AUTHOR COMMENTARIES - 2008

Currently, Dr. Sheehan is Professor of Psychiatry and Behavioral Medicine and Director of the Depression and Anxiety Disorders Research Institute at the University of South Florida College of Medicine, as well as Professor of Psychology at the University of South Florida College of Arts and Sciences.

In the interview below, he talks with ScienceWatch.com correspondent Gary Taubes about his work on the MINI project.

What prompted you to develop the Mini-International Neuropsychiatric Interview (MINI) and your highly cited 1998 Journal of Clinical Psychiatry paper?

Back in the early 1990s, I was asked to help set up studies, mainly in Europe, on anxiety disorders. One pharmaceutical company in particular wanted to recruit patients with panic disorder. I recommended it use a structured diagnostic interview so that the company could be sure that all of the patients it enrolled had this specific disorder. Several of the European researchers, however were hesitant about the need for a structured diagnostic interview. It was their impression that structured interviews were too long and were a burden on the efficient implementation of a treatment study. They didn’t have the time, the resources or the staff, they said, to do all these complicated screenings. Wasn’t there an easier way of accomplishing the same goal?

I agreed that that the existing structured diagnostic interviews were long and difficult to navigate. The company wanted to know if I could make a new interview that was shorter, easier to navigate, and more elegant. I said, "Yes, that’s possible." Another person in that advisory group, Yves Lecrubier, said, "I agree with David. There should be a way to make a shorter, more elegant one." The medical director for the drug company asked if I would go ahead and develop such an instrument. In a weak moment, I said, "Yes, why not." And I added, "Yves also seems interested. Maybe the two of us can get together this evening and report back in the morning."

So Yves and I had dinner. We sketched out how we could create a new interview that would be a fraction of the size of the existing one, and we talked about how to make it much more user-friendly— so that it could be used in non-research clinic settings as well as in research. We reported back, saying that it seemed this project could be done, and we’d be happy to do it. And the company then said they’d be interested in funding us to do a validation study if we developed the structured interview.

How did it proceed from there, particularly since you and Dr. Lecrubier work on different sides of the Atlantic?

We corresponded back and forth and met together. Yves is at INSERM in Paris (France’s National Institute for Health), and he’s also an expert in psychiatric diagnosis and psychopharmacology. It took us a while, but we developed our first draft and the company gave us the funds to do a validation study, to see how it stacked up against the two other most widely used structured diagnostic interviews. The long and the short of it is that we were able to show that our diagnostic interview had very similar psychometric properties to the others, but was much easier to use and considerably shorter. The MINI was developed to assist in the precise diagnosis of the 16 most common psychiatric disorders, conditions like major depression, bipolar disorder, panic disorder, psychotic disorders and substance use disorders.

Was that a surprise to you—that it performed just as well?

Not really, because we were able to build on the shoulders of giants, so to speak. We knew what the weaknesses of the other interviews were. That meant that we could make the MINI strong where the others were weak. For example, if one of the then current interviews was strong on anxiety disorders but weak on psychotic disorders and bipolar depression, and the other was strong on the more severe psychiatric disorders, we tried to capture the strengths of each and circumvent the weaknesses. So, in that sense, the study showed that it came out exactly as we had planned. The MINI was strong where each of the others were weak.

That study then was the subject of several publications. Probably the most important one was the one in the Journal of Clinical Psychiatry, which is the highly cited paper. That paper came out years after the study was actually done - because it took a large number of patients, with several hundred thousand datapoints of information, to collate and analyze. It was an arduous task.

What did you do after that first validation study and the collating job?

Subsequently we did other validation studies in Europe, in the Middle East and in Asia, where we tested the MINI against expert opinion. People then approached us to translate the MINI into other languages. We co-opted a group of wonderful, bright, highly experienced diagnosticians from all over the world to help us in the translation efforts. Then we got one of the companies to fund a big international meeting where all these translation teams came together. This was about a decade ago, and we tried to get everybody on the same page so that all of the translations captured the spirit of the MINI and were not just literal translations of words.

Can you give an example of what you mean by that?

Here’s one. I don’t think that this particular item is in the MINI anymore in the exact same way. Still, it’s a good illustration. Let’s say that you’re asking a patient in a southern state, "Are you feeling down-hearted, blue?" The individual may respond, "Yeah, sure, I feel down-hearted, blue" quite a lot of the time. If you translate those words literally into French, the patient is likely to look at you and ask, "What do you mean? My heart is down and I’m feeling the color blue?" There’s a great deal lost in the translation. Many languages have no word for depression. In others—Arabic, for instance—there are multiple words for depression. In Pashto, a language used in Afghanistan and western Pakistan, they use the same work for both anxiety and depression.

The issue for us was how to capture all these subtle nuances from one language to the next, not just the experience of the disorder, but also the experience of the patient, and try to get that constant across all the languages. That way everybody can understand the spirit of each question and the experience at a level almost beyond language.

