Keywords: generalized linear mixed models, ordinal endpoint, battery cage study, intestinal lesion score
Establishing the efficacy of an anti-coccidial drug entails conducting multiple battery studies, where the target animals are inoculated with single and mixed Eimeria species under controlled laboratory conditions. One of the primary outcomes in a battery study is the intestinal lesion score defined on a discrete ordinal scale of zero to four. So far, the linear mixed model (LMM) has been routinely used in the regulatory evaluation of anti-coccidial drugs. This present work proposes to apply the generalized linear mixed model (GLMM) with the cumulative logit link to analyze coccidial lesion scores. Upon applying this new approach on nine real-world datasets, it is observed that the proposed GLMM fitted the data well and, at the 0.05 level, generated statistical results in complete concordance with the traditional LMM approach. Advantages of the GLMM over the LMM are discussed from statistical and clinical standpoints.