I still remember being taken aback by how young she was. “She” was Mary — a 28-year-old woman who had completed chemotherapy for stage II breast cancer. She was treated elsewhere and had moved cities when her husband got a promotion.

“I’m still getting used to this area, but I am happy my hair came back before we had to move. I can’t imagine trying to set up a new house and have to go through my cancer history with strangers. I just don’t want to be that young mother with cancer.”

We reviewed her medical history, the details of her pathology, her prior treatment and current endocrine therapy.

“How are you doing on tamoxifen?” I asked.

“Fine. I think it agrees with me, fortunately!” she said.

We discussed surveillance after treatment, the frequency of follow-up, and what that should entail, and I told her:

“Based on what we know, checking your labs or doing imaging in the absence of symptoms or a reason don’t help us help you. I don’t recommend testing if you feel well otherwise. How do you feel about that?”

All of a sudden, she appeared uncomfortable:

“Well, my doctors gave me the same advice, but I am not comfortable with surveillance. I don’t want to not get any tests.”

I tried to reassure her, tried to explain my rationale, and that it was based on ASCO guidelines. However, she stopped me before I could get any further.

“I know you are trying to help me, and I understand what you are saying. I have read the guidelines, done my reading. I’ve gone to every site that anyone has told me to read, and I get it — extensive testing doesn’t help women live longer after breast cancer,” she stated.

“But, I need to do more. My gut tells me I should do more. That’s why my prior doctors checked tumor markers for me every three months. That’s why I do a chest x-ray every six months. I can’t just stand by and do nothing. I just can’t.”

After speaking with multiple colleagues, I know this experience is not uncommon. I often wondered why patients took a course of action contrary to what best evidence told us.

On my way to work last week, I heard a story that might shed some light on this. As reported on NPR’s Morning Edition by Shakar Vedantam, a study by Brendan Nyhan and colleagues at Dartmouth evaluated public health messaging around vaccines and whether they helped to address parental attitudes (and reluctance in some) to vaccination. They randomly assigned almost 1,800 patients to four interventions aimed at highlighting the evidence underlying the risks of autism (or lack thereof) and benefits of the measles, mumps, and rubella (MMR) vaccination, and the dangers posed to children who are not vaccinated.

At the end of the study, they found that while interventions reduced misperceptions about the MMR and its link to autism, parents who expressed the least favorable views on vaccination were no more likely to vaccinate. As summarized by Mr. Vedantam, while one would hope that attitudes are shaped by facts, it might be that “beliefs about facts are shaped by attitudes.”

So, what does this study on vaccines have to do with my patient? It suggests that if we are to improve the quality of cancer care, it is not enough to present evidence or even guidelines when we talk to our patients (or to our colleagues for that matter). Instead, as Nyham et al. emphasize in their study, we must start by “measuring beliefs and behavioral intents when assessing health interventions,” and I agree. Only by understanding the ground on which our patients stand can we begin to walk that path with them, and perhaps, point them towards the road perhaps best traveled.

Until then, I will listen to my patient, her values and preferences. Knowing that she has taken best evidence under consideration, I will do what I did here: Proceed with the surveillance testing that both she and I were comfortable with.

Don S. Dizon is an oncologist who blogs at ASCO Connection, where this post originally appeared.