This editorial needs a preface, its title seems to imply that a discussion on why clinical research is important is going to take place and it should be clarified that no such discussion will occur. I believe that every reader of this journal will already believe that clinical research is important, so further convincing them of this will have no real effect and no real impact. Just like the articles that this journal publishes, this editorial, too, seeks to have an impact. Its goal is to convince more clinicians to engage in clinical research.
Given our assumed shared belief in clinical research as a worthwhile pursuit, we can hypothesize that there exists some barrier that prevents more clinicians from engaging in research. My belief is that this barrier
normally occupies one of three forms. The first form would be a perceived lack of time to devote to research, for this, I can only quip that a lack of time is true for all things one could do and that the only time we ever have time for something is when we make time for it. The second form is one of uncertainty, where the lack of clarity in how to help the research community has led to the clinician’s interest stalling out. The third form originates from the acknowledgement that performing clinical research is its own area of mastery and a craft that requires many skills, and this daunting perception has left the clinician unable to make the required leap of faith into the arena of
clinical research. This editorial is set to be part of a series. Today, we will discuss what type of research is needed and try to add clarity to our shared goal. The remaining editorials of this series will be devoted to the planning, performing and writing of a research project. To start off our discussion, I’ll share some words of T. S. Eliot that I have appropriated from his play, “The Rock”:1
We live in a world where our access to information is greater than it ever has been before. And yet, even with all of this information, we know that there is still so much left for us to learn. What we need are clinicians that can take the information that is before them and interpret it within its context to produce new knowledge and other clinicians that can then take that knowledge and distill it into wisdom that can be more easily passed on. Today, I’ll highlight three areas where we can improve upon our ability to process the information available to us into wisdom. It is my hope that you, my reader, will take interest in one of these problems and commit to furthering humanity’s wisdom. The first problem relates to medical decision making being based on physician preference instead of evidence-based medicine. Now, in many cases, there are no suitable evidence-based guidelines and the only thing a physician can rely on is their own judgment and we affectionately refer to this as the “Art of Medicine.” But art can be messy and our patients are not canvases, they are the people that have placed their faith in us, a faith founded on the belief that we know what we are doing. It is impossible to know everything, but when we recognize that we don’t know something, then we owe it to our patients to start to correct that deficit. I should note that I appreciate physician preference and that I see it as a valuable source of information. An experienced local physician will often have a feeling for patterns in their patient population that a nationally-instituted guideline can’t compete with. Taking all of this into account, our true goal should be to ensure that the physician’s preference is evidence-based, as opposed to anecdote-based, and we can realize this goal through the effort of local clinicians that test, keep track of and report on the outcomes of the various local physician-preference treatment regimens. This data could then be used for the foundation of an ethically-approved randomized controlled trial.
The first problem relates to medical decision making being based on physician preference instead of evidence-based medicine. Now, in many cases, there are no suitable evidence-based guidelines and the only thing a physician can rely on is their own judgment and we affectionately refer to this as the “Art of Medicine.” But art can be messy and our patients are not canvases, they are the people that
have placed their faith in us, a faith founded on the belief that we know what we are doing. It is impossible to know everything, but when we recognize that we don’t know something, then we owe it to our patients to start to correct that deficit. I should note that I appreciate physician preference and that I see it as a valuable source of information. An experienced local physician will often have a feeling for patterns in their patient population that a nationally-instituted guideline can’t compete with. Taking all of this into account, our true goal should be to ensure that the physician’s preference is evidence-based, as opposed to anecdote-based, and we can realize this goal through the effort of local clinicians that test, keep track of and report on the outcomes of the various local physician-preference treatment regimens. This data could then be used for the foundation of an ethically-approved randomized controlled trial. documented new ideas in health care, with additional care taken to ensure that we don’t stonewall ideas that conflict with that which we want to be true. The third problem is our oversimplification of statistical significance and how we imply clinical significance from it. There is a greater truth found in statistical significance because there are robust objective methods used to determine it. Our perception is that statistical significance gives our studies validity, and that’s true, just not in as absolute of a sense as we’d like to think. What the p-value really tells us is how certain we can be that our findings are not due to random chance. The p-value does not say that our findings are important or otherwise meaningful. Instead, this determination of clinical significance is left to the clinician who will normally rely on their own expert opinion regarding how to categorize the result. At face value, this doesn’t seem bad, but this method introduces into our research a hidden reliance on what is inherently a subjective determination and this subjectivity enables bias to creep into our research findings. To mitigate this, we have to first change how we perceive the notions of statistical and clinical significance. Furthermore, we must maximize our usage of objective means to determine clinical significance using methods like the Number Needed to Treat or the Cohen’s Effect Size and we must demand to see these values in our literature, just as we do for the p-value. That said, when we consider a clinical significance in a more holistic sense, issues such as cost and adverse effects come into play and these issues are harder to handle in a systematic and objective manner, as such, some amount of subjective analysis will still occur. Nevertheless, our research’s potential for bias will be greatly reduced.