Geometric means provide a biased efficacy result when conducting a faecal egg count reduction test (FECRT)

Area of Science:

• Veterinary parasitology
• Statistical modeling in biological systems
• Anthelmintic drug efficacy evaluation

Background:

• Faecal egg count reduction tests are crucial for assessing anthelmintic drug efficacy.
• Estimating efficacy using arithmetic or geometric means can be influenced by the distribution of worm egg counts.
• Understanding the impact of host population variation and sample variation is key to accurate efficacy assessment.

Purpose of the Study:

• To compare the accuracy of arithmetic versus geometric means in estimating anthelmintic efficacy.
• To evaluate the influence of worm egg count aggregation on the performance of different statistical means.
• To provide recommendations for improving the methodology of faecal egg count reduction tests.

Main Methods:

• Simulation of faecal egg count reduction tests using Monte Carlo methods (10,000 runs).
• Modeling host population variation with the negative binomial distribution (NBD).
• Modeling sample variation with a Poisson distribution; geometric means calculated with constants (C=1, 12, 25).

Main Results:

• Arithmetic means consistently provided the best estimate of true efficacy across all levels of worm aggregation.
• Geometric means showed substantial overestimation of efficacy with moderate aggregation (C=1) and underestimation with high aggregation, irrespective of C.
• Increasing C reduced bias in geometric means but did not improve accuracy over arithmetic means.

Conclusions:

• Arithmetic means are superior to geometric means for estimating anthelmintic efficacy in faecal egg count reduction tests.
• Current guidelines for assessing anthelmintic resistance may need revision to incorporate arithmetic means.
• Accurate efficacy estimation is vital for managing anthelmintic resistance in livestock and other animal populations.