Common decisions made and actions taken during small-animal consultations at eight first-opinion practices in the United Kingdom

In order for veterinary surgeons to undertake an evidence-based approach to making decisions about theirpatients, it is important that new evidence is generated to support the clinical decision-making process.Many of the decisions are likely to be around the actions taken to treat or manage health problemsdiscussed during the consultation, and little is currently known about the factors which affect the typeof action taken. The aim of this study was to determine the decisions made and actions taken for healthproblems discussed during first-opinion small-animal consultations, as well as identifying factors whichmay affect the decision-making process.Data were gathered during direct observation of small-animal consultations conducted by 62 veterinarysurgeons in eight first-opinion practices in the United Kingdom. For each patient presented, data weregathered on all health problems discussed during the consultation. The decision made (whether an actionwas taken or not) and the action taken where applicable (e.g. therapeutic treatment with antibiotics) wasalso recorded. A three-level multivariable logistic-regression model was developed, with problem (Level1) nested within patient (Level 2) nested within consulting veterinary surgeon (Level 3), and a binaryoutcome variable of action versus no action.At least one action was taken for 69% (n = 2203/3192) of all problems discussed. Therapeutic treatmentwas the most common action taken (n = 1286/3192 problems; 40.3%), followed by management advice(n = 1040/3192; 32.6%) and diagnostic work-up (n = 323/3192; 10.1%). The most common therapeutictreatment was antibiotics (n = 386/1286; 30%), while the most common management advice given wasdietary advice (n = 509/1040; 48.9%). The three explanatory variables remaining in the final model werewhether the problem was a presenting or non-presenting problem, the type of diagnosis made, andthe body system affected. Explanatory variables which did not remain in the final model were patientsignalment, problem history, consultation type, clinical examination type, and who raised the problem(veterinary surgeon or owner).For over two-thirds of problems discussed, an action was taken which suggests these problems maybe seen as important by the veterinary surgeon and/or pet owner. No action was taken for almost a thirdof cases which could represent ‘watchful waiting’, which has been highlighted as important in humanhealthcare. Future research should focus on the common actions taken, further exploring the complexdecision-making process, and examining the effect of the decisions made on long-term patient outcomes