A few years ago, my mother called with an urgent request. She was agitated. She had just come from a meeting of her Jewish women’s group, where she had learned about BRCA1 and BRCA2, the so-called breast cancer genes. She wanted me to get tested. She insisted. I immediately said no.

BRCA1 and 2 are some of the most powerful cancer markers scientists have discovered—archetypes of an ever expanding pool of potentially lifesaving genetic tests. A mutation in one of those genes more than quadruples your risk of breast cancer to between 45 and 65 percent. For ovarian cancer, the risk hits 10 to 39 percent. Treatment is ruthless but effective: Cut away your breasts and ovaries to cut down your risk.

When I got that call from my mom, though, I shut her down. Because while BRCA is a powerful tool to estimate risk, it’s not a crystal ball. As with every mutation in our DNA, its effects vary depending on where exactly the gene is broken and on a family’s cancer history. It’s never “You will get cancer.” As a science journalist, I long ago internalized that uncertainty. And I didn’t want the results of one inscrutable test to propel me into irreversible surgery.

It took me more than a year to step back from that gut reaction. The science of genetics is constantly evolving, and I wasn’t sure exactly where I stood—and if I wasn’t sure, then what did that mean for the millions of people getting results from other, even more inscrutable tests?

I decided to dig into the science.

My mom’s group had good reason to talk about breast cancer: Ashkenazi Jews like us have a 2 to 3 percent chance of inheriting a BRCA mutation, 10 times higher than average. My mom also made sure to point out that two of my paternal relatives—Great-aunt Ann and my dad’s cousin Fran—died of breast cancer.

I came armed with that information to the National Cancer Institute in Maryland, where senior researcher Mark Greene plugged it into an online BRCA risk-assessment tool—a first step for any woman wondering whether to get the genetic test. After building a family tree, a web of cancers and no cancers, the program calculated what I expected: My risk of having the mutation is small. “Basically, you have two distant relatives with cancer, only one of whom is a young case,” said Greene. “That’s not the kind of pattern we typically associate with mutations in BRCA1 or 2.” The data wasn’t screaming at me to get tested.

I left Greene’s office feeling relieved. Still, low risk isn’t no risk. When I talked to a half-dozen researchers and genetic counselors, nearly every one of them thought I should play it safe and get the test.

So let’s say I did. If I tested positive, the statistics would push me toward removing my breasts and ovaries. Data suggests that surgery can increase survival in people with confirmed BRCA mutations. Removing my ovaries would reduce my ovarian cancer risk by as much as 90 percent.

There’s a problem with those rates, though: They’re largely based on families with a history of breast and ovarian cancer. The rates might not apply to other women. Ann and Fran died of breast cancer—but I don’t know if it was BRCA-linked.

Every BRCA-positive woman has to weigh the strength of that survival data against the repercussions of surgery. I wouldn’t be able to have children. I’d lose sensation in my breasts—a crucial part of my sexual enjoyment—and possibly strength in my arms. Oophorectomy also kick-starts menopause: That can mean early bone weakening, cardiovascular disease, even dementia.

Not worth it for many women. Plan B isn’t bad: twice-yearly breast scans, which can prevent some breast cancer deaths in BRCA-positive patients.

But here’s where things get truly tricky—and where the initial choice to test becomes fraught. Frequent mammograms and MRIs usually find something, even if the shadows on the screen may never become life-threatening. But patients tend to pounce on those shadows. After years of tests and biopsies, some women give in and remove their breasts.

The data backs that up: Humans just can’t manage fear. Barry Kramer, director of the NCI’s division on cancer prevention, explains that mastectomies have increased steeply in the past few decades as women find early signs of cancer. “But there’s pretty strong evidence that there’s no difference in overall survival,” Kramer says. Patients are petrified. Then they take drastic, potentially harmful—and sometimes unnecessary—action.

Genetic tests could mislead patients in the same way. The torrent of information buried in the tightly packed twists of our chromosomes will become more and more powerful, but it will never provide simple answers.

It will always be monstrously hard to calculate cancer risk. It’s even harder to turn that risk into helpful preventive care while facing the fear of death. I spoke to a woman who decided, after much agonizing, to take the BRCA test and then remove her breasts and ovaries. She didn’t regret it.

But her decision isn’t mine. Knowing what I do about my family’s history—after Ann and Fran, there’s no more breast or ovarian cancer in my family—a positive test wouldn’t convince me to get surgery. It would just feel like hefting a sword, dangling it by a frayed thread above my head, and waiting for it to fall. I’m more scared of living like that than I am of cancer.