As an advocate of patient-centered care, I have to recognize that some varieties of patient-centrism make me more comfortable than others. If I really want the patient to do X, and the patient doesn’t want to, I generally feel okay about that. Frustrated, sure, and often times convinced that my way is the right way and the patient’s way is some sort of detour. Most often, though, I am able to put aside those feelings and encourage the patient to make their own decision. I am a less-is-more kind of guy, after all. Plus, encouraging the patient’s decision making can leave behind the pleasing glow of low-grade self-righteousness.

But sometimes things go the other way. You have garden-variety back pain, I tell Ms. X, without any alarm signs that might indicate infection or cancer. Yet she still wants the scan. Or say you come to me and want to check your “basic labs,” to make sure “everything is okay with your blood.”

Neither of these is necessary. MRIs for garden-variety back pain, as we have discussed before in this space, can lead to real harm, as can blood tests for no good reason. And this, the asking for things that I don’t think have a point, is very difficult for me to deal with. I think this because it goes against the grain of my personality (I would rather do less, and avoid iatrogenic harm, than do more and cause it) and such requests reinforce a real, justified expectation that we all have when we go the doctor: we should leave with something, even if it is just a prescription slip. It’s as if the X-ray order or the antibiotic we leave the doctor’s office with is an objective correlative for the care they are supposed to give us.

When the patient asks for something that might harm them without discernible use to me, I try to explain my views – and then people, being who they are, sometimes want to make that same decision. Which leaves me with a range of unpalatable options. I can say no; I can say yes; or I can temporize. Saying no makes me feel good and dissatisfies the patient. Often, if the contraindicated care is provided by a broad range of providers anyway (e.g., antibiotics for viral upper respiratory infections), my refusal does no good. If I say yes I have implicated myself in the patient’s potential self-harm.

Part of a solution – apart from me biting the bullet and saying no more often – is to change expectation, so that leaving the visit “with something in hand” can be not just a script, or a procedure, but a plan of action or a symptom diary to be filled out. Until then, disagreement will still get my hackles up, especially if the patient wants something I don’t.