A few weeks ago, a healthy patient in her early 40s came to my office requesting a CT scan of the chest. She was a smoker and wanted to be sure that she didn't have lung cancer.

I explained to her that a CT, or computerized tomography, scan was not a good test for lung cancer. I even pulled up the U.S. Preventive Services Task Force guidelines on my computer to show her that there was no compelling evidence to support screening.

She refused to accept this explanation. "I want to know that I don't have cancer," she repeated. I told her that a negative CT scan could not give her that reassurance, but this issue of false negatives was ignored. I then explained about a CT's high dose of radiation and its dye's risks of kidney failure, but the dangers did not sway her.

We went back and forth for the better part of an hour, at which point I was getting worn down. I finally agreed to order a chest X-ray. She didn't need a chest X-ray. There was no scientific evidence that a chest X-ray was a beneficial screening test in asymptomatic people, but I decided that the harms and costs of an X-ray outweighed the harms of a continued fight with this patient.

When I saw her the following week and told her about the clean X-ray, she was elated. "I feel so much better," she said, her voice earnest and engaging. "Now I can go on, knowing that I don't have cancer."

I repeated my entreaty that this was not a reassurance, that she could still have cancer despite the negative X-ray. My words floated past her like vapor. This conviction of good health based on negative screening tests is a common reaction, but not one grounded in reality. Screening tests have inherent and significant limitations. Cancers are not straightforward.

But people extract a psychological reassurance from screening tests. We want, and need, to believe these tests protect us. Doing so gives us a sense of control, even if it's illusory. This disconnect is what is at the root of the current commotion over mammography.

When I first read of the new mammography screening recommendations offered by the U.S. Preventive Services Task Force, I didn't think they seemed so radical. I was completely taken aback by the fierce storm of controversy that arose.

Whenever there are such heated emotions regarding a medical recommendation, it's a safe bet that there's a lot going on beyond the actual science. (And when politicians start to weigh in, you can be sure that the hard data have been left far behind.)

The group that wrote the report is highly respected by doctors precisely because it focuses only on the data. Because there is no pharmaceutical money involved, no patents or research grants at stake, no promotions or tenure to be considered, the task force is able to concentrate strictly on the science, nitpicking through the reams of numbers, applying the hard-nosed analytical skills we all wish we possessed.

What the task force actually did was review the longer-term follow-up on some of the major mammography trials. Its members slogged through the data to analyze whether mammograms were saving lives (a "hard" clinical end-point) and weighed this benefit against the risks.

The conclusion - that women in their 40s don't derive much benefit from mammograms - shouldn't have surprised anyone. The data were never very strong to begin with, but the idea of the annual mammogram from age 40 on took hold in our society, and now no one wants to give it up.

The question is: Why are we so afraid of the data? It should be obvious that any medical procedure that offers more harm than good should be abandoned.

But cancer screening tests have crystallized in our society as one of the positive things we can do to prevent disease. They have taken on an almost saintly image in a way that vaccines (which actually do prevent disease) have not. The idea that mammograms - or prostate-specific antigen testing - may not save lives is nearly impossible to conceptualize. The suggestion that they may cause harm seems blasphemous.

Don't get me wrong: I'm not against screening. As a primary-care internist, screening tests are my stock in trade. Screening tests have offered many benefits, but that doesn't mean we should be afraid to openly consider the harms.

As we physicians help interpret the task force analysis for our patients, we need to consider the emotional reactions of our patients. A scientific explanation of multivariate analysis is not necessarily helpful when a person is terrified of a disease.

However, we cannot fall into the trap of offering absolute reassurances. We need to help our patients (and our politicians) understand that uncertainty is a given. We can modify risk, but we can't remove it. Screening tests are no more perfect than any other medical intervention.

Doctors shouldn't administer antibiotics when we think it would be harmful. We shouldn't perform heart transplants, colonoscopies or mammograms on patients in whom they might be harmful. We have to consider individual risk and prescribe these interventions only to the appropriate patients. This is called clinical analysis. Politicians who call it rationing ought to be ashamed of themselves. The more thoughtfully we examine the data, ultimately the better off we all are.

But we also have to take into account our emotional attachments to screening tests, even if they can be irrational at times.

Ofri is an internist in New York City, editor of the Bellevue Literary Review and author of "Medicine in Translation: Journeys With My Patients."