This conference must have been a unique experience.

Absolutely! I’ve given lectures and attended in conferences in 65 countries around the world, but this was the most fascinating conference I’ve ever attended. Nothing else was ever like it. What we began to realize was that in translating the MINI we were harnessing the research and academic experience of psychiatrists all over the world. Prior to this, clinicians and researchers used different diagnostic systems. We were bringing them all onto the same page and pointing them all to magnetic north at a very practical level. In the collaboration, we were providing each other with a practical recipe for psychiatric diagnosis in a way that everyone could use in their everyday work and in a way that they could all understand. That in turn greatly increases the level of input and expertise you bring to solving the puzzles of these psychiatric disorders. At the same time you are harnessing all this international energy and getting them all to speak a common diagnostic language so that they can collaborate together. This sets the stage for giant international collaborations in solving these illnesses and finding better treatments.

Do you think this contributes to the huge influence your paper and the MINI have had?

Yes. If you look at the structured interview itself, the MINI, you’d say maybe anybody could have done that. But as the first very widely translated, global, psychiatric interview, it was bringing everybody into the same diagnostic system. Everybody had to be part of it, and everybody had to communicate with each other at international meetings. Somebody could come to a big meeting in the US and say, "I did a study with 500 people with bipolar depression and 200 with major depressive disorder, and we’re doing it the exact same way that you’re doing it."

So it allowed everybody to take everybody else’s work seriously, and it allowed the world’s regulatory agencies, like the FDA, to be able to use research from other countries, because they’re all using the same system. And, hopefully, this also decreases the cost of psychiatric medical research, because now we can pool resources. We don’t have to replicate everything in every country. So this brought everything and everybody together, and that’s been very exciting. The reason people now use the MINI so much is that it allows everybody in every country to be able to participate in this common enterprise.

Was this a deliberate plan from the beginning, or something you realized as the project evolved?

We tried from the get-go to make this more than just a U.S. system. We realized that we were at the dawn of global medical research and this would require international collaboration and harmonization. This was a deliberate plan. We were trying to bring everybody together with the MINI. And we always said this is not a completed product. It is an exercise in progress, and we will continue to refine and tweak and improve it with continued input from people and data from studies all over world to make it better and better. We have relentlessly and ceaselessly done that.

What was the hardest or most challenging aspect of the MINI?

The interesting thing, I have to say, is that it never has really felt like work. The pleasure we have had in collaborating on this venture has far outweighed the challenges. Yves and I have both worked on the MINI mainly at night and on weekends. We don’t have time during the day to work on it. Accepting that the development of the MINI is not a destination but a process has made the challenge easier to manage. It is an exercise in paying attention to fine, tiny little details, subtle nuances most people would totally miss on first reading. We tried to capture the essence of the experience of each of the key symptoms and signs of these psychiatric disorders. There are layers of these in the MINI structured diagnostic interview.

You said originally that you set out to incorporate the strengths of the other structured diagnostic interviews, while simultaneously making it considerably shorter. How did you go about doing that?

When we looked at the existing structured interviews, Yves and I both realized very quickly that not all questions had equal weight. Some are more potent at hitting the bull’s eye right off the bat than others. Many of the questions in other interviews are just frill, icing on the cake. They’re not relevant to making clinical decisions.

Our task was to cut everything down to the essentials so that the MINI could make a valid and accurate diagnosis with as few questions as possible. This is important for researchers and clinicians who have to make a clinical decision—for example include or exclude an individual from a study or to start treatment. You look for the items that are loaded up most heavily toward that end. Some are very potent; some are very weak. We used what’s called decision tree logic: if you ask one or two screen questions and the answer is no, then you skip over the next items and go to the next disorder, much as clinicians with all illnesses in their busy practices.

Is there one last point you’d like to make about the MINI or your work in general?

About the MINI, when I’m training people on it, I always emphasize a lesson I learned from classical music. I have a great interest in music, and there’s a famous quote by Isaac Stern, who said that it’s not the notes that count, it’s the pauses between the notes.

When you are reading the MINI questions to a patient and administering the interview, don’t just read off the words—blah, blah, blah, in this flat way—you have to present the information in the question, pausing appropriately between phrases, while you watch the patient digest what you’ve said. This is all about cadence and using language.

One of the things people occasionally spot with the MINI is that many of the questions in English are phrased in such a way that they flow in a musical sense. They have a rhythm and a cadence to them. And that rhythm is there to help the patient get it and digest it and not be overwhelmed all at once by too much detail.

I have to say, we didn’t dream all this up in one night. It came from years of thinking about it, and fiddling with it, always improving it.

David V. Sheehan, M.D. MBA
Professor of Psychiatry and Director,
Depression and Anxiety Disorders Research Institute
University of South Florida College of Medicine
Tampa, FL, USA

Keywords: MINI, structured diagnostic interviews, psychological diagnosis