Abstract
Viral pathogenesis and therapeutic screening studies that utilize small mammalian models rely on the accurate quantification and interpretation of morbidity measurements, such as weight and body temperature, which can vary depending on the model, agent and/or experimental design used. As a result, morbidity-related data are frequently normalized within and across screening studies to aid with their interpretation. However, such data normalization can be performed in a variety of ways, leading to differences in conclusions drawn and making comparisons between studies challenging. Here, we discuss variability in the normalization, interpretation, and presentation of morbidity measurements for four model species frequently used to study a diverse range of human viral pathogens - mice, hamsters, guinea pigs and ferrets. We also analyze findings aggregated from influenza A virus-infected ferrets to contextualize this discussion. We focus on serially collected weight and temperature data to illustrate how the conclusions drawn from this information can vary depending on how raw data are collected, normalized and measured. Taken together, this work supports continued efforts in understanding how normalization affects the interpretation of morbidity data and highlights best practices to improve the interpretation and utility of these findings for extrapolation to public health contexts.