Maybe I had convinced myself of something that was not true. Maybe I had smooth-talked the neurology consultant into agreeing with my findings.

These plaguing questions unanswered, I had to take action in order to muffle the crescendo of lingering doubts. So, I picked up the phone and I called the patient at home (using the translator phone, no less; I don’t speak Mandarin).

I had to know how my patient was feeling. To my relief, he was better but the vertigo had not fully resolved. A few days later I called again to check in. He was feeling much better. A week later, I called yet again. Finally, he said, “Thank you, but you really don’t have to call me any more.”

One year later, do I still believe what I wrote in these pages about practicing evidence-based medicine? Yes.

But I have learned that this style of medicine requires me to follow up with my patients in a way that practicing defensive medicine would normally not.

I truly believe sending a low-risk chest pain patient home (without elevated cardiac markers, no concerning ECG findings, and excellent follow-up plans) is better for the patient than ordering further invasive testing, especially if the patient agrees with the plan and understands the facts. This is in part because I have engaged in the literature and in part because I’ve seen the negative consequences of over-testing.

It is real and can be devastating. But now, when I make evidence-based decisions to avoid defensive medicine, a key part of my plan is to actively follow up with the patient within a couple of days by phone. I tell patients that I might call them and ask them to return for further testing if symptoms have not improved or have changed or worsened.

I also assess the likelihood of the patient following through with this plan before the discharge, as I do not want to lose patients to follow-up. While this kind of longitudinal care is not standard among many emergency physicians, I believe it is a small price to pay in exchange for the ability to avoid reflexively pursuing invasive, potentially harmful, and costly work-ups in every patient presenting with certain alarming chief complaints.

So, a year later I can happily say “Yes,” to the extent that it’s appropriate, I am doing my best to walk the walk of evidence-based medicine.

But when I do, it means I may have to pick up that phone a day or two later, call the patient, and, in a slightly different way, talk the talk.