Statistical methods for unidirectional switch designs: Past, present, and future.

Area of Science:

• Clinical Trials Methodology
• Biostatistics
• Epidemiology

Background:

• Sequential treatment introduction in clinical trials is used to assess efficacy, driven by ethical, logistical, or disease-modifying effect considerations.
• Stepped wedge designs (SWDs) and delayed start designs (DSDs) are common unidirectional switch designs (USDs), often employed in large-scale pragmatic trials and drug development for cognitive decline, respectively.
• Existing statistical methods for USDs are complex, varied, and lack standardization, hindering research comparability.

Purpose of the Study:

• To provide a comprehensive overview of statistical developments for unidirectional switch designs (SWDs and DSDs).
• To unify various statistical approaches under a common framework for easier comparison of switch designs.
• To focus on statistical models, estimation, sample size calculation, and optimal designs for treatment effect estimation in USDs.

Main Methods:

• Formulating all USDs within a unified framework of treatment patterns.
• Reviewing past and current statistical developments for SWDs and DSDs.
• Discussing statistical models, estimation methods, sample size calculations, and optimal designs.

Main Results:

• A unified framework is proposed to facilitate comparison between different switch designs.
• The paper consolidates and clarifies complex statistical methodologies for USDs.
• Key aspects like estimation, sample size, and optimal design are addressed.

Conclusions:

• Standardizing statistical approaches for USDs is crucial for advancing clinical trial research.
• The unified framework aims to reduce barriers for researchers in selecting and comparing methods.
• Further research is suggested to address open issues in unidirectional switch designs.