I never have understood clinicians who see clinical research as unrelated or even opposed to what they do. This attitude – that clinical research and clinical practice are opposites – even is enshrined in the US government’s official federal policy regarding research ethics (the 1976 “Belmont Report”). That policy stipulates that Institutional Review Boards (IRBs) have to give approval for research, with the standard aspects of informed consent forms and such. What is deemed clinical research is based on the criterion that the researcher is making decisions based on obtaining knowledge, not based on what is in the patient’s best interest. Clinical practice, in contrast, is marked by making decisions based on what is in the patient’s best interest, not based on obtaining knowledge.

The illogic of this distinction is clear: If a clinician is practicing, as opposed to doing research, based on what he/she thinks is in the patient’s best interest, then on what basis is that clinician making the judgment that is supposed to be in the patient’s best interest? On personal opinion? Well, that’s a weak basis to make decisions. Your opinion is different than my opinion; how is that somehow scientific or better than differences of opinion about religion or politics or a myriad of clearly non-scientific and non-medical activities. On personal experience? That’s weak too. Ditto about how that differs from religion or politics. On the combined experience of many (i.e., a professional consensus)? That’s a little better, but not much; in religion and politics there are groups with differences in consensus.

We pretend that in medicine we practice scientifically. We say that our standard is science, not just opinion or experience. Of course the latter have roles, but mainly when the science is unclear or weak. We prefer science as the main basis for our medical clinical practice. This applies to psychiatry as to any field of medicine.

If that’s the case, if our clinical practice should be based on science, then it would need to be based on scientific research, which is really a synonym for science. In other words, clinical practice is, or should be, based on research. If so, then when we say that clinical practice is based on what is in the patient’s best interest, we are basing that judgment on the best scientific research. In other words, clinical practice should be based on research. And yet the US government insists that practice and research are opposites. What is in the patient’s best interest is to make judgments based on research. Yet the government says that research is claimed to be the activity that seeks knowledge, not what is in the best interest of the patient. But practice cannot be in the patient’s best interest unless it is based on knowledge, which is established by research. So, in my view, even though the US government appears to disagree, clinical practice without a basis in research is unethical. And clinical research is not just about obtaining knowledge, because the point of that knowledge is to practice in the patient’s best interest.

The opposition, then, between research and practice is both false and unethical, in my viewpoint. You cannot be a good clinician unless you focus on research; in fact, the best clinicians are those who do research too. You also cannot be a good researcher unless you focus on clinical practice; in fact, the best researchers are active clinicians.

So for me the question of whether psychiatric residents should engage in research during their training, and when they should do so, is not an optional question. I don’t think that residents can become as good clinicians as they can if they do not learn about research, and preferably, engage in it.

What I really loved about research was that every time I ran into a clinical problem, or an issue of clinical uncertainty, I could collect some cases about it, create some data, and publish a paper about it. In other words, my clinical dilemmas could be translated into steps of progress in knowledge. Otherwise, the inevitable problems of clinical practice just sit there, unmoving, unimproved, festering in the mind of the clinician as insoluble mysteries that forever impede progress in practice. For me, I can’t imagine being a clinician without doing research. I’d be too frustrated and unhappy about my lack of knowledge. Obviously there are far more clinicians than researchers; so this problem doesn’t seem to bother a lot of other people. Maybe I’m unusual. I don’t know. But there is, and has been, no progress in clinical practice without research, so I don’t understand the unwillingness among most colleagues to do both.

As to when and how to do research, my own experience was that I started in the third year of psychiatry residency, and picked it up strongly in my fourth and final year, and then did a research fellowship for one year after residency. I would recommend this approach, since the first two years of residency are too full with high-intensity clinical work (as medical interns and then first year psychiatry residents) for most people to do any research. The third year of residency (second in psychiatry) is outpatient, and allows for some time to start research activity, and the fourth year is mostly elective and allows for even more research time.

