Closed-form confidence intervals for saved time using summary statistics in Alzheimer's disease studies

Area of Science:

• Neuroscience
• Clinical Trials Methodology
• Biostatistics

Background:

• Saved time is a key metric in Alzheimer's disease (AD) trials for communicating treatment benefits.
• Current methods for estimating saved time and confidence intervals (CI) have limitations, including not accounting for outcome correlations and lacking a practical closed-form CI.

Purpose of the Study:

• To derive a closed-form confidence interval (CI) for saved time in Alzheimer's disease (AD) trials.
• To address the gap in practical CI methods for researchers in AD clinical studies.

Main Methods:

• Derivation of a closed-form confidence interval (CI) for saved time.
• Comparison of the new CI method with existing approaches regarding coverage probability and interval width.
• Utilizing data from the phase 3 donanemab trials for practical illustration.

Main Results:

• A novel closed-form confidence interval (CI) for saved time was successfully derived.
• The new CI method was evaluated against existing techniques under diverse disease progression patterns common in AD trials.
• The application of the new CI methods was demonstrated using real-world data from the donanemab trials.

Conclusions:

• The developed closed-form CI offers a practical and potentially more accurate tool for analyzing saved time in Alzheimer's disease (AD) clinical trials.
• This advancement can improve the interpretation and communication of treatment benefits to stakeholders in AD research.