Sample size calculation using Markov chains for a one-arm study of heroin administration routes

Area of Science:

• Biostatistics
• Clinical Trial Design
• Pharmacoeconomics

Background:

• Standard sample size calculations for time-to-event outcomes assume single events per unit.
• Repeating events, such as disease status changes or treatment switches, necessitate alternative modeling approaches.
• Classical survival time models are inadequate for trials with recurrent events.

Purpose of the Study:

• To develop a novel sample size calculation method for studies with repeating transition events.
• To apply Markov chain modeling for determining sample size in a feasibility study of a new drug intake route.
• To extend the method for power-based sample size calculations in multi-arm trials.

Main Methods:

• Modeling repeating transition events using homogeneous finite-state, higher-order Markov chains.
• Translating transition matrix assumptions into multinomial distributions of preferred administration routes.
• Calculating required sample size based on these distributions and the study's specific objectives.

Main Results:

• A method for sample size calculation in one-arm feasibility studies involving repeating events (e.g., drug intake route switches) was established.
• The method utilizes Markov chain transition probabilities to inform sample size requirements.
• The proposed methodology is adaptable for power calculations in multi-arm clinical trials.

Conclusions:

• Markov chain modeling provides a robust framework for sample size determination in trials with recurrent events.
• The developed method addresses limitations of traditional survival analysis for complex event patterns.
• This approach enhances the accuracy and feasibility of clinical trial design for specific research questions.