Regression analysis of group-tested current status data

Area of Science:

• Biostatistics
• Epidemiology
• Public Health

Background:

• Group testing reduces costs and time for large-scale infectious disease screening by pooling specimens.
• Current status data, where individuals are assessed once for a time-to-event endpoint, is common in such studies.
• Analyzing group-tested current status data presents unique statistical challenges.

Purpose of the Study:

• To develop and evaluate statistical methods for analyzing group-tested current status data.
• To estimate proportional hazard regression models using pool test outcomes.
• To provide a computationally efficient and statistically rigorous approach for group testing data.

Main Methods:

• A sieve maximum likelihood estimation approach was developed, approximating the cumulative baseline hazard function.
• A computationally efficient expectation-maximization algorithm using data augmentation was derived.
• Modern empirical process theory was applied to establish asymptotic properties of the estimator.

Main Results:

• The proposed sieve estimation method accurately analyzes group-tested current status data.
• Simulation studies demonstrated the method's nominal performance and advantages over individual testing analysis.
• The approach was successfully applied to a real-world chlamydia dataset.

Conclusions:

• The developed statistical methods offer an effective way to analyze complex group-tested current status data.
• This approach enhances the efficiency and accuracy of infectious disease screening analysis.
• The findings have significant implications for public health surveillance and research.