Cancer cells are seen on a large screen connected to a microscope at the CeBit computer fair in Hanover, March, 6, 2012Fabian Bimmer / Reuters

If you had a disease, and you could find out sooner rather than later, why wouldn’t you?

Medicine has long focused on early detection of diseases as part of a move toward preventive care. But imperfect tests, false positives, and overdiagnosis mean that sometimes the tests do more harm than good, and in recent years, there have been more recommendations to reduce some kinds of screening, including pap smears, colonoscopies, mammograms, and even annual pelvic exams.

“This is something we all need to understand, the two sides of early detection. It does help people, but it’s almost guaranteed to harm others,” said H. Gilbert Welch, a professor of medicine, public policy, and business administration at Dartmouth College, and author of the book Should I Be Tested for Cancer? (He reveals his answer in the book’s subtitle: “Maybe not.”)

The more you look for disease, the more you find it. And in the case of cancer, it’s hard for doctors to know if what they find is dangerous and needs to be addressed, or if it’s just a small tumor that won’t grow and poses no threat. “We can’t be sure which is which, so we treat everybody,” Welch explained at the Aspen Ideas Festival’s Spotlight Health session. “That means we’re treating people who will never experience problems from their disease.”

But they may experience problems from the treatment.

The panel gave the example of prostate cancer, which is very common in men—one in seven American men will be diagnosed with it in their lifetimes. “But it turns out a lot of these cancers are very indolent,” said Jessica Herzstein, a preventive-medicine consultant and member of the U.S. Preventive Services Task Force. Around 30 to 40 percent of men who’ve been treated for prostate cancer likely had “slow-growing tumors that would never have become a threat to the man’s lifespan or health,” according to the Prostate Cancer Foundation.

In other words, “you’re going to die with them, not of them,” Herzstein said, “and the treatments are very very harmful.” Radiation therapy, for example, can cause incontinence and erectile dysfunction, and hormone therapy can cause osteoporosis and depression.

The possibility of a false positive is another downside. Not only could it lead to more invasive follow-up tests or treatments that aren’t needed, but it can also give patients unnecessary anxiety.

“If we resolve the test by saying ‘The test was wrong, you’re fine!’, that’s one thing,” Welch said. “But most false alarms aren’t resolved that way. [It’s more like] ‘You don’t have cancer, but you have some abnormality that possibly puts you at a higher risk for cancer, but we’re not going to do anything about it. I think that’s where there can be [mental] harm.”

Ultimately, it comes down to a weighing of the benefits and the harms, and, in the absence of clear evidence, the preferences of the patient. The U.S. Preventive Services Task Force helps identify which tests are beneficial by evaluating and grading them. It gives tests an A if there’s a high certainty of substantial benefit, a B if there’s moderate certainty of substantial benefit, a C if there’s moderate certainty of a small benefit, a D if there’s moderate or high certainty of no benefit, and an I if the evidence is just too insufficient to say.

The task force gave prostate cancer screening a D. HIV screening got an A. For breast cancer screening, an always-controversial topic, the results vary. Breast self-exams got a D. Mammograms got a B, but only for women between 50 and 74 years old. For women in their 40s, the grade is a C, meaning the task force recommends patients and physicians discuss and decide together.

Before getting a screening test, patients should think about what would happen if they get a positive result, and if they’d be ready for it, Welch advised.

“If I were to go through this, and have this diagnosis, would I want to have this surgery?” Herzstein asked, posing a hypothetical. Would you want to undergo the biopsy, the chemo, whatever treatments come next? “Maybe you don’t even want to go there if there is no treatment for the disease,” she added.

Welch gives an example. “With Alzheimer’s disease that’s a fundamental question: What are you going to do with a positive result?” he asked. “What’s the downstream plan? It’s easy to screen a population, it’s hard to know what the right thing to do is.”

We want to hear what you think about this article. Submit a letter to the editor or write to letters@theatlantic.com.