bbssr 1.0.2
Minor Updates
- Fixed title case in DESCRIPTION file for CRAN submission
- Updated from “Re-estimation” to “Re-Estimation” as requested by
CRAN
bbssr 1.0.1
Minor Updates
- Function Removal: Removed
ClopperPearsonCI()
function as it was not being used in the
main BSSR functionality
- Documentation Updates: Updated all documentation to
reflect the removal of confidence interval functionality
- Package Optimization: Streamlined package to focus
on core BSSR methods
bbssr 1.0.0
Initial Release
This is the first release of bbssr
, a comprehensive R
package for blinded sample size re-estimation (BSSR) in two-arm clinical
trials with binary endpoints.
Main Features
- Blinded Sample Size Re-estimation: Implement
adaptive trial designs with
BinaryPowerBSSR()
- Multiple Exact Statistical Tests: Support for 5
different exact tests:
- Pearson chi-squared test (
'Chisq'
)
- Fisher exact test (
'Fisher'
)
- Fisher mid-p test (
'Fisher-midP'
)
- Z-pooled exact unconditional test (
'Z-pool'
)
- Boschloo exact unconditional test (
'Boschloo'
)
- Flexible Design Options: Choose between restricted,
unrestricted, and weighted BSSR approaches
- Traditional Methods: Calculate power
(
BinaryPower()
) and sample sizes
(BinarySampleSize()
) for fixed designs
- Exact Confidence Intervals: Clopper-Pearson
confidence intervals (
ClopperPearsonCI()
)
- Rejection Regions: Compute exact rejection regions
(
BinaryRR()
)
Design Approaches
- Restricted Design: Conservative approach ensuring
final sample size ≥ initial sample size
- Unrestricted Design: Flexible approach allowing
both sample size increases and decreases
- Weighted Design: Advanced approach using weighted
averaging across interim scenarios
Documentation
- Comprehensive documentation with examples for all functions
- Detailed vignettes explaining methodology and usage:
vignette("bbssr-introduction")
- Getting started
guide
vignette("bbssr-statistical-methods")
- Statistical
methodology
- Complete README with practical examples
Statistical Validity
- All methods maintain exact Type I error control at specified α
level
- Exact statistical tests rather than asymptotic approximations
- Suitable for small to moderate sample sizes common in clinical
trials
Dependencies
- Base R (≥ 3.5.0)
fpCompare
for robust floating-point comparisons
stats
for statistical functions
Development
- Package follows R package development best practices
- Comprehensive documentation with roxygen2
- Ready for CRAN submission