Optimal sample size allocation for two-arm superiority and non-inferiority trials with binary endpoints

• Verfasst von: Kirchner, Marietta [VerfasserIn]
• Verfasst von: Schüpke, Stefanie [VerfasserIn]
• Verfasst von: Kieser, Meinhard [VerfasserIn]
• Titel: Optimal sample size allocation for two-arm superiority and non-inferiority trials with binary endpoints
• Verf.angabe: Marietta Kirchner, Stefanie Schüpke, Meinhard Kieser
• E-Jahr: 2024
• Jahr: 12 March 2024
• Umfang: 9 S.
• Fussnoten: Erstveröffentlichung: 12.03.2024 ; Gesehen am 23.07.2024
• Titel Quelle: Enthalten in: Pharmaceutical statistics
• Ort Quelle: New York, NY : Wiley, 2002
• Jahr Quelle: 2024
• Band/Heft Quelle: (2024) early view, Seite 1-9
• ISSN Quelle: 1539-1612
• DOI: doi:10.1002/pst.2375
• Datenträger: Online-Ressource
• Sprache: eng
• Sach-SW: allocation ratio
• Sach-SW: clinical trial
• Sach-SW: power
• Sach-SW: sample size
• K10plus-PPN: 1896099092

Abstract

The sample size of a clinical trial has to be large enough to ensure sufficient power for achieving the aim the study. On the other side, for ethical and economical reasons it should not be larger than necessary. The sample size allocation is one of the parameters that influences the required total sample size. For two-arm superiority and non-inferiority trials with binary endpoints, we performed extensive computations over a wide range of scenarios to determine the optimal allocation ratio that minimizes the total sample size if all other parameters are fixed. The results demonstrate, that for both superiority and non-inferiority trials the optimal allocation may deviate considerably from the case of equal sample size in both groups. However, the saving in sample size when allocating the total sample size optimally as compared to balanced allocation is typically small.