Videotaped behavior as a predictor of clinical outcome in rhesus macaques (Macaca mulatta)

Understanding the behavior of laboratory NHP facilitates health assessment and clinical care. We sought to characterize the behavior of critically ill rhesus macaques (Macaca mulatta) and determine whether specific behaviors or behavioral changes might facilitate the determination of prognosis and clinical endpoints. Twenty-two critically-ill subjects were videorecorded after they were removed from the outdoor breeding colony for diagnostic work-up and treatment. Subjects were categorized as survivors (n = 15) and those that were euthanized according to existing clinical endpoints (n = 7). Behavior before, during, and after cageside examination was compared between these groups with regard to the presence or absence of direct observation. This approach allowed us to determine whether these settings revealed differences between groups or masking of behaviors during direct observation. Before cageside examination, several behaviors (for example, self-grooming and anxiety behaviors) were significantly more common in surviving subjects than in euthanized subjects. Few significant differences in behavior were detectable during or after the examination. Subjects that were eventually euthanized showed more illness-related behaviors; however, not all animals requiring euthanasia showed these signs when an observer was present. Furthermore, euthanized animals spent more time in an alert posture during direct observation than at other times. Therefore, direct observation of critically ill rhesus macaques may not yield the most accurate assessment of illness severity, and using video to assess behavior may be helpful for prognosis.