BIOMS Seminar: Biostatistics and Diagnostic Tests in Non-Domestic Veterinary Species

SPEAKER: 

Hugues Beaufrere, DVM, PhD, ACZM, ABVP(Avian), ECZM(Avian)

Assistant Professor, Department of Clinical Studies, Ontario Veterinary College, University of Guelph

ABSTRACT:

The ability of exotic veterinary specialists to diagnose diseases relies on the accuracy, reliability, and interpretation of diagnostic tests such as blood work in a given species. Discriminating between normal and abnormal findings may be difficult in non-domestic veterinary species such as birds, reptiles, and wild or zoo animals because the overall diagnostic accuracy of medical tests is hampered by a general lack of information, numerous sources of biological and analytical variability, diversity of species seen, and difficulties in cross-validating tests developed for domestic species.

As in conventional domestic species, the validation and interpretation of diagnostic tests in uncommon veterinary species rely heavily on statistical tools and principles. However, the high number of species encountered precludes the establishment of high-quality reference values in all species and brings a higher level of uncertainty in diagnostic tests interpretation. In addition, laboratory variability (pre-analytical, analytical, and post-analytical) and biological variability are greater in these species than in domestic carnivores. Reference intervals studies, validation studies, and method comparison studies are complicated by small sample sizes and unique species-specific characteristics inherent to working with exotic species.

This presentation will cover applied statistical principles underlying the establishment of reference values, interpretation of diagnostic tests, and validation of diagnostic methods in the context of non-domestic species. In order to maximize diagnostic accuracy given the limited available data and the extreme variability of target populations, a working knowledge of statistical principles and their limitations pertaining to diagnostic tests is essential to taking appropriate clinical decisions.