Blindrecalc - an R package for blinded sample size recalculation
Besides the type 1 and type 2 error rate and the clinically relevant effect size, the sample size of a clinical trial depends on so-called nuisance parameters for which the concrete values are usually unknown when a clinical trial is planned. When the uncertainty about the magnitude of these paramet...
Saved in:
| Main Authors: | , , |
|---|---|
| Format: | Article (Journal) |
| Language: | English |
| Published: |
June 21, 2022
|
| In: |
The R journal
Year: 2022, Volume: 14, Issue: 1, Pages: 137-145 |
| ISSN: | 2073-4859 |
| DOI: | 10.32614/RJ-2022-001 |
| Online Access: | Verlag, lizenzpflichtig, Volltext: https://doi.org/10.32614/RJ-2022-001
|
| Author Notes: | by Lukas Baumann, Maximilian Pilz, and Meinhard Kieser |
| Summary: | Besides the type 1 and type 2 error rate and the clinically relevant effect size, the sample size of a clinical trial depends on so-called nuisance parameters for which the concrete values are usually unknown when a clinical trial is planned. When the uncertainty about the magnitude of these parameters is high, an internal pilot study design with a blinded sample size recalculation can be used to achieve the target power even when the initially assumed value for the nuisance parameter is wrong. In this paper, we present the R-package blindrecalc that helps with planning a clinical trial with such a design by computing the operating characteristics and the distribution of the total sample size under different true values of the nuisance parameter. We implemented methods for continuous and binary outcomes in the superiority and the non-inferiority setting. |
|---|---|
| Item Description: | Gesehen am 21.12.2022 |
| Physical Description: | Online Resource |
| ISSN: | 2073-4859 |
| DOI: | 10.32614/RJ-2022-001 |