Opportunities and challenges when pooling milk samples using ELISA.

Area of Science:

• Veterinary diagnostics
• Laboratory automation
• Biostatistics

Background:

• High-throughput sample testing is costly and time-consuming.
• Sample pooling offers a potential strategy to optimize testing efficiency.
• The effectiveness of pooling depends on test sensitivity, specificity, and cost parameters.

Purpose of the Study:

• To evaluate and compare hierarchical and non-hierarchical sample pooling schemes against standard pooling.
• To determine the efficiency and potential cost savings of different pooling strategies based on prevalence and pool size.
• To assess the impact of analytical sensitivity changes in pooled samples on diagnostic accuracy.

Main Methods:

• Developed and tested two novel pooling frameworks (hierarchical and non-hierarchical).
• Compared pooling schemes with standard methods using simulated data and real-world milk sample assays (ELISA).
• Incorporated cost parameters (retesting, equipment, labor) to calculate total savings.

Main Results:

• Analytical sensitivity decreased with increasing pool sizes in milk assays for Salmonella Dublin, Mycobacterium avium spp. paratuberculosis, and bovine virus diarrhea.
• Both tested pooling schemes showed potential optimality depending on prevalence, pool size, and retesting costs.
• A three-stage hierarchical scheme was often more efficient in test numbers, while a two-stage scheme offered better savings when re-pooling costs were high.

Conclusions:

• Sample pooling can be cost-effective, but efficiency varies with pooling strategy, prevalence, and laboratory costs.
• The choice between hierarchical and non-hierarchical pooling depends on specific economic factors and desired outcomes (test reduction vs. overall savings).
• A web application was developed to aid laboratories in selecting optimal pooling strategies based on their parameters.