• Artists and Empiricists

    This is a cross-post from Ezra Klein’s blog, where Austin and I are guest-blogging this week.

    For better or for worse, when I hear people using the phrase “the art of medicine”, I hear an excuse for doing whatever you like. Personally, I’d prefer a little more science and a little less art

    Since reading Atul Gawande’s graduation address this weekend, I’ve been thinking about it far too much. That’s partly because I’m jealous of how talented he is, and partly because I don’t think his point can be emphasized strongly enough. He suggested that, when it comes to physicians, we need fewer cowboys and more pit crews. In other words, we as physicians imagine ourselves as heroes working individually, instead of as part of a team, perhaps in a more supporting role.

    I think there’s something to this, but it’s only a part of it. We in medicine are obsessed with the new. Whether it’s the latest drug, the newest procedure, or the latest homeopathic cure, we are sure that the thing that’s going to make everything better is just around the corner. This focus, however, means that we spend too little time fixing what we already have.

    Take asthma. It’s not like we don’t know how to care for it; we’ve known for quite some time. There are guidelines and evidence, and it’s all there for anyone who wants it. The medicines we’ve used for asthma aren’t new either. Yet, things have not improved:

    There is evidence that applying guidelines in clinical practice has not been highly successful. Certainly, examination of available data do not support a positive impact for improving asthma outcomes in the U.S pediatric population at large. Neither hospitalizations, emergency department visits, nor deaths have decreased in children since the inception of the Guidelines in 1991. Because younger children, particularly, have had the highest rate of hospitalization for asthma, and asthma most frequently has its onset during the early years, these data indicate that this 16-year effort to improve asthma outcome in children has thus far been ineffective.

    We don’t follow directions. It’s not that we don’t know what to do. It’s that we don’t do it consistently.

    Another passion of mine is ADHD. Despite progress in our assessment, diagnosis, and treatment of children with ADHD, researchers have documented wide variations in clinicians’ practices for children suspected of having ADHD. A 1999 national survey of pediatricians and family physicians found that, although about 60% of doctors used formal diagnostic criteria, less than 30% used criteria consistent with the DSM-IV. Another study found that that less than 40% of children suspected of having ADHD were formally assessed with behavioral questionnaires, and DSM-IV criteria were used in less than 40% of cases.  Most importantly, despite strong empirical evidence documenting the benefits of medication use, wide variation was seen in the amount of medications prescribed.

    It’s not that we don’t know what to do. It’s that we don’t do it consistently.

    Guidelines are meant to decrease practice variation and ensure that assessments and treatments are based on the best available empirical evidence.  Unfortunately, mere publication of practice guidelines does not ensure physician use of or adherence to them.  The survey mentioned above found that 14% of pediatricians and 5% of family physicians in this survey reported adopting a practice guideline for ADHD.

    Physicians are often resistant to guidelines because they make them feel unnecessary. They don’t talk about it, but some fear that if a guideline can tell you how to care for a child, you might no longer need a doctor.

    We need guidelines, though. We need doctors to learn to practice more consistently, and to use evidence and data. We need them to focus more on implementing what we already know, than in worrying about what’s around the corner.

    It’s not as sexy nor as exciting as what you might see on TV, but it’s far more likely to do more for more people than almost anything else you can think of.

    We don’t just need fewer cowboys and more pit crews. We need fewer artists and more empiricists.


    • Dr. Carroll:

      Would you say this is one of the reasons that those who practice in the VA system are so successful at what they do? In this case, the US Government wants to take care of some Very Important People, many of whom can’t afford “bleeding edge” solutions. So they provide a fair amount of Guidelines…

      Are there any “Lessons Learned” papers written about the VA system that should be on the reading list and absorbed more willingly by those seeking solutions for today’s problems? I’m not saying The Solution has to be socialized medicine as is the case for the VA. But we are mortal fools to walk away from Things That Work (if that is indeed what we are doing from time to time).


    • Whenever I read something like this I think of medical licencing. If it was easier to become a doctor could companies hire them like technicians and use systems making the system more important than the Doctor. Why is a doctor supposed to know so much, could not engineers design a system that works with less trained people?

    • @Floccina-No. People are not cars.

      @Aaron- I have heavily pushed guideliens in our group. It has met a lot of resistance, as have checklists, but people are mostly following them. It has been my belief that a good, well read doc will not be afraid of guidelines. They will give you the right answer the large majority of the time. If you know your stuff, you should know when you need to leave the guidelines and be able to explain why you are doing so.


      • But doctors pretty much aren’t well read. It doesn’t pay the bills. Otherwise why would they presribe brand name drugs in place of generics, suggest procedures that are ineffective, over test, etc.? Unless you want to say that they aren’t good. If that’s the case then we have a much larger problem.

        I think that Floccina has a point. Doctors are highly skilled, educated and paid trades people. Medical school is a trade school. They need to be viewed as part of a system. Doctors, like patients, aren’t nearly as unique and special as they think they are.

        • Some are not well read, but the problem is bigger. The problem lies in trying to transfer large studies to individual patients. At that level, docs are often more influenced by their own experience, or the anecdotal evidence of others. It is also difficult for some docs to tell patients they have no answer for their problem.


      • Guidelines are great when they stay as guides and do not become mandates–or even a carrot or a stick.

    • I am not against guidelines, but I am opposed to trivializing the doctor- patient relationship. Maybe for consultants, technicians are desirable. However, if you want care over time as your health and symptoms evolve, better to stay with a doctor you know and trust or at least a small group of docs that know you well. If an asthmatic has been to urgent care, ER, a few docs, maybe even a specialist and cannot remember the drugs he’s been prescribed or the number of nights he coughs per month, he needs continuity more than any guideline. All guidelines assume compliant patients in an ideal world. They are the starting point.