Defining the Primary Research Question in Veterinary Clinical Studies

Veterinary clinical trials are experiments that test medical treatments in client-owned animals with naturally occurring health conditions. As veterinarians become increasingly familiar with the principles of evidence-based medicine, they are looking to the results of clinical trials to guide their practice. However, the methodological quality of published veterinary clinical trials is often poor,^1–4 which means that readers could draw incorrect conclusions from trial results and in turn make unfounded changes in patient care. Results of a recent review of study power in reports of small animal trials indicated that not only did most trials contain methodological flaws, but only a handful of studies even stated the primary research question.⁴ Whereas this may seem impossible (after all, how can an experiment have no question?), it is actually a well-recognized problem in clinical research: the general idea is clear, but the specific study question is inadequately defined.⁵ In trying to address large knowledge gaps with minimal resources, veterinary investigators often pose a broad research objective, design a study that will maximize the number and types of comparisons that can be made, comb the data to find interesting associations, and then devise explanations as to what these findings mean. Although well intentioned, trials that utilize this exploratory data-driven approach can result in biased or misleading results and are typically improperly designed to answer the questions of greatest clinical relevance. Of particular concern is the potential to arrive at false conclusions because of inappropriate study design, inadequate sample size, multiple statistical comparisons, or a combination of these issues. These are serious problems that affect the validity of clinical trials and other common research designs in veterinary medicine, such as retrospective case series and cohort studies as well as preclinical studies (eg, those involving purpose-bred animals or cadavers).

Development of a primary research question is essential to producing clinically relevant study results that can be applied to evidence-based practice.^5–8 A well-defined and specific research question helps determine the appropriate study design, sample population, and methods of data collection and analysis necessary to meaningfully address the knowledge gaps of greatest importance.⁸ In this article, I will review how to define a research question, explain how the research question determines key methodological elements of a study, and discuss common pitfalls of clinical studies that fail to clearly define a primary research question. Although the discussion will focus on clinical trials, the principles can and should also be applied to other types of study designs commonly encountered in veterinary medicine, including retrospective and preclinical studies. It is hoped that this review and discussion will encourage veterinary investigators in all disciplines to successfully apply these principles and to incorporate a clearly defined primary research question when designing and conducting their studies.