IN recent years, the opportunities from veterinary research that uses electronic patient record (EPR) data have been increasingly recognised and, especially in the UK, application of these methods is now gathering momentum (O'Neill 2014). Previously, veterinary clinical research had focused mainly on clinical data from referral centres, but the UK now has three major practice-based research (PBR) projects that collect and analyse primary care clinical data to answer questions relevant to primary care practitioners: the VetCompass Programme at the Royal Veterinary College, the Centre for Evidence Based Veterinary Medicine at the University of Nottingham, and SAVSNET at the University of Liverpool in partnership with the British Small Animal Veterinary Association. These projects offer a new research paradigm that actively engages practising veterinary surgeons in different ways and aims to effectively bridge between the research and practising arms of the profession. However, to get the best from these novel PBR systems, it is important to fully understand the strengths and weaknesses of research from primary and secondary-care clinical data (O'Neill and others 2014). Referral clinical caseloads can facilitate certain types of clinical research by generally providing good diagnostic precision, detailed clinical workup and access to specialist clinicians and equipment. These features ensure that studies based on referral caseloads can describe presentations, diagnostics and outcomes that may …