Return of the Righting Reflex Does Not Portend Recovery of Cognitive Function in Anesthetized Rats

As the number of individuals undergoing general anesthesia rises globally, it becomes increasingly important to understand how consciousness and cognition are restored after anesthesia. In rodents, levels of consciousness are traditionally captured by physiological responses such as the return of righting reflex (RORR). However, tracking the recovery of cognitive function is comparatively difficult. Here we use an operant conditioning task, the 5-choice serial reaction time task (5-CSRTT), to measure sustained attention, working memory, and inhibitory control in male and female rats as they recover from the effects of several different clinical anesthetics. In the 5-CSRTT, rats learn to attend to a five-windowed touchscreen for the presentation of a stimulus. Rats are rewarded with food pellets for selecting the correct window within the time limit. During each session we tracked both the proportion of correct (accuracy) and missed (omissions) responses over time. Cognitive recovery trajectories were assessed after isoflurane (2% for 1 h), sevoflurane (3% for 20 min), propofol (10 mg/kg I.V. bolus), ketamine (50 mg/kg I.V. infusion over 10 min), and dexmedetomidine (20 and 35 μg/kg I.V. infusions over 10 min) for up to 3 h following RORR. Rats were classified as having recovered accuracy performance when four of their last five responses were correct, and as having recovered low omission performance when they missed one or fewer of their last five trials. Following isoflurane, sevoflurane, and propofol anesthesia, the majority (63-88%) of rats recovered both accuracy and low omission performance within an hour of RORR. Following ketamine, accuracy performance recovers within 2 h in most (63%) rats, but low omission performance recovers in only a minority (32%) of rats within 3 h. Finally, following either high or low doses of dexmedetomidine, few rats (25-32%) recover accuracy performance, and even fewer (0-13%) recover low omission performance within 3 h. Regardless of the anesthetic, RORR latency is not correlated with 5-CSRTT performance, which suggests that recovery of neurocognitive function cannot be inferred from changes in levels of consciousness. These results demonstrate how operant conditioning tasks can be used to assess real-time recovery of neurocognitive function following different anesthetic regimens.