Find a research mentor. Join that person’s work. Or pick a topic of your own interest, and have a mentor guide you to research on that topic. When I was a third year resident at McLean Hospital, I was in supervision with Andrew Stoll, a clinical research fellow; I told him about one of my psychotherapy patients, recently hospitalized for mania, who had poor insight, which made psychotherapy very difficult. “Insight!” he exclaimed. “Why don’t you study insight?” My research topic came from my clinical practice. I then did a literature review, and found that no one had published a study on insight in mania (though there had been many studies of it in schizophrenia). This was my chance to add to our knowledge. I found a creative mentor at McLean Hospital in Harrison Pope Jr, who taught me how to plan a study – collecting a certain number of patients, interviewing them using a certain insight scale, collecting basic data from charts on their clinical course. I planned the study in my third year, and then conducted it in my fourth year, walking around the wards for about 6 months recruiting patients, and then, with the help of Stoll and Pope, organizing and analyzing and writing up my data. I published the paper the following year after graduating residency, my first research study. In the meantime, I wrote up my literature review with Pope’s help and published it in my senior residency year, my first published paper. Over the next decade, I conducted 2-3 more studies on insight in bipolar illness and depression, and published perhaps half a dozen more papers on the topic.

I had gotten the research bug, and it improved my clinical practice. I was very attuned to insight in mood patients; I realized that since insight was absent in mania but present in depression, many patients misreported their histories to deny mania, and were misdiagnosed as having unipolar depression rather than bipolar illness. I then designed a study, while working on an inpatient unit as an attending for a year, to document this misdiagnosis. I found that it was associated also with high rates of antidepressant use, which had been found to worsen bipolar illness. All of it came together: low insight, misdiagnosis, mistreatment with antidepressants, poor clinical outcomes. It led to changes in my clinical practice: correction for poor insight by getting history from family and friends, decreased misdiagnosis of bipolar illness, reduction in antidepressant use, improved clinical outcomes.

I then taught these findings to my residents and colleagues, teaching them about insight, and misdiagnosis, and antidepressants, and I was off in my academic career. It wasn’t just about “research” in isolation. It was about clinical practice identifying problems, which research clarified, which I then used to improve my clinical practice, which I then taught to the larger profession.

Practice, research, teaching – That’s how it all comes together. They call it the “triple threat” for academics. I don’t see it as optional. I think the best clinicians are researchers, and they are also the best teachers. Teachers who don’t do research are limited; clinicians who don’t do research are limited; researchers who don’t practice and don’t teach are limited. If you can do it all, you will do it all better than those who only do some.

For me, I wouldn’t have been a happy clinician if I wasn’t a researcher, nor would I have felt that I could teach anything worth teaching. For me, it was all or nothing.

This was my course; it’s not the only one. There are other variations. Obviously non-clinical research (with animals or in laboratories) may not require as much clinical activity to be effective. Most people who do formal research spend two years or more in a research fellowship. I later obtained a MPH to get more formal training in research methods and statistics, which greatly improved the quality of my research activity.

All trainees in psychiatry and allied mental health fields should ask themselves how much science matters in their planned activities. If science is relevant, then they are obligated, in my view, to learn as much about research as they can learn. They should do a research study themselves, if they can, in the middle to end of their training. At the very least, they should learn sufficiently about research methods and statistics to interpret research studies well (SN Ghaemi, A Clinician’s Guide to Statistics and Epidemiology in Clinical Practice, Cambridge University Press, 2009); this is the goal of the evidence-based medicine movement. I share these goals, and I realize how much we fall short of them in psychiatry.

I believe that most psychiatry residencies fail in not teaching research methods and statistics anywhere near the needed level. One problem is that most of the faculty are undereducated on these topics, hence they can’t teach them well. It’s unfair, but residents are left to themselves. That doesn’t work. You can’t teach yourself mathematics; you took many classes over many years of school to learn basic math. It’s the same with statistics; it can’t be self-taught. It has to be taught by a teacher in front of a blackboard.

In the meantime, my best advice would be to find a mentor and do a study, any study, but preferably something that comes from your clinical practice and that can advance knowledge, even a step, to improve that clinical practice. And then trust the rest to the gods of science, who, over the course of your clinical lifetime, might teach you a few more secrets about knowledge, if you realize you need them to learn them, and are willing to listen.