Maximising the impact of evidence-based medicine on equine health and welfare

Evidence-based veterinary medicine (EVM) has the potential to improve welfare of equine patients in primary care practice by ensuring that decision making at the point of care is underpinned by credible research. Patient focussed care requires consideration of a combination of evidence, veterinary judgement, experience and owner associated variables to ensure the best possible outcomes for the animal, whilst maximising the impact of EVM on equine health and welfare. A holistic approach to development and implementation of EVM, that involves horse owners, other equine stakeholders and veterinary surgeons, will ensure that all parties involved in decision making on behalf of equine patients are engaged in evidence-based decisions. The central aim of the thesis was to determine how advances in evidence-based veterinary medicine had the potential to achieve the greatest impact on equine health and welfare. The objectives were to: • Review methods of implementing evidence-based medicine in clinical practice and determine their suitability for veterinary medicine. • Describe the prevalence and outcome of conditions seen for ‘out of hours’ primary assessment in equine practice. • Investigate horse owner perception, understanding and approach to the most common emergency conditions. • Identify paucities in the knowledge and resources available to equine owners and highlight the information required to implement and facilitate evidence-based decision making. • Use an evidence-based approach to develop and disseminate an educational resource for horse owners to improve understanding of an emergency condition and equine welfare associated with that specific disease. The first section of the thesis aimed to identify a common emergency condition with a major welfare impact, which would become a focus for research in subsequent studies. A combination of methods were used in these chapters including a retrospective case study of ‘out of hours’ case records from two practices over a three year period, and a mixed methods online survey of opinions and experiences of equine yard owners. In the retrospective case study, colic (abdominal pain) was the most common condition seen as an ‘out of hours’ emergency at both veterinary practices, accounting for approximately a third of the ‘out of hours’ caseload. The three most common reasons for ‘out of hours’ visits for both practices were colic (35%), wounds (20%) and lameness (11%). The majority of cases seen ‘out of hours’ required a single treatment for resolution of the presenting problem (58%), whilst 26% needed multiple treatments. An online survey consisting of open and closed questions was distributed to UK livery yard owners. The survey investigated participants’ demographics and experiences, their opinion of the most common emergency conditions in the horse, and decision-making in emergency situations. Descriptive analysis of data included frequency ranking and categorisation of free text responses. There were 104 survey participants. The majority had kept horses for more than ten years (97%), and had previous experience of emergency conditions (99%), predominantly of colic (96%) and wounds (92%). Participants considered that the most common emergency conditions were colic (98%), wounds (49%) and fractures (22%), and the most concerning conditions were colic (94%), lameness (36%) and wounds (21%). Factors important in emergency decision making were: degree of pain, likelihood of condition resolving, and severity of disease. The results of these studies provided the impetus for using colic as the focus for the next phases of work within this thesis. The second section of the thesis investigated veterinary practitioners’ and horse owners’ approaches to horses with abdominal pain, with the aim of identifying where knowledge or resources were good or were variable or lacking. The subset of horses with abdominal pain from the ‘out of hours’ retrospective case series were explored to describe how these animals presented to veterinary surgeons, how they were diagnosed, treated and the outcomes. The clinical signs associated with a ‘critical’ outcome of colic were determined using univariate logistic regression. In the retrospective case series, 941 cases presented with signs of abdominal pain; 23.9% (n=225/941) cases were categorised as ‘critical’, and 18% (n=168/941) were euthanased. Fifteen variables from the case presentation significantly correlated with a critical case (p<0.01) and were therefore incorporated into the multivariate model. The final multivariable model included three variables significantly associated with the likelihood of a case being classified as ‘critical’: increased heart rate (p<0.01), abnormal mucous membrane colour (p<0.01) and absence of borborygmi in at least one quadrant (p