Map of life expectancy at birth from Global Education Project.

Friday, October 03, 2014

Oh yeah, about the conference . . .

I'm not going to bore you with any of the specific presentations (okay, except for mine) but I will try to give a sense of the overall state of research into communication in health care. Yes, this is a scientific field and we do empirical research, as rigorously as our usually quite limited funding will allow. But it's impossible to separate facts from values.

The goal of health care is to make money improve patients' well being. But that presumably means different things to different people. Do you want to live as long as possible or enjoy yourself more? Do you want to take a chance on surgery that will probably fix your arthritic hips and let you get back out on the golf course, but might just possibly go wrong and leave you worse off? And so on, including many even more complicated and stress inducing situations.

Furthermore, your doctor might think that you will unambiguously be better off if you take some course of action, but for various reasons of your own you don't do it, one possibility being that if you take pills you will have to think of yourself as sick but you don't want to be. Is it the doctor's job to use our research to manipulate you into doing what the doctor thinks is best? To take a more concrete example, I feel that my orthopedic surgeon somewhat downplayed the pain and disability I would endure after thumb arthroplasty, but I'm not mad at him. I now feel that it was worth it in the long run but I might not have gone through with it if he'd been totally honest. Of course he also had the ulterior motive that he gets paid to do surgery.

The fact is that most patients will never understand the biomedical reasoning behind treatment decisions, and even if you strip the decision down to relative probabilities of understandable outcomes, most people don't really grasp the concepts of mathematical probability and they have all sorts of cognitive biases. A 90% chance of a good outcome and a 10% chance of a bad outcome are evaluated very differently, even though they are of course identical.

So, the value now is "shared decision making," which even has a standard presentation as SDM. The doctor is supposed to be the expert the probabilities of various outcomes of various courses of action, and the patient is supposed to be the expert on what she values the most among that landscape of probabilities and so we decide. But that is merely aspirational, Nobody knows how to accomplish it consistently in the real world in a meaningful way.

It's only over the last 20 or 30 years that we have even had this aspiration as the standard in medicine. It used to be that doctor knew best and we were supposed to follow doctor's orders. As you can imagine, it is a very complex problem to establish the relationship between features of clinical communication and outcomes because, pretty much by definition, what constitutes a good outcome is undefined going in. We can arbitrarily say it's life expectancy, or quality adjusted life years, or some specific reduction in some pre-specified symptom. But if we don't get there because the patient chose a course that didn't maximize that outcome, is that a failure? And asking people to evaluate decisions retrospectively is not very helpful; we have powerful cognitive biases in doing that.

So, we have to start by trying to measure what people understand after a clinical encounter, whether they feel their questions were answered, and how confident they feel in their ability to make choices that affect their health. Then, if we can define a patient centered good outcome, we can eventually find out if it's achieved --but we might have to follow them for years, and that costs a lot of money, and they aren't giving it to us, yet, for this kind of research.

I'll tell you exactly what I'm doing anon.


robin andrea said...

It would be interesting to know how doctors communicate based on their specialties. It seems that surgeons might have more at stake for a patient to follow a particular protocol than a family practitioner who wants a patient to take medication (or lose weight or start exercising, etc) with no financial incentive. Can you imagine a doctor/patient first appointment scenario where the patient is asked what his/her health goals are? Doctors and patients are strangers to each other, and despite the intimacy involved, remain so. It's an interesting dynamic.

Cervantes said...

Yes. In the case of prostate cancer, for example, radiologists recommend radiation, surgeons recommend surgery, and oncologists recommend chemo.