Veterinary research in clinical practice

While randomisation is an ideal for good statistical analysis of the outcomes being evaluated, in veterinary surgery randomisation is almost always not applicable to clinical research because subjecting patients to novel treatments where previous treatments have become established and accepted would be considered experimentation. Often in surgical research the research method is necessarily suboptimal from a statistical point of view, but optimal from a surgical development point of view, in which the various aspects of bias are brought into sharp focus to achieve a rigorous conclusion. Raw evidence gathered can be of various levels of quality, but we should not denigrate this raw data for the reason it is 'lower quality evidence' on the basis that it fails to meet the quality assurance measures, because it is a critically important part of the whole research process and effort.