Anthelmintic resistance is an increasing challenge in the control of parasites in livestock.

The fecal egg count reduction test (FECRT) is the practical gold standard method for

evaluating resistance, but the interpretation is complicated due to high levels of

variability. Several factors contribute to cause this high variability and these must be

taken into account to accurately identify a reduction in anthelmintic efficacy. To address

this problem, we developed a hierarchical statistical model for analysis of FECRT data

from multiple farms. The model includes animal effect and farm clusters as random

effects. Resistance classifications are based on model adjusted lower confidence limit

(LCL) values of predicted mean efficacies on each farm. The model was used to evaluate

the efficacy of pyrantel embonate paste from 64 Danish horse farms. On these farms 614

out of 1644 horses had egg counts = 200 eggs per gram (EPG) and were treated. Post

treatment samples and information on age, gender and farm zip code were collected for

each horse. In addition, individual coprocultureswere performed on all pretreatment fecal

samples to determine the presence of Strongylus vulgaris, with 31farms (48.4 %) testing

positive. The efficacy of pyrantel was unaffected by the presence of S. vulgaris on the

farm. The cutoff LCL values used for classifying farms as pyrantel resistant were: >92%:

no resistance, 88-92%: suspect resistance, and <88%: resistance. Using model-adjusted

LCLs, we classified 7 (10.93 %) farms as pyrantel resistant, 5 (7.81 %) as suspect

resistant and the remainder of farms (81.25%) as not resistant. In comparison with

unadjusted LCLs, the model provided a more stable classification of farms with a 1.1 %

12

false discovery rate. The statistical model presented here can be adapted to handle

FECRT data obtained from other livestock species, drug types, and parasite species.