Pool testing with dilution effects and heterogeneous priors

Area of Science:

• Epidemiology
• Biostatistics
• Public Health

Background:

• Pooled testing, like the Dorfman scheme, combines individual samples to reduce screening costs.
• Traditional pooling methods may not optimize test efficiency or accuracy.
• The choice of how to group individuals (partitioning) significantly impacts pooled testing performance.

Purpose of the Study:

• To analyze the performance of ordered partitions in Dorfman pooled testing.
• To identify conditions where ordered partitions minimize expected tests, false negatives, and false positives.
• To propose and evaluate an equity measure for pooled testing strategies.

Main Methods:

• Mathematical analysis of ordered partitions within the Dorfman pooled testing framework.
• Derivation of sufficient conditions for performance optimization.
• Development and analysis of an equity metric for population screening.

Main Results:

• Ordered partitions can outperform other partitioning strategies in minimizing expected tests and classification errors.
• Sufficient conditions for the superiority of ordered partitions are derived and verifiable.
• Conditions for maximizing an equity measure using ordered partitions are established.

Conclusions:

• Ordered partitions offer an efficient and potentially more equitable approach to population screening using Dorfman pooled testing.
• The findings provide practical guidelines for implementing optimized pooled testing strategies.
• Further research into dilution effects and equity is warranted for refining pooled testing protocols.