Twenty years later: a single-centre, repeat retrospective analysis of equine perioperative mortality and investigation of recovery quality.

Abstract: To determine the mortality rates associated with equine anaesthesia for elective and emergency (colic and non-colic) cases in one equine, university teaching hospital and to investigate the effect of several horse- and anaesthetic-related variables on anaesthetic recovery quality. Methods: Retrospective data analysis. Unassigned: In total, 1416 horses undergoing anaesthesia between May 2010 and December 2013. Methods: Patient information and details of the anaesthetic, recovery period and immediate complications were extracted from an archiving database. Statistical evaluation of factors affecting mortality included chi-squared tests and binary logistic regression. Factors affecting recovery quality were investigated using univariable and multivariable ordinal logistic regression. Statistical significance was set at p < 0.05. Results: Anaesthesia/recovery-related mortality was 1.1% for all cases, 0.9% for elective cases, 1.6% for colics and 0% for non-colic emergencies. Fractures and dislocations accounted for the majority (71.4%) of deaths. No intra-operative deaths occurred during the study period. Risk factors for mortality included increasing age and American Society of Anesthesiologist's (ASA) status but these and other factors were confounded by 'colic'. Non-fatal complications in the immediate recovery period included postanaesthetic myopathy/neuropathy and postanaesthetic respiratory obstruction. Recovery quality was associated with body mass (p = 0.016), ASA status 3 and 4 (p = 0.020 and 0.002, respectively), duration of anaesthesia (p < 0.001) and out-of-hours anaesthesia (p = 0.013). Although recovery quality was also influenced by age and breed-type, these factors were removed from the final model as age was highly associated with both ASA status (p < 0.001) and colic surgery (p < 0.001), and breed-type was a determinant of body mass. Conclusions: Anaesthetic/recovery-associated mortality was comparable to previously reported figures except intra-operative deaths were not reported. Fractures remained responsible for the largest proportion of recovery-associated deaths. Improvements to the recovery process that can reduce fracture occurrence are still required.