planned sample size N_total.

References

Zhao, J., and Mu, R. (2026). Introduction to Bayesian Decision Method and Its Application in Medical Research. Manuscript in preparation (in Chinese).

Thall, P. F., and Simon, R. M. (1994). Practical Bayesian guidelines for phase IIB clinical trials. Biometrics, 50, 337–349. doi:10.2307/2533377.

Lee, J. J., and Liu, D. D. (2008). A predictive probability design for phase II cancer clinical trials. Clinical Trials, 5, 93–106. doi:10.1177/1740774508089279.

See Also

oc_singlearm_binary

Examples

posterior <- posterior_predictive_singlearm_binary(
  x_obs = 9, n_current = 20, N_total = 40
)
posterior$p_null

predictive <- posterior_predictive_singlearm_binary(
  x_obs = 9, n_current = 20, N_total = 40,
  type = "predictive"
)
predictive$ppos_null

both <- posterior_predictive_singlearm_binary(
  x_obs = 9, n_current = 20, N_total = 40,
  type = "both"
)
both