Effects of inter-individual variability in experimenters' sensitivity and training on behavioral assessment of animal models of vestibular disorders

ROMAIN BOULARAND ¹ Bérénice HATAT ¹ CLAIRE BRINGUIER ¹ NICOLAS CHANUT ¹ Abdessadek El AHMADI ² STEPHANE BESNARD ³ BRAHIM TIGHILET ² CHRISTIAN CHABBERT ^2*

• ¹ Vertidiag, Montpellier, France
• ² UMR7077 Centre de Recherche en Psychologie et Neurosciences (CRPN), Marseille, Provence-Alpes-Côte d'Azur, France
• ³ UR Vertex 7480, Université de Caen Normandie, CHU Caen, Caen, France

This study was designed to explore the correlation between animal behavioral assessment quality and rater's individual sensitivity and training. We selected different raters to form a panel to rate the severity of posturo-locomotor deficits in animals displaying excitotoxic or ototoxic lesions-induced vestibular syndrome. All raters, regardless of their scientific level, received brief training based on videos and tutorial files. They then had to score videos of rats with different types and stages of vestibular syndromes. All data were collected and analyzed. Inter-individual variability in raters significantly altered the results of behavioral assessment of posturo-locomotor deficits in vestibulo-lesioned animals.Neither gender nor scientific level had an impact on the results. In contrast, the sensitivity of the individual to animal welfare impacted the mean score in the ototoxic lesion model.Raters with high sensitivity tended to exaggerate the symptomatology. The use of automated assessments of posturo-locomotor deficits in vestibulo-lesioned rodents, is the best solution to limit these assessment biases.Identifying and reducing experimental bias is part of the scientific process to better control animal models predictability. Sensitivity of raters to animal welfare impacts behavioral assessment in ototoxic lesion model. Small raters team with similar training can avoid bias linked to inter-individual variability.